Securitized Products

Highlights Politics will inject further volatility into risk assets, but stocks will outperform bonds and cash on a 6-12 month horizon. The health of the economy and earnings matter more than Trump's political woes for investors and the Fed. The consumer - a key driver of the U.S. economy and corporate earnings - will provide a solid backdrop for the economy through 2017 and beyond. The combination of balance sheet shrinkage and Fed rate hikes will lead to higher bond yields than are currently discounted in the market. It is a different story for the mortgage market, where spreads will be biased to widen during Fed runoff. Feature The Economy Matters More Than Politics The health of the economy and earnings matter more than Trump's political woes for investors and the Fed. When the economy and earnings backdrop was favorable during presidential scandals in the 1920s and the 1990s, the equity markets performed well. In the early 70s, amid soaring inflation and the worst recession since the Great Depression, there was a bear market in equities (Chart 1). Today, the backdrop for the economy and earnings - while not as robust as in the 1920s or late 1990s - provides support for higher stock prices, two more Fed rate hikes and higher Treasury bond yields. Trump's political woes may slow, but not completely halt the GOP's legislative agenda1. Support for Trump among his GOP base remains high at 85%, making impeachment a long shot until after the November 2018 mid-term elections (Chart 2). If the Democrats take the House, they are likely to impeach Trump in 2019. For the Trump and the Republicans in Congress, this means the impetus is even greater to make progress now on tax cuts, tax reform and infrastructure. However, the embattled White House will slow the process as the president's staff often acts as a coordinator among the various factions in Congress. With Trump's team preoccupied with political woes, they will not be effective in this role. Chart 1Economy Will Trump Politics ##br## For Financial Markets Economy Will Trump Politics For Financial Markets Economy Will Trump Politics For Financial Markets Chart 2GOP Base Not Yet Willing To ##br## Impeach Trump The Economy Trumps Politics The Economy Trumps Politics The Fed will look through the politics and focus on the health of the economy and will continue to raise rates gradually this year, with the next hike coming in June. Financial conditions have eased since the Fed's 25 basis point rate hike in December, and that alone should be enough to keep the Fed on track to tighten next month. As we have noted in recent reports, even without fiscal stimulus, the U.S. economy will still grow near its long-term potential, tighten the labor market and push up wages and inflation. The Fed has been reticent to include any impact from fiscal stimulus into their policy deliberations thus far. The minutes of the March FOMC meeting noted that "members continued to judge that there was significant uncertainty about the effects of possible changes in fiscal and other government policies". Bottom Line: The lack of progress on legislation may result in a pullback in U.S. equity prices, but absent a material weakening of the U.S. economy or profit picture, the pullback will not turn into a bear market. Checking In On The Consumer The consumer - a key driver of the U.S. economy and corporate earnings - will provide a solid backdrop for the economy through 2017 and beyond. This backdrop will allow the Fed to pursue two rate hikes this year. The weakness in several indicators has worried some investors that the economy may be on the verge of a slowdown or even a collapse. However, a firming economy should sustain corporate earnings growth and, ultimately, higher stock prices. Consumer spending's share of GDP is 68% and increasing (Chart 3). GDP growth excluding consumer spending is more volatile than overall GDP growth. The household sector has contributed 75% to growth since the end of the recession, which is the best performance of any sector. The key drivers of spending point to further gains in the sector, and the imbalances that were present ahead of prior downturns are not evident today. Chart 3Household Share Of GDP Is At An All Time High And Rising Household Share Of GDP Is At An All Time High And Rising Household Share Of GDP Is At An All Time High And Rising Chart 4Consumer Spending Remains In An Uptrend Consumer Spending Remains In An Uptrend Consumer Spending Remains In An Uptrend Household spending growth has softened but remains in an uptrend. Broad measures of consumer spending tend to peak two to four years prior to the start of a recession. The lead time is even longer in a long-cycle expansion.2 Investors should not dismiss the weakness altogether, but position portfolios for the late-cycle environment. Personal consumption expenditure growth peaked at 4% year-over-year in Q1 2015. Auto sales, a timelier measure of spending although not as comprehensive, peaked in December 2016 (Chart 4). Applying the 2 to 4 year lead time noted above - and making the assumption that spending has indeed peaked - this points to a recession commencing in the middle of 2019 at the earliest. Household net worth is at an all-time high, and the overall wealth effect on consumer spending has been positive for some time. Our forecast for financial markets and the housing market, though modest, imply that the positive wealth effect will continue. Debt-financed spending remains a viable option for consumers, which was not the case in late 2007 before the onset of the recession. Banks have not changed their lending standards for most consumer loans and demand for these loans will stay solid despite the Fed rate increases that we expect. The Bank Credit Analyst's March 2017 report showed that even a 100-basis point rate rise from the current levels would not lift the interest payments to burdensome levels by historical standards. Incomes will continue to climb and importantly, consumer income expectations have also hit new highs. With the economy at the Fed's assessment of full employment, wage growth is accelerating, albeit more modestly than in previous recoveries. Our recent report3 found that wages tend to rise about two years after the output gap has formed a bottom. A narrowing output gap leads to a tighter labor market and higher incomes. As measured by the quit rate, job security is at a fresh cycle high (not shown). Many consumer indicators are in better shape today than they were in 2007 or at similar points in the other long cycles4 (Charts 5 and 6). We define the long cycle economic expansions as those lasting 8-10 years. The two expansions that meet the definition are 1981-1990 and 1992-2001.5 Consumer spending is running in line with incomes, unlike in the mid-2000s. Chart 5Key Consumer Metrics ##br## Remain Favorable Key Consumer Metrics Remain Favorable Key Consumer Metrics Remain Favorable Chart 6There Is Still Plenty Of Support ##br## For Solid Consumer Spending There Is Still Plenty Of Support For Solid Consumer Spending There Is Still Plenty Of Support For Solid Consumer Spending Mortgage equity withdrawal, a crucial source of debt-fueled consumer spending prior to 2007, has been non-existent in this cycle. Spending on essentials are close to all-time lows. In 2007 they were at record highs and had moved up dramatically in the prior half-decade amid escalating debt levels, rising energy prices and consumer interest rates. We are concerned by the historically high percentage of household incomes (17%) dedicated to medical care. An aging population, ever rising healthcare costs and uncertainty surrounding the future of Obamacare may drive medical spending even higher. Household debt levels as a percentage of disposable income peaked in 2008 at over 120%, but are back under 100%, i.e. at the level that existed prior to the 2007-2009 recession. The level of household debt compares favorably to similar points in the long cycles of the 1980s and 1990s. Financial obligations are at multi-decade lows (Chart 6, bottom panel). Bottom Line: The fundamentals supporting consumer spending remain solid. A healthy consumer means the economy can meet the Fed's modest GDP forecast for 2017, keeping the central bank on track to tighten twice more in 2017. This outlook supports our view for stocks over bonds in the next 6-12 months. The Fed's Balance Sheet: It's Diet Time Chart 7Fed Set To Begin Tapering In Early 2018 Fed Set To Begin Tapering In Early 2018 Fed Set To Begin Tapering In Early 2018 The minutes from the March FOMC meeting indicated that a change in the Fed's reinvestment policy will likely be appropriate "later this year". The minutes suggested that the FOMC is split on whether to simply terminate all reinvestment for both Treasurys and MBS, or to "taper" reinvestment over time. Our base case is that the Fed will follow up a June rate hike with another one in September, at which point policymakers will provide some details on their plans for balance sheet runoff to begin in January of 2018. Investors are rightly concerned about the potential impact of the runoff, especially given that memories of the 2013 "taper tantrum" are still fresh. There is disagreement among academics about whether quantitative easing (QE) directly depressed bond yields by restricting the supply of high-quality fixed income assets, or whether the impact on yields was solely via the "signaling effect" (i.e. that QE implied that short-rates will be held at a low level for a very long time). Either way, balance sheet runoff will likely have some impact on bond yields. A good starting point is to employ an empirical estimate of the impact of QE. The IMF has modeled long-term Treasury yields based on a number of economic and financial variables, including inflation expectations, demographics, growth, current accounts and budget balances. The model also includes the stock of assets held by the Fed as a share of GDP. If the Fed were to begin running off its holdings of both Treasurys and MBS at the beginning of 2018 by terminating all reinvestment, then the amount of bank reserves held at the Fed would likely evaporate by 2021. This represents a fall of roughly 10 percentage points of GDP (Chart 7). Given the IMF interest rate model's coefficient of -0.9, it implies that long-term Treasury yields and mortgage rates would rise by 90 basis points from the "portfolio balance" effect alone. However, it is more complicated than that. The impact on yields is likely to be tempered by three factors: The Fed may opt to avoid going "cold turkey" on reinvestment, choosing instead to scale back gradually. Fed President William Dudley recently commented that the Fed wants balance sheet reduction to "run in the background", such that it is not a major event for markets. Some academic experts are recommending that the Fed maintain a fairly large balance sheet by historical standards because of the need in financial markets for short-term, risk-free assets that would diminish if there are fewer excess bank reserves available. Banks, for example, are required by regulators to hold more high-quality assets than they did in the pre-Lehman years. The implication is that the balance sheet may never fully revert to historic norms relative to GDP. As the FOMC dials back monetary stimulus it will be concerned with overall monetary conditions, including short-term rates, long-term rates and the dollar. If long-term rates and/or the dollar rise too quickly, policymakers will moderate the pace of rate hikes and use forward guidance to talk down the long end of the curve so as to avoid allowing financial conditions get too tight, too quickly (i.e. the term premium would rise, but would be partly offset by a lower expected path for the fed funds rate). Thus, the path of short-term rates is dependent on the dollar and the reaction of the long end of the curve. It is difficult to estimate how it will shake out, but a recent report from the Federal Reserve Bank of Kansas City estimated that a $675 billion reduction in the size of the Fed's balance sheet is equivalent to a 25 basis point increase in the fed funds rate (although the authors admit that the confidence band around this estimate is extremely wide).6 We expect that the impact of runoff alone will be much less than the 90 basis point estimate discussed above. Still, the combination of balance sheet shrinkage and Fed rate hikes will lead to higher bond yields than is currently discounted in the market. We could also see some upward pressure on global term premia when the ECB announces the next tapering of its QE purchase program, possibly this autumn. However, it will be years before the ECB will be in a position to reduce the size of its balance sheet. As for the Bank of Japan, we doubt that the central bank will ever shed its JGB holdings. What about the shape of the Treasury curve? Our fixed-income strategists believe that the shape of the curve will be determined by the normal cyclical dynamics we have seen in the past. We are still in a window in which the Treasury curve will steepen as yields rise. A little later in the Fed cycle, the curve will bear-flatten as the long-end begins to rise at a slower pace than the front end. We do not see balance sheet adjustment as changing these dynamics much. Similarly, with respect to credit spreads, the state of nonfinancial corporate sector balance sheets and the overall stance of monetary policy will continue to be the main drivers of the credit cycle. If unwinding the balance sheet leads to a premature tightening of financial conditions, then the Fed will proceed more slowly on rate hikes. The crucial indicator to watch is core PCE inflation. Credit spreads will remain fairly well contained until core PCE inflation reaches the Fed's 2% target. At that point, the pace of monetary normalization will ramp up, putting spreads at risk of widening. It is a different story for the mortgage market, where spreads will be biased to widen during Fed runoff. While spreads have already widened a bit, in our view they still do not adequately compensate for the additional MBS supply that will hit the market when the Fed takes a step back. Historically, there is a reasonably tight correlation between MBS spreads and the spread between mortgage rates and Treasury yields (Chart 8). Thus, it is reasonable to expect mortgage rates to rise by more than Treasury yields. Chart 8MBS Spreads Set To Widen As Fed Tapers MBS Spreads Set To Widen As Fed Tapers MBS Spreads Set To Widen As Fed Tapers While the Fed's balance sheet reduction by itself may not have a big impact on the dollar, we still believe the currency has more upside because of the divergence in the overall monetary policy stance between the U.S. on one side and the ECB and Bank of Japan (BoJ) on the other. The BoJ will hold the 10-year JGB near to zero for quite some time. The ECB will also not be in a position to tighten for a long time, outside of removing negative short rates and tapering QE purchases a bit further in 2018. Meanwhile, we think the Fed will tighten by more than is currently discounted. Admittedly, the economic data have disappointed so far in 2017 and CPI inflation has softened which, at the margin, would cause some FOMC members to back away from rate hikes. Nonetheless, policymakers are focused more on the labor market than GDP to gauge the health of the expansion and the amount of economic slack. Despite the dismal Q1 GDP figures, following unimpressive growth in 2016, the unemployment rate has already fallen below what the FOMC expected the rate will be at the end of this year! A tightening labor market means that the economy is still growing above a trend pace. Unless there is a clear deceleration in wage growth as measured by the ECI or the Productivity and Cost report, the FOMC will likely hike rates by more than the 38 basis points currently discounted over the next 12 months. John Canally, CFA, Senior Vice President U.S. Investment Strategy johnc@bcaresearch.com Mark McClellan, Senior Vice President The Bank Credit Analyst markm@bcaresearch.com 1 See Geopolitical Strategy Report, "Break Glass In Case Of Impeachment," May 17, 2017. Available at gps.bcaresearch.com 2 See The Bank Credit Analyst, March 2017. Available at bca.bcaresearch.com 3 See U.S. Investment Strategy Weekly Report, "Still Awaiting the Next Pullback", May 15, 2017. Available at usis.bcaresearch.com 4 See The Bank Credit Analyst, March 2017. Available at bca.bcaresearch.com 5 We did not include the 1960s in this analysis because the Fed waited too long to tighten and allowed inflation to get out of hand. 6 Forecasting the Stance of Monetary Policy Under Balance Sheet Adjustments. The Macro Bulletin, Federal Reserve Bank of Kansas City. Troy Davig and A. Lee Smith. May 10, 2017.

Highlights Chart 1Rate Hikes Lagging Wage Growth Rate Hikes Lagging Wage Growth Rate Hikes Lagging Wage Growth Last Friday's GDP report showed that the U.S. economy grew a meagre 0.7% (annualized) in the first quarter of 2017, well below levels necessary to sustain an uptrend in inflation. However, our forward looking indicators still point to U.S. growth of around 2% during the next few quarters. It is likely that faulty seasonal adjustments suppressed Q1 GDP growth. Q1 growth has averaged -0.1% during the past 10 years, while Q2 growth has averaged more than 2%. Q2 growth has also exceeded Q1 growth in 8 of the last 10 years. For its part, the Bloomberg Barclays Treasury index has provided an average return of close to 1% during the past 10 Q1s and an average return of 0.4% during the past 10 Q2s. Treasury returns have been greater in the first quarter than in the second quarter in 6 out of the past 10 years. Investors would be wise to ignore Q1 GDP and stay focused on the uptrends in wage growth and inflation that are likely to persist (Chart 1). With the market priced for only 38 bps of rate hikes between now and the end of the year, there is scope for the Fed to send a hawkish surprise. Stay at below-benchmark duration and short January 2018 Fed Funds Futures. Feature Investment Grade: Overweight Chart 2Investment Grade Market Overview Investment Grade Market Overview Investment Grade Market Overview Investment grade corporate bonds outperformed the duration-equivalent Treasury index by 23 basis points in April. The index option-adjusted spread tightened 2 bps on the month and, at 116 bps, it remains well below its historical average (134 bps). While supportive monetary policy will ensure excess returns consistent with carry, investors should not bank on further spread compression as spreads have already discounted a substantial improvement in leverage (Chart 2). In a recent report,1 we noted that net leverage (defined as: total debt minus cash, as a percent of EBITD) is positively correlated with spreads, and also that it has never reversed its uptrend unless prompted by a recession. In other words, the corporate sector never voluntarily undertakes deleveraging, it only starts to pay down debt when forced by a severe economic contraction. We conclude that debt growth will likely continue to outpace profit growth (panel 4), even as profits rebound over the course of this year. If our anticipated timeline plays out, we will be looking to scale back on credit risk in 2018, when inflationary pressures are more pronounced and the Fed steps up the pace of tightening. Energy related sectors still appear cheap after adjusting for differences in credit rating and duration (Table 3). Further, our commodity strategists expect OPEC production cuts will be extended through to the end of the year, and that $60/bbl remains a reasonable target for oil prices. Table 3ACorporate Sector Relative Valuation And Recommended Allocation* Time Of The Season Time Of The Season Table 3BCorporate Sector Risk Vs. Reward* Time Of The Season Time Of The Season High-Yield: Overweight Chart 3High-Yield Market Overview High-Yield Market Overview High-Yield Market Overview High-Yield outperformed the duration-equivalent Treasury index by 62 basis points in April. The index option-adjusted spread tightened 12 bps on the month and, at 371 bps, it is currently 27 bps above its 2017-low. Wider junk spreads in recent months appear to be largely related to flight-to-safety flows driven by elevated global political uncertainty. We find it notable that spreads tightened following the market-friendly result of the first round of the French election. While political uncertainty remains, we view current spreads as attractive on a 6-12 month horizon. In a recent report,2 we tested a strategy of "buying dips" in the junk bond market and found that it produced favorable results in a low-inflation environment. With the St. Louis Fed's Price Pressures Measure still suggesting only a 6% chance of PCE inflation above 2.5% during the next 12 months, we think this strategy will continue to work. Moody's recorded 21 defaults in Q1 (globally) down from 41 in the first quarter of 2016, with the improvement attributable to recovery in the commodity sectors. While commodity sectors still accounted for half of the defaults in Q1, Moody's predicts that the retail sector will soon assume the mantle of "most troubled sector." According to Moody's, nearly 14% of retail issuers are trading at distressed levels. Moody's still expects the U.S. speculative grade default rate to be 3% for the next 12 months, down from 4.7% for the prior 12 months. Based on this forecast we calculate the High-Yield default-adjusted spread to be 207 bps (Chart 3), a level consistent with positive excess returns on a 12-month horizon more than 70% of the time. MBS: Underweight Chart 4MBS Market Overview MBS Market Overview MBS Market Overview Mortgage-Backed Securities outperformed the duration-equivalent Treasury index by 2 basis points in April. The conventional 30-year MBS yield fell 10 bps on the month, driven by an 11 bps decline in the rate component. The compensation for prepayment risk (option cost) rose by 2 bps, but this was partially offset by a 1 bp tightening in the option-adjusted spread (OAS). Since the middle of last year the MBS OAS has widened alongside rising net issuance, but this has been offset by a falling option cost (Chart 4). This is exactly the price behavior we would expect to see in an environment where mortgage rates are moving higher and the market is starting to discount the Fed's eventual exit from the MBS market. Higher mortgage rates suppress refinancings, and this will ensure that the option cost component of spreads remains low. However, higher mortgage rates are also unlikely to halt the uptrend in net MBS issuance, since the main constraint on housing demand this cycle has been insufficient household savings, not un-affordable mortgage payments.3 This means that OAS still have room to widen alongside greater net issuance. The winding down of the Fed's mortgage portfolio - a process that is likely to begin later this year - will only add to the supply that the market needs to absorb. How will the opposing forces of low option cost and widening OAS net out? The option cost component of spreads is already close to its all-time low, while the OAS is still 16 bps below its pre-crisis mean. We think it is unlikely that a lower option cost can fully offset OAS widening. Government-Related: Underweight Chart 5Government-Related Market Overview Government-Related Market Overview Government-Related Market Overview The Government-Related index outperformed the duration-equivalent Treasury index by 2 basis points in April, bringing year-to-date excess returns up to 75 bps. The high-beta Sovereign and Foreign Agency sectors outperformed by 8 bps and 1 bp, respectively. Meanwhile, the low-beta Domestic Agency and Supranational sectors outperformed by 7 bps each. Local Authorities underperformed the Treasury benchmark by 23 bps. Since the beginning of the year, excess returns from the Sovereign sector have been supported by a weakening U.S. dollar (Chart 5). Mexican debt, in particular, has benefited from a 10% appreciation of the peso relative to the U.S. dollar (panel 3). A stronger peso obviously makes Mexico's USD-denominated debt easier to service and has led to year-to-date excess returns of 402 bps for Mexican sovereign debt relative to U.S. Treasuries. Mexican debt accounts for 21% of the Sovereign index. Our Emerging Markets Strategy service thinks that Mexico's central bank could deliver another 50 bps of rate hikes, because inflation is above target, but also maintains that further rate hikes will soon start to squeeze consumer spending.4 Conversely, the Fed has scope to hike rates much further. Sovereigns no longer appear expensive on our model, relative to domestic U.S. corporate sectors. But we still expect them to underperform as the dollar resumes its bull market. Local authorities and Foreign Agencies still offer lucrative spreads on our model, and we remain overweight those spaces within an overall underweight allocation to the Government-Related index. Municipal Bonds: Neutral Chart 6Municipal Market Overview Municipal Market Overview Municipal Market Overview Municipal bonds underperformed the duration-equivalent Treasury index by 12 basis points in April (before adjusting for the tax advantage). The average Municipal / Treasury (M/T) yield ratio was flat on the month, but has fallen 15% since peaking shortly after the U.S. election (Chart 6). The sparse details of the Trump administration's proposed tax reform plan, released last week, did not include any specific mention of the municipal bond tax exemption, but did call for the elimination of "targeted tax breaks" leaving some to wonder if the tax exemption is in play. It is too soon to tell whether repealing the tax exemption will be part of the final tax reform plan, although its repeal would be at odds with the President's stated desire to spur infrastructure spending. For this reason, we suspect the tax exemption will ultimately survive. Assuming the tax exemption survives, the proposed repeal of the Alternative Minimum Tax and of the state & local government income tax deduction should both increase demand for tax-exempt municipal bonds. However, this positive impact will be offset by lower tax rates. All in all, it is too soon to know how this will all shake out, but the considerable uncertainty makes us reluctant to take strong directional bets in the municipal bond market for now. Meanwhile, Muni mutual fund inflows have totaled more than $9 billion since the beginning of the year, while total issuance is at a 12-month low. Strong inflows and low supply likely explain why yield ratios are testing the low-end of their post-crisis trading range. Treasury Curve: Favor 5-Year Bullet Over 2/10 Barbell Chart 7Treasury Yield Curve Overview Treasury Yield Curve Overview Treasury Yield Curve Overview The Treasury curve shifted lower in April, with the 2/10 slope flattening by 12 basis points and the 5/30 slope steepening by 6 bps. The 5-year Treasury yield declined 12 bps on the month, while the 10-year yield fell 11 bps. The 2-year yield actually ticked 1 bp higher. Significant outperformance in the 5-year part of the curve means that our recommendation to favor the 5-year bullet over a duration-matched 2/10 barbell has returned 27 bps since inception on December 20, 2016. This 5-year bullet over duration-matched 2/10 barbell trade is designed to profit from 2/10 curve steepening, which has not yet materialized. Instead, the trade has performed well because the 2/5/10 butterfly spread has moved much closer to our estimate of fair value (Chart 7). The 5-year bullet still looks moderately cheap on the curve, but no longer offers an exceptional valuation cushion. For our trade to outperform from here we will likely need to see some 2/10 curve steepening. We continue to hold the 5-year bullet over duration-matched 2/10 barbell trade, because we still expect the 2/10 slope to steepen. This steepening will be driven by wider long-maturity TIPS breakevens which should eventually catch up to leading pipeline inflation measures (see next page). In a recent report,5 we outlined the main drivers of the slope of the yield curve on a cyclical horizon and concluded that wider breakevens can cause the nominal curve to steepen even with the Fed in the midst of hiking rates. TIPS: Overweight Chart 8TIPS Market Overview TIPS Market Overview TIPS Market Overview TIPS underperformed the duration-equivalent nominal Treasury index by 25 basis points in April. The 10-year TIPS breakeven rate declined 5 bps on the month and, at 1.92%, it remains well below its pre-crisis trading range of 2.4% to 2.5%. Our Financial Model of TIPS breakevens - which models the 10-year TIPS breakeven rate using the stock-to-bond total return ratio, the price of oil and the trade-weighted dollar - attributes the recent decline in breakevens to weakness in the stock-bond ratio and the fact that the 10-year breakeven rate was already quite elevated compared to fair value (Chart 8). Both core and trimmed mean PCE inflation dropped sharply in March, and are now running at 1.6% and 1.8% year-over-year, respectively (bottom panel). This decline is likely to reverse in the coming months. Crucially, pipeline inflation measures, such as the ISM prices paid index, are holding firm at high levels (panel 4). We remain overweight TIPS versus nominal Treasuries on the view that growth will be strong enough to keep measures of core inflation on a steady upward trajectory, eventually converging with the Fed's 2% inflation target. In that environment, TIPS breakevens should eventually return to their pre-crisis range. In last week's report,6 we considered the possibility that TIPS breakevens might not return to their pre-crisis trading range, even if measures of core inflation remain strong. The most likely reason relates to structural rigidities in the repo market that have made it more costly to arbitrage the difference between real and nominal rates. For now, we consider this simply a risk to our overweight view. ABS: Overweight Chart 9ABS Market Overview ABS Market Overview ABS Market Overview Asset-Backed Securities outperformed the duration-equivalent Treasury index by 11 basis points in April, bringing year-to-date excess returns up to +33 bps. Aaa-rated issuers outperformed the Treasury benchmark by 10 bps on the month, while non-Aaa issues outperformed by 13 bps. The index option-adjusted spread for Aaa-rated ABS tightened 1 bp on the month, and remains well below its average pre-crisis level. Banks are now tightening lending standards on both auto loans and credit cards. While we do not expect this recent development to have much of an impact on consumer spending,7 it is usually an indication that there is growing concern about ABS collateral credit quality. This concern is echoed by the fact that net losses on auto loans are trending sharply higher (Chart 9). Credit card charge-offs remain subdued for now - and we continue to recommend that investors favor Aaa-rated credit cards over Aaa-rated auto loans - but even in the credit card space quality concerns are starting to mount. Capital One reported a 20% drop in earnings in Q1 versus the same quarter in 2016, and noted that it has been tightening underwriting standards against a back-drop of credit card loans growing faster than income. We remain overweight ABS for now, as the securities still offer attractive spreads compared to other high-quality spread product, but we are closely monitoring credit quality metrics for signs of rising stress. Non-Agency CMBS: Underweight Chart 10CMBS Market Overview CMBS Market Overview CMBS Market Overview Non-Agency Commercial Mortgage-Backed Securities outperformed the duration-equivalent Treasury index by 3 basis points in April, bringing year-to-date excess returns up to +19 bps. The index option-adjusted spread for non-agency Aaa-rated CMBS widened 1 bp on the month, and is fast approaching its average pre-crisis level. Apartment and office building prices are growing strongly, but as in the corporate space, the retail sector is a major drag (Chart 10). Tighter lending standards and falling demand also suggest that credit stress is starting to mount, but while office and retail delinquencies are rising multi-family delinquencies remain low (panel 5). Agency CMBS: Overweight Agency CMBS outperformed the duration-equivalent Treasury index by 3 basis points in April, bringing year-to-date excess returns up to +19 bps. The index option-adjusted spread for Agency CMBS widened 1 bp on the month, and currently sits at 54 bps. The option-adjusted spread on Agency CMBS looks attractive compared to other high-quality spread product: Agency MBS = 35 bps, Aaa consumer ABS = 46 bps, Agency bonds = 17 bps and Supranationals = 20 bps. We continue to recommend an overweight position in Agency CMBS. Treasury Valuation Chart 11Treasury Fair Value Models Treasury Fair Value Models Treasury Fair Value Models The current reading from our 2-factor Treasury model (which is based on Global PMI and dollar sentiment) places fair value for the 10-year Treasury yield at 2.59% (Chart 11). Our 3-factor version of the model, which also includes the Global Economic Policy Uncertainty Index, places fair value at 2.43%. The lower fair value results from the large spike in the uncertainty index last November, which has only been partially unwound (bottom panel). Large spikes in uncertainty that do not coincide with deterioration in other economic indicators tend to mean revert fairly quickly. So we are inclined to view the fair value reading from our 2-factor model as more indicative of true fair value at the moment. It should also be noted that the fair value readings from both the 2-factor and 3-factor models are calculated using FLASH PMI estimates for April. These estimates will be revised later today when the actual PMI data are released. For further details on our Treasury models please refer to the U.S. Bond Strategy Weekly Report, "The Message From Our Treasury Models", dated October 11, 2016, available at usbs.bcaresearch.com At the time of publication the 10-year Treasury yield was 2.32%. 1 Please see U.S. Bond Strategy Weekly Report, "The Payback Period In Corporate Bonds", dated April 11, 2017, available at usbs.bcaresearch.com 2 Please see U.S. Bond Strategy Weekly Report, "Keep Buying Dips", dated March 28, 2017, available at usbs.bcaresearch.com 3 Please see U.S. Bond Strategy Weekly Report, "Keep Buying Dips", dated March 28, 2017, available at usbs.bcaresearch.com 4 Please see Emerging Markets Strategy Weekly Report, "A Time To Be Contrarian", dated April 5, 2017, available at ems.bcaresearch.com 5 Please see U.S. Bond Strategy Weekly Report, "The Yield Curve On A Cyclical Horizon", dated March 21, 2017, available at usbs.bcaresearch.com 6 Please see U.S. Bond Strategy Weekly Report, "Will Breakevens Ever Recover?", dated April 25, 2017, available at usbs.bcaresearch.com 7 Please see U.S. Bond Strategy Weekly Report, "The Odds Of March", dated February 21, 2017, available at usbs.bcaresearch.com Fixed Income Sector Performance Recommended Portfolio Specification Corporate Sector Relative Valuation And Recommended Allocation Total Return Comparison: 7-Year Bullet Versus 2-20 Barbell (6-Month Investment Horizon)

Highlights High Conviction Views: The global cyclical backdrop remains negative for government bond markets, and the recent declines in yields will not be sustained. We continue to recommend a below-benchmark overall duration stance, favoring U.S. corporate debt with underweight exposures to U.S. Treasuries and Italian government debt, as our highest conviction views. Medium Conviction Views: Staying overweight global inflation protection, French government bonds versus Germany, and Japanese Government Bonds (JGBs) versus the rest of the developed bond markets, while remaining underweight U.S. Mortgage Backed Securities, are recommendations that we hold with a more moderate conviction level. Euro Area Bond Distortions: The ECB's negative interest rate and asset purchase programs have created significant distortions in the German bond yield curve that are not as evident in the Euro Area swap rate curve, especially at shorter maturities. ECB tapering will be the trigger for a reversal of these trends. Feature Chart of the WeekWhy Are Yields Falling? Why Are Yields Falling? Why Are Yields Falling? After publishing two Special Reports in the past two weeks, this Weekly Report is our first opportunity to comment on the markets in April. We find it somewhat surprising that government bonds in the developed world have rallied as much as they have since the most recent peak last month, with the benchmark 10-year U.S. Treasury and German Bund seeing yield declines of -29bps and -22bps, respectively. Most of the move in Treasuries has been in the real yield component, while Bunds have seen a more even split between declines in real yields and inflation expectations. This has occurred despite minimal changes in actual growth or inflation pressures in either the U.S. or Europe (Chart of the Week). The price action in the Treasury market after last week's U.S. Payrolls report is a sign that the bond backdrop remains bearish. Yields initially fell all the way to 2.26% after the March increase in jobs fell short of expectations, before subsequently rebounding sharply to end the day at 2.38%. While intraday yield reversals on Payrolls Fridays are as typical as the sun setting in the west, a 12bp swing is one of the larger ones in recent memory (perhaps because investors eventually noticed the weather-related distortions in the data or, more importantly, that the U.S. unemployment rate had fallen to 4.5%). We continue to favor a pro-growth bias for bond investors, staying below-benchmark on overall duration and selectively overweight on corporate credit (favoring the U.S.). Ranking Our Current Market Views, By Conviction We have seen little in the economic data over the past few weeks to change our main strategic market views and portfolio recommendations. We summarize our main opinions below, ranked in order of our conviction level: Highest conviction views: Below-benchmark on overall portfolio duration exposure (for dedicated bond investors). Global bond yields have more room to rise alongside solid economic growth, tightening labor markets, inflation expectations drifting higher and central banks moving to slightly less accommodative monetary policies, on the margin. While the sharp upward momentum in coincident bond indicators like the global ZEW sentiment index has cooled of late, the solid upturn in the BCA Global Leading Economic Indicator continues to point to future upward pressure on real yields (Chart 2). The recent pullback in yields also appears to have run too far versus the trend in global data surprises, which remain elevated (bottom panel). One factor that we see having a potentially huge negative impact on global bond markets is the European Central Bank (ECB) announcing a move to a less accommodative policy stance later this year. A taper of asset purchases starting in 2018 is the more likely outcome than any hike in policy interest rates, which we see as more of a story for 2019. This should help push longer-dated bond yields higher within the Euro Area, and drag up global bond yields more generally. Underweight U.S. Treasuries. We still expect the Fed to deliver at least two more hikes this year, and there is still room for U.S. inflation expectations to rise further and put bear-steepening pressure on the Treasury curve. Our two-factor model for the benchmark 10-year Treasury yield, which uses the global purchasing managers index (PMI) and investor sentiment towards the U.S. dollar as the explanatory variables, indicates that yields are now about 18bps below fair value. From a technical perspective, the Treasury market no longer appears as oversold as it did after the rapid run-up in yields following last November's U.S. elections. The large short positions indicated by the J.P. Morgan duration survey and the Commitment of Traders report for Treasury futures have largely been unwound, while price momentum has flipped into positive territory (Chart 3). This removes one of the largest impediments to a renewed decline in Treasury prices, and we expect that the 10-year yield to rise to the upper end of the recent 2.30%-2.60% trading range in the next couple of months, before eventually breaking out on the way to the 2.80%-3% area by year-end. Chart 2Maintain A Defensive Duration Posture Maintain A Defensive Duration Posture Maintain A Defensive Duration Posture Chart 3Stay Underweight U.S. Treasuries Stay Underweight U.S. Treasuries Stay Underweight U.S. Treasuries Underweight Italian government bonds, versus both Germany and Spain. Italian government debt continues to suffer from the toxic combination of sluggish growth and weak domestic banks. The OECD leading economic indicator for Italy is declining, in contrast to the stable-to-rising trends in Germany and Spain (Chart 4). Meanwhile, the 5-year credit default swaps (CDS) for the major banks in Italy remain elevated around 400bps, in sharp contrast to the declining CDS in Germany and Spain which are now at 100bps. It is no coincidence that the widening trend in Italy-Germany and Italy-Spain spreads began around the same time last year that Italian bank CDS started to disengage from the rest of Europe (bottom panel). Markets understand that the undercapitalized Italian banking system will need government assistance at some point, which will add to the Italian government's already huge debt/GDP ratio of 133%. Political uncertainty in Italy, with parliamentary elections due by the spring of 2018 and populist parties like the anti-euro Five-Star Alliance holding up well in the polls, will also ensure that the risk premium on Italian bonds stays wide both in absolute terms and relative to other Peripheral European markets. Overweight U.S. corporate bonds, versus both U.S. Treasuries and Euro Area equivalents. The positive case for U.S. corporate debt is built upon two factors - the cyclical decline in default risk and the marginal improvement in balance sheet metrics. The latest estimates from Moody's are calling for a decline in the U.S. speculative grade corporate default rate to 3.1% this year. This leaves our measure of default-adjusted spreads in U.S. high-yield at levels that our colleagues at our sister publication, U.S. Bond Strategy, have shown to have a high probability of delivering positive excess returns over Treasuries in the next 12 months.1 Add to that the recent change in trend of our U.S. Corporate Health Monitor (CHM), which appears largely driven by some more positive numbers coming from lower-rated issuers in the Energy space given the recovery in oil prices, and the optimistic case for U.S. corporate debt is compelling. This is in contrast to our Euro Area CHM, which shows that the improving trend in balance sheet metrics has stalled of late (Chart 5, top panel). Chart 4Stay Underweight Italy Stay Underweight Italy Stay Underweight Italy Chart 5Stay Overweight U.S. Corporates vs Europe Stay Overweight U.S. Corporates vs Europe Stay Overweight U.S. Corporates vs Europe The difference between the U.S. and European CHMs has proven to be a good directional indicator for the relative return performance between the two markets, and is currently pointing to continued outperformance of both U.S. investment grade and high-yield debt versus European equivalents (bottom two panels). The threat of an ECB taper also hangs over the Euro Area investment grade corporate bond market, given the large buying of that debt by the central bank over the past year that has helped dampen both yields and spreads. Chart 6Stay Overweight Inflation Protection Stay Overweight Inflation Protection Stay Overweight Inflation Protection Medium-conviction views: Overweight inflation protection (both inflation-linked bonds and CPI swaps) in the U.S., Euro Area and Japan. In the U.S., the breakeven inflation rate on 10-year TIPS looks a bit too wide relative to our shorter-term model based on financial variables. However, underlying U.S. inflation pressures remain strong (Chart 6, top panel), particularly given the evidence that conditions in the labor market are getting progressively tighter. We expect inflation expectations to eventually rise back to levels consistent with the Fed's 2% inflation target on headline PCE inflation (which is around 2.5% on 10-year TIPS breakevens that are priced off the CPI index). The reflation story is somewhat less compelling in Europe and Japan, although CPI swaps are now at levels consistent with the underlying trends in realized inflation in both regions (bottom two panels). We continue to view long positions in CPI swaps in Europe and Japan as having a positive risk/reward skew given the tightening labor market in the former and the yen-negative monetary policies in the latter. Long France government bonds (10yr OATs) versus Germany (10yr Bunds). This is purely a call on the upcoming French election, which our political strategists believe will not end in a victory for the populist Marine Le Pen. While Le Pen has seen a recent bump in support heading into the first round of voting on April 23rd, her strong anti-euro position will eventually prove to be her undoing in the run-off election on May 7th (Chart 7). We first made this recommendation back in early February, and even though France-Germany spreads have been volatile since then as both Le Pen and the far-left candidate Jean-Luc Melenchon have seen a pickup in their poll numbers, the yield differentials are essentially at the same levels.2 We take this as a sign that the market believes current spreads are enough to compensate for the likely probability that either candidate could win the French presidency. Overweight JGBs Vs. the Global Treasury index. The argument here is a simple one - in an environment where there is cyclical upward pressure on global bond yields, favor the lowest-beta bond market (Chart 8). Persistently low inflation will prevent the Bank of Japan (BoJ) from making any changes to its current hyper-accommodative policies this year, especially the 0% cap on the benchmark 10-year JGB yield.3 The lack of yield limits the prospects for JGBs on a total return basis, but relative to other government bond markets, JGBs should outperform over the next 6-12 months as non-Japanese yields rise further. Chart 7Stay Overweight France Vs Germany Stay Overweight France Vs Germany Stay Overweight France Vs Germany Chart 8Stay Overweight Low-Beta JGBs Stay Overweight Low-Beta JGBs Stay Overweight Low-Beta JGBs Underweight U.S. Agency Mortgage-Backed Securities (MBS). Investors should remain underweight U.S. MBS, as spreads remain tight by historical standards. Our colleagues at U.S. Bond Strategy note that nominal MBS spreads have been flat in recent weeks as the option cost, which is the compensation for expected prepayments, has tightened to offset a widening in the option-adjusted spread (OAS).4 Chart 9Stay Underweight U.S. MBS Stay Underweight U.S. MBS Stay Underweight U.S. MBS We tend to think of the OAS as being influenced by trends in net issuance while the option cost is linked to mortgage prepayments (Chart 9). Looking ahead, the supply of MBS should increase further when the Fed starts to shrink its balance sheet later this year (as was mentioned in the minutes of the March FOMC meeting that were released last week), leading to a wider OAS. At the same time, refinancing applications should stay low as Treasury yields and mortgage rates rise. This will keep downward pressure on the option cost component of spreads. But with the option cost already near its historical lows, it is unlikely to completely offset the widening in OAS going forward. We see little value in U.S. MBS at current spread levels. Bottom Line: The global cyclical backdrop remains negative for government bond markets, and the recent declines in yields will not be sustained. We continue to recommend a below-benchmark overall duration stance, favoring U.S. corporate debt with underweight exposures to U.S. Treasuries and Italian government debt, as our highest conviction views. Staying overweight global inflation protection, French government bonds versus Germany, and Japanese Government Bonds (JGBs) versus the rest of the developed bond markets, while remaining underweight U.S. Mortgage Backed Securities, are recommendations that we hold with a more moderate conviction level. How Much Has The ECB Distorted The European Bond Market? Last week, Benoit Coeure of the ECB Executive Board gave a speech entitled "Bond Scarcity and the ECB Asset Purchase Program."5 That title piqued our interest, as that exact topic has come up in several of our conversations with clients this year. In his speech, Coeure discussed how the huge rally at the short-end of the German government bond curve over the past year has been at odds with what has occurred in the Euro swap curve, where interest rates are much higher for shorter-maturity swaps. Typically, German yields and Euro swap rates move in tandem, with the only differences being a function of technical factors like fixed-rate corporate debt issuance or government bond repo rates - and, on occasion, shifts in the perceived health of Euro Area banks that are the counterparties to any interest rate swap. The latter has become much less of an issue in recent years given the regulatory changes to the swap market, where trading has moved to centralized exchanges to reduce counterparty risks. In this environment, the difference between German bond yields and Euro swap rates, a.k.a the swap spread, should be relatively modest. Yet as can be seen in Chart 10, there has been a notable divergence at the shorter-maturity portions of the respective yield curves, where swap rates are rising but bond yields remain subdued. We can also see the divergences in the slopes of the relative yield curves, with the Euro Area swap curve much flatter than the German bond curve, particularly at longer maturities (Chart 11). Chart 10Large Distortions At The Front End Of The German Curve Large Distortions At The Front End Of The German Curve Large Distortions At The Front End Of The German Curve Chart 11Euro Area Swap Curves Are Generally Flatter Euro Area Swap Curves Are Generally Flatter Euro Area Swap Curves Are Generally Flatter Coeure argued that part of this distortion can be attributed to ECB asset purchases, especially after the decision taken last December to allow bond buying at yields below the -0.4% ECB deposit rate. This created a more favorable demand/supply balance for German debt, especially given the dearth of short-dated issuance. In addition, Coeure noted that there have been substantial safe-haven flows into shorter-dated German bonds (including treasury bills) by non-Euro Area entities. Some of this demand comes from large institutional investors like sovereign wealth funds and currency reserve managers, who are worried about political risks in France and Italy, and about the general rising trend in global bond yields, and are thus seeking the safety of low duration German debt. But some of the demand for short-dated German paper also comes from non-Euro Area banks, who have excess liquidity that needs to be parked in Euros but do not have access to the ECB deposit facility for the excess reserves of Euro Area banks. We can see this in Chart 12, which shows ECB data for the relative government bond ownership trends for Germany, France and Italy. The data is broken into holdings for bonds with maturities of one year or less (short-term) and bonds with maturities greater than one year (long-term). It is clear that the non-Euro area buyers own a much larger share of short-term German paper, around 90%, than in France and Italy, while Euro Area entities own nearly 80% of long-term bonds in all three countries. Coeure is correct in pointing out that there is an excess demand condition for short-dated core European debt, exacerbated by foreigners who need Euro-denominated safe assets - particularly GERMAN safe assets, if those investors are at all worried about redenomination risks given the rise of anti-euro populist parties in Europe.6 It is clear that the economic messages sent by looking at the German bond and Euro swap curves are very different. The flatter swap curve is more consistent with a steadily growing Euro Area economy where economic slack is being steadily absorbed and inflation pressures are building (albeit slowly). Also, the sovereign spread differentials within Europe do not look as problematic using swaps as the reference rate rather than German bonds. That is the case in France, where spreads versus swaps look in line with the averages of the past few years (Chart 13). This contrasts with the yield differentials versus Germany, which have reportedly gone up as investors have priced in a higher sovereign risk premium before the French presidential election. Chart 12French Bond Valuations Look More Subdued vs Swaps The Song Remains The Same The Song Remains The Same Chart 13French Bond Valuations Look More Subdued vs Swaps French Bond Valuations Look More Subdued vs Swaps French Bond Valuations Look More Subdued vs Swaps The story is a little different for Italy, where bond spreads versus both German bonds and Euro Area swaps have risen for all but the shortest maturities (Chart 14). This could be consistent with an interpretation that Italy's banking sector woes will add to the nation's longer-term fiscal stresses (as discussed earlier in this report), but not in a way that raises immediate default risks (which is why the 2-year Italy vs swap spread is well-behaved). Regardless of the "bias of interpretation", one thing that is clear is that the ECB's extraordinary monetary policies have created distortions in Euro Area bond markets. These may start to unwind, though, if the ECB begins to signal a shift towards a tapering of asset purchases next year, as we expect. The distortions in Euro area government bond yields (and, by association, swap spreads) have occurred alongside both the cuts in ECB policy rates into negative territory and the expansion of its balance sheet to purchase government bonds (Chart 15). As the ECB moves incrementally towards less accommodative monetary policy, we would expect to see front-end Euro swap spreads narrow in absolute terms and relative to longer-tenor spreads, and the German bond curve to flatten toward levels seen in the swap curve. Chart 14Only Short-Dated Italian Bond Valuations Look More Subdued vs Swaps Only Short-Dated Italian Bond Valuations Look More Subdued vs Swaps Only Short-Dated Italian Bond Valuations Look More Subdued vs Swaps Chart 15ECB Policies Have Caused The Distortions In Euro Swap Spreads ECB Policies Have Caused The Distortions In Euro Swap Spreads ECB Policies Have Caused The Distortions In Euro Swap Spreads Bottom Line: The ECB's negative interest rate and asset purchase programs have created significant distortions in the German bond yield curve that are not as evident in the Euro Area swap rate curve, especially at shorter maturities. ECB tapering will be the trigger for a reversal of these trends. Robert Robis, Senior Vice President Global Fixed Income Strategy rrobis@bcaresearch.com 1 Please see BCA U.S. Bond Strategy Weekly Report, "Buy The Back-Up In Junk Spreads", dated March 14, 2017, available at usbs.bcaresearch.com 2 Please see BCA Global Fixed Income Strategy Special Report, "Our Views On French Government Bonds", dated February 7, 2017, available at gfis.bcaresearch.com 3 Please see BCA Global Fixed Income Strategy Weekly Report, "Staying Behind The Curve, For Now", dated March 21, 2017, available at gfis.bcaresearch.com 4 Please see BCA U.S. Bond Strategy Weekly Report, "The Payback Period In Corporate Bonds", dated April 11, 2017, available at usbs.bcaresearch.com 5 http://www.ecb.europa.eu/press/key/date/2017/html/sp170403_1.en.html 6 Coeure noted that, at the time that the ECB began its asset purchase program in March 2015, the share of German bonds of less than TWO years maturity held by foreigners was 70%, but that rose to 90% by the 3rd quarter of 2016. The GFIS Recommended Portfolio Vs. The Custom Benchmark Index The Song Remains The Same The Song Remains The Same Recommendations Duration Regional Allocation Spread Product Tactical Trades Yields & Returns Global Bond Yields Historical Returns

Highlights Duration: Bond market positioning is no longer at a bearish extreme and the economy is quickly approaching full employment. We expect Treasury yields will soon break through the upside of their post-election trading range. Maintain below-benchmark duration. Fed's Balance Sheet: The unwinding of the Fed's balance sheet is only important for Treasury yields if it impacts the market's rate hike expectations. However, the extra supply of MBS should lead to wider MBS spreads. Credit Cycle: Corporate spreads are in a "payback period" from 2014's energy shock that will allow them to tighten as corporate profits rebound, even though corporate leverage continues to trend higher. The weakening state of corporate balance sheets means spreads are at risk once monetary policy turns less accommodative. Feature The bond bear market has been on pause for the past few months, with Treasury yields confined to a trading range since last November's post-election sell off. While yields have not moved meaningfully higher during this time, firm floors have also formed beneath both the 5-year and 10-year yields (Chart 1). Even after last Friday's disappointing payrolls number, the 10-year did not move below 2.3% and the 5-year did not move below 1.8%. Trading Range About To Break? Our sense is that the current consolidation phase in Treasuries is approaching its end and yields will soon head higher. Global growth indicators have continued to improve during the past few months, and as we noted in last week's report,1 our 2-factor Treasury model, based on Global PMI and U.S. dollar sentiment, pegs fair value for the 10-year yield at 2.54%. We attribute the recent leveling-off in yields to technical shifts in bond positioning and sentiment. Earlier this year, net positions in Treasury futures and asset manager duration allocations were deep in "net short" territory (Chart 2). Extreme short positioning usually leads to a period of bond market strength until short positions are washed out. Now that bond market positioning is closer to neutral, a key impediment to further yield increases has been removed. Chart 1Poised For A Breakout? Poised For A Breakout? Poised For A Breakout? Chart 2Positioning Has Normalized Positioning Has Normalized Positioning Has Normalized The elevated level of economic surprises has also been flagged as a potential roadblock to the bond bear market. Extended readings from the economic surprise index tend to mean revert as investor expectations are revised higher in the face of improving data. However, our research suggests that the change in Treasury yields tends to lead the economic surprise index by 1-2 months (Chart 2, bottom panel). Given this relationship, we suspect that the bond market has already discounted a lot of mean reversion in the economic surprise index. Chart 3Approaching Full Employment Approaching Full Employment Approaching Full Employment Finally, last week's employment report should not be taken as a signal that U.S. economic growth is weakening. Bad weather in the northeast played a key role in the low March payrolls number - only 98k jobs added. But more importantly, at this stage of the cycle we should expect payroll growth to slow and wage pressures to increase as we approach full employment. As can be seen in Chart 3, the late cycle trends of slowing payroll growth and rising wages are very much in place. Further, even broad measures of labor market tightness, such as the U6 unemployment rate,2 are quickly approaching levels that suggest the economy is operating at full employment. Increasingly it is measures of labor market utilization, wage growth and inflation that will guide the Fed's decision making, and these measures continue to improve. It was even noted in the minutes from the March FOMC meeting that "tight labor markets [are] increasingly a factor in businesses' planning". The minutes also reported that: Business contacts in many Districts reported difficulty recruiting workers and indicated that they had to either offer higher wages or hire workers with lower qualifications than desired Accordingly, surveys show that households are increasingly describing jobs as "plentiful" (Chart 3, panel 3) and small businesses are indeed ramping up their compensation plans (Chart 3, bottom panel). At this stage of the cycle, continued progress on measures of labor market utilization, wage growth and inflation will be sufficient for the Fed to continue lifting rates, pushing Treasury yields higher. Bottom Line: Bond market positioning is no longer at a bearish extreme and the economy is quickly approaching full employment. We expect Treasury yields will soon break through the upside of their post-election trading range. Maintain below-benchmark duration. The Fed Will Shrink Its Balance Sheet This Year Last week's release of the minutes from the March FOMC meeting also contained some new information about how the Fed plans to deal with its large balance sheet. To summarize, we learned that: The Fed intends to start shrinking its balance sheet later this year (assuming growth maintains its current pace). The Fed will shrink its balance sheet by ceasing the reinvestment of both its MBS and Treasury holdings at the same time. Still no decision has been made about whether reinvestments will stop entirely or whether they will be phased out over time ("tapered"). On February 28, we published a detailed report about the Fed's balance sheet policy.3 In that report we explained why the winding down of the balance sheet will not have much of an impact on Treasury yields, but could lead to a material widening in MBS spreads. The new information received last week does not change either of these conclusions. The minutes did make clear that the Fed favors what Governor Lael Brainard recently called a "subordination strategy" for dealing with its balance sheet.4 [A subordination strategy] would prioritize the federal funds rate as the sole active tool away from the effective lower bound, effectively subordinating the balance sheet. Once federal funds normalization meets the test of being well under way, triggering an end to the current reinvestment policy, the balance sheet would be set on autopilot, shrinking in a gradual, predictable way until a "new normal" has been reached, and then increasing in line with trend increases in the demand for currency thereafter. Under this strategy, the balance sheet might be used as an active tool only if adverse shocks push the economy back to the effective lower bound. Essentially, the Fed is trying to de-emphasize the size of the balance sheet and would rather investors focus on the fed funds rate to assess the stance of monetary policy. For our part, we think it would be unwise to "fight the Fed" on this issue. For Treasury yields, we observe that the real 10-year Treasury yield closely tracks changes in the expected number of rate hikes during the next 12 months, while the inflation component of the 10-year yield tracks changes in realized inflation (Chart 4). These two relationships will continue to determine trends in bond yields going forward, and Fed balance sheet shrinkage is only important if it impacts the expected pace of rate hikes or inflation. The Fed's "subordination strategy" should ensure that the act of winding down the balance sheet does not have much of an impact on the expected pace of rate hikes. Ironically, if Treasury yields were to rise sharply following the announcement of balance sheet runoff, then the ensuing tightening of financial conditions would probably lower the expected pace of rate hikes and bring Treasury yields back down again. The story for MBS is somewhat different. Nominal MBS spreads remain tight by historical standards and closely track implied interest rate volatility (Chart 5). But we can also think of nominal MBS spreads as being split between the option cost, which is the compensation for expected prepayments, and the option-adjusted spread (OAS), which tends to correlate with net supply (Chart 5, panel 2). Chart 4Focus On Rate Expectations Focus On Rate Expectations Focus On Rate Expectations Chart 5Stay Underweight MBS Stay Underweight MBS Stay Underweight MBS In recent weeks, the OAS has widened alongside rising net issuance, but this has been offset by a sharp decline in the option cost. This is generally the pattern we would expect to play out as the Fed lifts rates and removes itself from the MBS market. The increased supply of MBS should lead to wider OAS, but refinancing applications should also stay low as Treasury yields and mortgage rates rise (Chart 5, bottom panel). However, netting it all out, the option cost component of MBS spreads is already near its historical lows and the OAS could move materially wider just to catch up to net issuance. In prior reports,5 we have also made the case that rate volatility should rise as the fed funds rate moves further away from the zero-lower-bound. Investors should stay underweight MBS. Bottom Line: The unwinding of the Fed's balance sheet is only important for Treasury yields if it impacts the market's rate hike expectations. However, the extra supply of MBS should lead to wider MBS spreads. Checking In On The Credit Cycle We continue to recommend overweight allocations to both investment grade and high-yield corporate bonds. This optimistic outlook is predicated on low inflation and a Fed that will support risk assets by remaining sufficiently accommodative until inflationary pressures are more pronounced. We think this "reflationary window" will stay open at least until core PCE inflation is firmly anchored around 2% and long-maturity TIPS breakevens reach the 2.4% to 2.5% range.6 Behind the scenes, however, leverage is building in the nonfinancial corporate sector. In this week's report we take a look at several different indicators of corporate credit quality and conclude that once the support from low inflation and accommodative monetary policy vanishes, it is very likely that corporate defaults will start to increase and corporate spreads will widen. If our anticipated timeline plays out, we will be looking to scale back on credit risk in 2018. Corporate Health Vs. The Yield Curve Our Corporate Health Monitor (CHM, see Appendix for further details) has been signaling deteriorating nonfinancial corporate health since late 2013 (Chart 6), and moved even deeper into 'deteriorating health' territory in Q4 of last year. Chart 6Corporate Health Is Deteriorating, But Monetary Policy Remains Supportive Corporate Health Is Deteriorating, But Monetary Policy Remains Supportive Corporate Health Is Deteriorating, But Monetary Policy Remains Supportive Periods when the CHM is in 'deteriorating health' territory are marked by shaded regions in Chart 6. We see that these regions usually correspond with periods when corporate spreads are widening. Even in the current episode, corporate spreads have yet to regain their mid-2014 tights. However, the bottom panel of Chart 6 shows that periods of deteriorating corporate health and wider corporate spreads are typically preceded by a very flat (often inverted) yield curve. This makes sense because a flat yield curve usually signals that interest rates are high and monetary policy is tight. Tight policy and elevated rates lead to more stringent bank lending standards and increase firms' interest burdens. With the curve still quite steep, we think the risk of sustained spread widening is minimal. However, if the CHM is still above zero when the yield curve is flatter, no support will remain for excess corporate bond returns. Net Leverage & The Payback Period We would further argue that the CHM will almost certainly be in 'deteriorating health' territory once the yield curve is close to flat. In Chart 7 we see that net leverage (defined as: total debt minus cash, as a percent of EBITD) is not only positively correlated with spreads, but also has never reversed its uptrend unless prompted by a recession. In other words, the corporate sector never voluntarily undertakes deleveraging, it only starts to pay down debt when forced by a severe economic contraction. Chart 7The Uptrend In Leverage Will Only Be Broken By Recession The Uptrend In Leverage Will Only Be Broken By Recession The Uptrend In Leverage Will Only Be Broken By Recession Closer inspection of Chart 7 reveals that the period between 1986 and 1989 is the only period when corporate spreads tightened even though leverage remained in an uptrend. In the late 1980s, leverage and corporate spreads both shot higher as a collapse in the energy sector caused overall corporate earnings to contract (Chart 7, bottom panel). But then the energy sector recovered just as quickly, and earnings growth bounced back. This caused spreads to tighten for a couple of years, even though the trend in net leverage only ever managed to flatten-off. Debt growth stayed robust during this time, despite the wild fluctuations in earnings. If any of this sounds familiar, it should. The energy sector collapse of 2014 caused net leverage and spreads to shoot higher, and now spreads have started to tighten again as earnings have rebounded. Notice that just like in the late-1980s, net leverage has not reversed its uptrend. We believe that corporate spreads have entered a "payback period" very similar to the late 1980s. Spreads can tighten as earnings rebound, but because the economy is not in recession, debt growth will remain solid and leverage will continue to trend higher. Once inflationary pressures start to bite and Fed policy becomes less accommodative, the payback period will end and spreads will head wider. Debt Growth Chart 8Bond Issuance Is Back Bond Issuance Is Back Bond Issuance Is Back Although we have made the case that the corporate sector does not delever unless prompted by a recession, it is notable that net corporate bond issuance was negative in Q4 of last year and the growth rate in bank lending to the corporate sector has slowed sharply. We do not think this cycle is different, and expect corporate debt growth (both bonds and loans) to rebound in the coming months. We chalk up weak corporate bond issuance in 2016Q4 to uncertainty surrounding the U.S. election. In fact, we see that gross corporate bond issuance has already rebounded strongly in January and February of this year (Chart 8). Turning to bank loans, we observe that the outright level of outstanding bank loans only contracts following a recession, and that the rate of increase follows bank lending standards with a lag (Chart 9). In other words, Commercial & Industrial (C&I) loan growth is still responding to the surge in defaults that resulted from the energy sector's 2014 collapse. Now that defaults have waned, this process will soon be thrown into reverse. In fact, our model of the 6-month rate of change in C&I lending - based on private non-residential fixed investment, small business optimism and corporate defaults - points to an imminent bottoming in C&I loan growth (Chart 10). Chart 9Loan Growth Follows Lending Standards Loan Growth Follows Lending Standards Loan Growth Follows Lending Standards Chart 10BCA C&I Loan Growth Model BCA C&I Loan Growth Model BCA C&I Loan Growth Model Bottom Line: Corporate spreads are in a "payback period" from 2014's energy shock that will allow them to tighten as corporate profits rebound, even though corporate leverage continues to trend higher. The weakening state of corporate balance sheets means spreads are at risk once monetary policy turns less accommodative. Ratings Trends & Shareholder Friendly Activities Chart 11Shareholder Friendly Activity Has Ebbed Shareholder Friendly Activity Has Ebbed Shareholder Friendly Activity Has Ebbed Our assessment of the cyclical back-drop for corporate spreads is primarily based on the combination of balance sheet quality - as determined by our Corporate Health Monitor and its underlying components - and the stance of monetary policy - as determined by the slope of the yield curve and C&I lending standards (among other factors). However, ratings migration and "shareholder friendly" activities have also historically provided advance notice of turns in the credit cycle. Net transfers to shareholders, i.e. payments to shareholders in the form of dividends and buybacks, are a direct transfer of capital from bondholders to equityholders. These transfers tend to rise late in the cycle, just before defaults start to increase and spreads start to widen (Chart 11). Net transfers to shareholders had been moving higher, but have recently rolled over. Similarly, ratings downgrades related to shareholder transfers have also moderated (Chart 11, panel 2). Historically, ratings migration related to "shareholder friendly" activities has been a more reliable indicator of the credit cycle than overall ratings migration. It has tended to move into "net downgrade" territory later in the cycle, closer to the onset of recession (Chart 11, panel 3). Ratings trends and transfers to shareholders are not flagging any imminent risk of spread widening. However, there is the additional risk that downgrades have simply not kept pace with the actual deterioration in credit quality of the nonfinancial corporate sector. Using firm-level data, we calculated the percent of high-yield rated companies with net debt-to-EBITDA ratios above 5. We see that actual ratings migration is too low relative to the number of highly-levered firms (Chart 11, bottom panel). It is possible that ratings agencies have already incorporated the rebound in energy prices and profit growth into their assessments while the actual debt-to-EBITDA data are lagging, but this is still a risk that bears monitoring. Ryan Swift, Vice President U.S. Bond Strategy rswift@bcaresearch.com 1 Please see U.S. Bond Strategy Portfolio Allocation Summary, "Reflation Window Still Open", dated April 4, 2017, available at usbs.bcaresearch.com 2 The U6 unemployment rate is a broader measure than the headline (U3) unemployment rate. It also includes those "marginally attached" to the labor force and those working part-time for economic reasons. 3 Please see U.S. Bond Strategy / Global Fixed Income Strategy Special Report, "The Way Forward For The Fed's Balance Sheet", dated February 28, 2017, available at usbs.bcaresearch.com 4 https://www.federalreserve.gov/newsevents/speech/brainard20170301a.htm 5 Please see U.S. Bond Strategy Weekly Report, "The Road To Higher Vol Is Paved With Uncertainty", dated February 14, 2017, available at usbs.bcaresearch.com 6 Please see U.S. Bond Strategy Weekly Report, "Keep Buying Dips", dated March 28, 2017, available at usbs.bcaresearch.com Appendix Chart 12Corporate Health Monitor Components Corporate Health Monitor Components Corporate Health Monitor Components Box 1: Corporate Health Monitor Components The BCA Corporate Health Monitor is a normalized composite of six financial ratios, calculated for the non-financial corporate sector as a whole (Chart 12). These six ratios are defined as follows: Profit Margins: After-tax cash flow as a percent of corporate sales Return on Capital: After-tax earnings plus interest expense, as a percent of capital stock Debt Coverage: After-tax cash flow less capital expenditures, as a percent of all interest bearing debt Interest Coverage: EBITDA (Earnings before interest, taxes, depreciation & amortization) divided by the sum of interest expense and dividends Leverage: Total debt as a percent of market value of equity Liquidity: Working Capital, excluding inventories, as a percent of market value of assets Fixed Income Sector Performance Recommended Portfolio Specification

Highlights Chart 1Is Inflation Heating Up? Is Inflation Heating Up? Is Inflation Heating Up? In past reports we have argued that as long as inflation (and inflation expectations) are below the Fed's target, then the "reflation trade" will remain in vogue. In other words, with inflation still too low, the Fed has an incentive to back away from its hawkish rhetoric whenever risk assets sell off and financial conditions tighten. But with inflation heating up - the last two monthly increases in core PCE are close to the highest seen in this recovery (Chart 1) - will the Fed become less responsive? Not yet! Year-over-year core PCE is still only 1.75% (the Fed's target is 2%) and the cost of inflation protection embedded in long-dated TIPS remains too low (panel 2). In fact, the uptrend in TIPS breakevens lost some of its momentum last month alongside wider credit spreads and the S&P 500's first monthly decline since October. In this environment, we are inclined to add credit risk as spreads widen and believe a "buy the dips" strategy will work until inflation pressures are more pronounced. On a 6-12 month horizon we continue to recommend: below-benchmark duration, overweight spread product, curve steepeners and TIPS breakeven wideners. Feature Investment Grade: Overweight Chart 2Investment Grade Market Overview Investment Grade Market Overview Investment Grade Market Overview Investment grade corporate bonds underperformed the duration-equivalent Treasury index by 15 basis points in March. The index option-adjusted spread widened 3 bps on the month and, at 118 bps, it remains well below its historical average (134 bps). While supportive monetary policy will ensure excess returns consistent with carry, investors should not bank on further spread compression as spreads have already discounted a substantial improvement in leverage (Chart 2). In fact, leverage showed a marked increase in Q4 2016 even though spreads moved tighter. The measure of gross leverage (total debt divided by EBITD) shown in Chart 2 increased in the fourth quarter even though total debt grew at an annualized rate of only 0.3%. However, EBITD actually contracted at an annualized rate of 7% in Q4 causing leverage to rise. The quarterly decline in EBITD looks anomalous, and the year-over-year trend is improving (panel 4). In fact, we would not be surprised to see leverage stabilize this year as profits rebound.1 But similarly, we also expect that the recent plunge in debt growth will reverse. Historically, it has been very rare for leverage to fall unless prompted by a recession. We will take up this issue in more detail in next week's report. Energy related sectors still appear cheap after adjusting for differences in credit rating and duration (Table 3), and we remain overweight. This week we also downgrade the Retailers and Packaging sectors, which have become expensive, and upgrade Cable & Satellite, which appears cheap. Table 3A Reflation Window Still Open Reflation Window Still Open Table 3B Reflation Window Still Open Reflation Window Still Open High-Yield: Overweight Chart 3High-Yield Market Overview High-Yield Market Overview High-Yield Market Overview High-Yield underperformed the duration-equivalent Treasury index by 25 basis points in March. The index option-adjusted spread widened 20 bps on the month and, at 383 bps, it is currently 136 bps below its historical average. Given the favorable policy back-drop described on page 1, we view the recent widening in junk spreads (Chart 3) as an opportunity to increase exposure to the sector. In fact, in a recent report2 we tested a strategy of "buying dips" in the junk bond market in different inflationary regimes. The strategy involved buying the High-Yield index whenever spreads widened by 20 bps or more in a month and then holding that position for 3 months. We defined the different inflationary regimes based on the St. Louis Fed's Price Pressures Measure (PPM).3 We found that our "buy the dips" strategy yielded positive excess returns 65% of the time in a very low inflation regime (PPM < 15%), 59% of the time in a low inflation regime (15% < PPM < 30%), 44% of the time in a moderate inflation regime (30% < PPM < 50%) and only 25% of the time in a high inflation regime (50% < PPM < 70%). Currently, the reading from the PPM is 13%. MBS: Underweight Chart 4MBS Market Overview MBS Market Overview MBS Market Overview Mortgage-Backed Securities outperformed the duration-equivalent Treasury index by 4 basis points in March. The conventional 30-year MBS yield rose 6 bps on the month, driven entirely by a 6 bps increase in the rate component. The compensation for prepayment risk (option cost) declined by 6 bps, but this was exactly offset by a 6 bps widening in the option-adjusted spread. As a result, the zero-volatility spread was flat on the month. The option-adjusted spread represents expected excess returns to MBS assuming that prepayments fall in line with expectations. On this basis, MBS look more attractive than they have for some time (Chart 4). However, net MBS issuance also surged in Q4 2016 (panel 4) and should remain robust this year despite higher mortgage rates.4 Interest rates have not been a deterrent to mortgage demand since the financial crisis. The limiting factors have been a lack of household savings and restrictive bank lending standards. Both of these headwinds continue to gradually fade. The option-adjusted spread still appears too low relative to issuance. Nominal MBS spreads are linked to rate volatility (bottom panel), and volatility should increase as the fed funds rate moves further off its zero-bound.5 The wind-down of the Fed's MBS portfolio - which we expect will begin in 2018 - should also pressure implied volatility higher as the private sector is forced to absorb the increased supply, some of which will be convexity-hedged. Government-Related: Underweight Chart 5Government-Related Market Overview Government-Related Market Overview Government-Related Market Overview The Government-Related index outperformed the duration-equivalent Treasury index by 22 basis points in March. The high-beta Sovereign and Foreign Agency sectors outperformed by 71 bps and 41 bps, respectively. Meanwhile, the low-beta Domestic Agency and Supranational sectors outperformed by 9 bps and 15 bps, respectively. Local Authorities underperformed the Treasury benchmark by 17 bps. The performance of Sovereigns has been stellar this year, as the sector has benefited from a 3% depreciation in the trade-weighted dollar (Chart 5). However, the downtrend in the dollar looks more like a temporary reversal than an end to the bull market. With U.S. growth on a strong footing, there is still scope for global interest rate differentials to move in favor of the dollar. Potential fiscal policy measures - such as lower tax rates and a border-adjusted corporate tax - would also lead to a stronger dollar, if enacted. As such, we do not believe the current outperformance of Sovereigns can be sustained. We continue to recommend overweight allocations to Foreign Agencies and Local Authorities, alongside underweight allocations to the rest of the Government-Related index. Municipal Bonds: Neutral Chart 6Municipal Market Overview Municipal Market Overview Municipal Market Overview Municipal bonds outperformed the duration-equivalent Treasury index by 24 basis points in March (before adjusting for the tax advantage). The average Municipal / Treasury (M/T) yield ratio declined 2% on the month and remains firmly anchored below its post-crisis average. This year's decline in M/T yield ratios has been concentrated at the short-end of the curve (Chart 6), and long maturities now offer a significant valuation advantage. This week we recommend favoring the long-end of the Aaa Muni curve (10-year maturities and beyond) versus the short-end (maturities up to 5 years). Overall, M/T yield ratios appear fairly valued on a tactical basis. While fund inflows have ebbed in recent weeks (panel 4), this has occurred alongside a plunge in gross issuance (bottom panel). The more concerning near-term risk for Munis is that yield ratios have already discounted a substantial improvement in state & local government net borrowing (panel 3). However, we expect net borrowing to decline during the next couple of quarters on the back of rising tax revenues. State & local government tax receipts decelerated throughout most of 2015 and 2016 alongside falling personal income growth and disappointing retail sales. However, both income growth and retail sales have moved higher in recent months, and this should soon translate into accelerating tax receipts and lower net borrowing. Longer term, significant risks remain for the Muni market.6 Chief among them is that state & local government budgets now finally look healthy enough to increase investment spending. Not to mention the significant uncertainty surrounding the potential for lower federal tax rates and plans to invest in infrastructure. Treasury Curve: Favor 5-Year Bullet Over 2/10 Barbell Chart 7Treasury Yield Curve Overview Treasury Yield Curve Overview Treasury Yield Curve Overview The Treasury curve underwent a small parallel shift upward in March, roughly split between a bear-steepening leading up to the FOMC meeting on March 15 and a bull-flattening from the meeting until the end of the month. Overall, the 2/10 Treasury slope flattened 1 basis point on the month and the 5/30 slope ended the month 1 bp steeper. Our recommended position long the 5-year bullet and short the 2/10 barbell - designed to profit from a steeper yield curve - returned +3 bps in March and is up 7 bps since inception on December 20. In addition, we also entered a short January 2018 fed funds futures trade on March 21.7 The performance of this trade has so far been flat. In a recent report,7 we identified the main cyclical drivers of the slope of the yield curve as: The fed funds rate (higher fed funds rate = flatter curve) Inflation expectations (higher inflation expectations = steeper curve) Interest rate volatility (higher volatility = steeper curve) Unit labor costs (higher unit labor costs = flatter curve) We concluded that even though the Fed is in the process of lifting the funds rate, the yield curve likely has room to steepen as long-maturity TIPS breakevens recover to levels more consistent with the Fed's inflation target (Chart 7). In addition, interest rate volatility has likely bottomed for the cycle and the uptrend in unit labor costs could level-off if productivity growth continues to rebound. The recent decline in bullish sentiment toward the dollar has also not yet been matched by a steeper 5/30 slope (bottom panel). TIPS: Overweight Chart 8TIPS Market Overview TIPS Market Overview TIPS Market Overview TIPS outperformed the duration-equivalent nominal Treasury index by 6 basis points in March. The 10-year TIPS breakeven rate declined 5 bps on the month and, at 1.97%, it remains well below its pre-crisis trading range of 2.4% to 2.5%. While the catalyst for the recent softening in TIPS outperformance seems to be the hawkish re-rating of Fed rate hike expectations, the uptrend in TIPS breakevens was probably due for a pause in any case. Breakevens had become stretched relative to our TIPS Financial Model - based on the dollar, oil prices and the stock-to-bond total return ratio. However, measures of pipeline inflation pressure - such as the ISM prices paid survey (Chart 8) - still point toward wider breakevens and, as was noted on the front page of this report, recent core inflation prints have been quite strong. All in all, growth appears strong enough that core inflation should continue its gradual uptrend and, more importantly, the Fed will be keen to accommodate an increase in both realized core inflation and TIPS breakevens, which remain below target. This means that in the absence of a material growth slowdown, long-maturity TIPS breakevens should continue to trend higher until they reach the 2.4% to 2.5% range that historically has been consistent with the Fed's inflation target. In a baseline scenario where the unemployment rate is 4.7% at the end of the year and the dollar remains flat, our Phillips curve model8 predicts that year-over-year core PCE inflation will be 2.02% at the end of this year. ABS: Maximum Overweight Overweight Chart 9ABS Market Overview ABS Market Overview ABS Market Overview Asset-Backed Securities outperformed the duration-equivalent Treasury index by 17 basis points in March, bringing year-to-date excess returns up to +22 bps. Aaa-rated issues outperformed the Treasury benchmark by 16 bps on the month, and non-Aaa issues outperformed by 26 bps. The index option-adjusted spread for Aaa-rated ABS tightened 5 bps on the month. At 48 bps, the spread remains well below its average pre-crisis level (Chart 9). Banks are now tightening lending standards on both auto loans and credit cards. While we do not expect this recent development to have much of an impact on consumer spending, it is usually an indication that there is growing concern about ABS collateral credit quality. As such, this week we scale back our recommended allocation to ABS from maximum overweight (5 out of 5) to overweight (4 out of 5). While credit card charge-offs remain well below pre-crisis levels, net losses on auto loans have started to trend higher (bottom panel). We continue to favor Aaa-rated credit cards over Aaa-rated auto loans, despite the modest spread advantage in autos (panel 3). Further, the spread advantage in Aaa consumer ABS relative to other high-quality spread product is becoming less compelling. Aaa ABS now only provide a 12 bps option-adjusted spread cushion relative to conventional 30-year Agency MBS and offer a slightly lower spread than Agency CMBS. Non-Agency CMBS: Underweight Chart 10CMBS Market Overview CMBS Market Overview CMBS Market Overview Non-agency commercial mortgage-backed securities underperformed the duration-equivalent Treasury index by 10 basis points in March, dragging year-to-date excess returns down to +16 bps. The index option-adjusted spread for non-agency Aaa-rated CMBS widened 4 bps on the month, but remains below its average pre-crisis level. Commercial real estate prices are still growing strongly, and prices in both major and non-major markets have re-gained their pre-crisis peaks (Chart 10). However, lending standards are tightening and, more recently, loan demand has rolled over (panel 4). This suggests that credit risk is starting to increase in commercial real estate, as do CMBS delinquencies which have put in a bottom (panel 5). Agency CMBS: Overweight Agency CMBS outperformed the duration-equivalent Treasury index by 14 basis points in March, bringing year-to-date excess returns up to +16 bps. The index option-adjusted spread for Agency CMBS tightened 2 bps on the month, and currently sits at 53 bps. The option-adjusted spread on Agency CMBS looks attractive compared to other high-quality spread product: Agency MBS = 36 bps, Aaa consumer ABS = 48 bps, Agency bonds = 18 bps and Supranationals = 22 bps. We continue to recommend an overweight position in Agency CMBS. Treasury Valuation Chart 11Treasury Fair Value Models Treasury Fair Value Models Treasury Fair Value Models The current reading from our 2-factor Treasury model (which is based on Global PMI and dollar sentiment) places fair value for the 10-year Treasury yield at 2.54% (Chart 11). Our 3-factor version of the model, which also incorporates the Global Economic Policy Uncertainty Index, places fair value at 2.28%. The lower fair value results from the large spike in the uncertainty index last November, which has only been partially unwound (bottom panel). Large spikes in uncertainty that do not coincide with deterioration in other economic indicators tend to mean revert fairly quickly. So we are inclined to view the fair value reading from our 2-factor model as more indicative of true fair value at the moment. For further details on our Treasury models please refer to the U.S. Bond Strategy Weekly Report, "The Message From Our Treasury Model", dated October 11, 2016, available at usbs.bcaresearch.com At the time of publication the 10-year Treasury yield was 2.35%. 1 For further detail on the medium-term profit outlook please see The Bank Credit Analyst, February 207, dated January 26, 2017, available at bca.bcaresearch.com 2 Please see U.S. Bond Strategy Weekly Report, "Keep Buying Dips", dated March 8, 2017, available at usbs.bcaresearch.com 3 A composite of 104 economic indicators designed to capture the probability of PCE inflation exceeding 2.5% during the subsequent 12 months. https://research.stlouisfed.org/publications/economic-synopses/2015/11/06/introducing-the-st-louis-fed-price-pressures-measure 4 Please see U.S. Bond Strategy Weekly Report, "Keep Buying Dips", dated March 28, 2017, available at usbs.bcaresearch.com 5 Please see U.S. Bond Strategy Weekly Report, "The Road To Higher Vol Is Paved With Uncertainty", dated February 14, 2017, available at usbs.bcaresearch.com 6 Please see U.S. Bond Strategy Special Report, "Trading The Municipal Credit Cycle", dated October 18, 2016, available at usbs.bcaresearch.com 7 Please see U.S. Bond Strategy Weekly Report, "The Yield Curve On A Cyclical Horizon", dated March 21, 2017, available at usbs.bcaresearch.com 8 Please see U.S. Bond Strategy Weekly Report, "The Odds Of March", dated February 21, 2017, available at usbs.bcaresearch.com Fixed Income Sector Performance Recommended Portfolio Specification Corporate Sector Relative Valuation And Recommended Allocation Total Return Comparison: 7-Year Bullet Versus 2-20 Barbell (6-Month Investment Horizon)

Highlights Spread Product: Any near-term correction in risk assets is likely to be fleeting. Investors should take the opportunity to increase credit exposure and maintain overweight spread product allocations on a 6-12 month horizon. Duration: Our 2-factor Global PMI model pegs fair value for the 10-year Treasury yield at 2.54%. Economy: U.S. economic growth will remain solidly above-trend this year, helped along by renewed strength in both residential and non-residential investment. Above-trend growth will ensure that inflation remains in its current gradual uptrend. Feature Chart 1Back Above 400 bps Back Above 400 bps Back Above 400 bps The reflation trade has come under question during the past couple of weeks. The S&P 500 is 1.7% off its recent high, the VIX has bounced and the average spread on the Bloomberg Barclays High-Yield index is back above 400 basis points (Chart 1). After such a move, it is reasonable to ask if the economic landscape has changed enough to warrant a reversal of our current overweight spread product allocation. We think not, and we advise investors to buy the dips, adding credit risk to their portfolios from more attractive levels. This week we examine why risk assets are vulnerable to a near-term correction, but also why these corrections are likely to be short lived. On a 6-12 month investment horizon we continue to recommend a pro-risk portfolio characterized by: below-benchmark duration, overweight spread product, curve steepeners and TIPS breakeven wideners. Three Catalysts For A Near-Term Sell Off... Three main factors suggest that risk assets might continue to correct in the near-term. The first is that Fed rate hike expectations might be increasing too quickly. Chart 2 shows the fed funds rate that is priced into the overnight index swap curve for the end of this year. The lower dashed horizontal line is the level consistent with one more rate hike between now and the end of the year. The higher dashed horizontal line is the level consistent with two more rate hikes between now and the end of the year. We see that risk assets were able to handle the shift in rate expectations up to the lower dashed line with no trouble. The yield curve steepened and the cost of inflation compensation rose (Chart 2, bottom panel). But now, as rate expectations approach the higher dashed line, the reflation trade is starting to fray. The yield curve has started to flatten and TIPS breakevens are rolling over. A second reason why risk assets might sell-off in the near-term is the still elevated level of economic policy uncertainty (Chart 3, top panel). Last Friday, markets hung on every word related to the likelihood of a new healthcare bill being passed. Now that the bill has failed, attention will turn quickly to tax reform. It is very likely that risk assets will suffer if it appears as though tax reform will be delayed or scrapped altogether. Importantly, it is the opinion of our Geopolitical Strategy service that tax reform will be passed before the end of the year.1 Chart 2How Much Hawkishness Can Markets Take? How Much Hawkishness Can Markets Take? How Much Hawkishness Can Markets Take? Chart 3Correction Catalysts? Correction Catalysts? Correction Catalysts? A third reason why risk assets are vulnerable to a near-term correction is that investors have bought into the reflation trade, and sentiment is extremely bullish (Chart 3, bottom panel). Surveys of investors conducted by Yale University show that 99% of investors expect the Dow to increase during the coming year, while simultaneously only 47% of investors characterize the stock market as "not too high" relative to its fundamental value. The divergence in itself suggests that the equity rally is built on a shaky foundation. It seems likely that either confidence needs to wane or valuations need to correct for the rally to be prolonged. ...But The Fed Cycle Trumps Them All In previous reports2 we outlined the four phases of the Fed Cycle (see Box), and observed that in all likelihood we are currently in Phase I. Box: The Four Phases Of The Fed Cycle Chart 4Stylized Fed Cycle Keep Buying Dips Keep Buying Dips The four phases of the Fed Cycle are illustrated in Chart 4 and defined as follows: Phase I represents the early stage of the withdrawal of monetary stimulus. This phase begins with the first hike of a new tightening cycle and ends when the fed funds rate crosses above its equilibrium (or neutral) level. Phase II represents the late stage of the tightening cycle, when the Fed hikes its target rate above equilibrium in an effort to slow the economy. Phase III represents the early stage of the easing cycle. It begins with the first rate cut from the peak and lasts until the Fed cuts its target rate below equilibrium. Phase IV represents the late stage of the easing cycle. It encompasses both the period when the fed funds rate descends to its cycle trough and the subsequent adjustment period when the Fed remains on hold in an effort to kick start an economic recovery. In Phase I, the Fed has begun to remove monetary accommodation but still needs inflation to rise back to target. In other words, if risk assets sell off and financial conditions start to tighten the Fed will adopt a more dovish policy stance to ensure that the recovery persists and inflation continues to trend higher. We note that core PCE inflation is running at 1.74% year-over-year, still below the Fed's 2% target. Further, the St. Louis Fed Price Pressures Measure3 is signaling only a 19% chance that PCE inflation will exceed 2.5% during the next twelve months, and market-based measures of inflation compensation are well below levels that are consistent with the Fed's inflation target (Chart 5). Chart 5Fed Still Needs Higher Inflation Fed Still Needs Higher Inflation Fed Still Needs Higher Inflation In this environment, if risk assets sell off because of overly aggressive rate hike expectations, fiscal policy disappointments or over-extended sentiment, the Fed will quickly adopt a more dovish policy stance, lending support to the reflation trade. Of course, if any of the catalysts for the market correction also cause a severe contraction in economic growth, then the reflation trade would face a more lasting setback. However, none of the three reasons for a market correction listed above seem likely to have significant pass-through effects on the economy. Even if fiscal stimulus turns out to be much less than was previously anticipated, there appears to be sufficient momentum in economic growth to maintain inflation on its upward trajectory (see section titled "Above-Trend Growth: Aided By Housing & Capex" below). It follows from this analysis of the Fed Cycle that a strategy of "buying the dips" should work whenever we are in an environment where the Fed needs inflation to move higher. It is only when inflation is more firmly anchored around the Fed's target that the Fed will be less willing to support markets, making a "buy the dips" strategy less effective. To test this theory, we devised a trading rule for high-yield bonds where we buy the High-Yield index whenever spreads widen by 20 bps or more during a month. We then hold that position for a period ranging from 1 to 3 months and calculate excess returns relative to duration-matched Treasuries during that period. Our goal is to see if the effectiveness of this "buy the dips" strategy differs depending on the stage of the Fed Cycle. For this test we define the stages of the Fed Cycle using the aforementioned St. Louis Fed Price Pressures Measure, which we split into four ranges: 0% to 15%: An environment of very limited inflation pressure most consistent with Phase IV of the Fed Cycle. 15% to 30%: Still muted inflation pressures. Roughly consistent with Phase I of the Fed Cycle. 30% to 50%: Rising inflation pressures, but still less than a 50% chance that PCE will exceed 2.5% in the coming 12 months. This likely coincides with some Phase I periods and some Phase II periods of the Fed Cycle. 50% to 70%: Strong inflation pressures, and a good chance of inflation overshooting the Fed's target. Most likely coincides with Phase II or Phase III of the Fed Cycle. We indeed find that a "buy the dips" strategy is more effective when inflation pressures are lower (Table 1). A strategy of buying the junk index after spreads widen by at least 20 bps and holding it for three months produces positive excess returns 65% of the time when the St. Louis Fed Price Pressures Measure is between 0% and 15%. This same strategy works 59% of the time when the Price Pressures Measure is between 15% and 30%, 44% of the time when the Measure is between 30% and 50% and only 25% of the time when the Measure is between 50% and 70%. Table 1High-Yield Corporate Bond Returns* Achieved By Holding The Junk Index Following ##br##A 20 BPs Widening In High-Yield Corporate OAS** Under Different Ranges##br## Of The St. Louis Fed Price Pressure Measure*** (February 1994 To Present) Keep Buying Dips Keep Buying Dips With the Price Pressures Measure at only 19% currently, we advise investors to increase exposure to spread product on any near-term correction. Bottom Line: Any near-term correction in risk assets is likely to be fleeting. Investors should take the opportunity to increase credit exposure and maintain overweight spread product allocations on a 6-12 month horizon. Above-Trend Growth: Aided By Housing & Capex For the analysis of the Fed cycle performed above to be applicable, we must have confidence in the view that GDP will continue to grow at an above-trend pace. That is, growth must at least be strong enough to remove slack from the labor market and cause inflation to trend gradually higher. This has mostly been the case since measures of core inflation bottomed in early 2015 and we see no evidence at the moment to suggest it is about to change. In fact, measures of global growth most relevant for Treasury yields have hooked up strongly in recent months, and our model now suggests that fair value for the 10-year U.S. Treasury yield is 2.54% (Chart 6). At the time of publication the 10-year yield was 2.40%. The fair value reading from our model moved higher during the past month even though PMIs in both the U.S. and Japan ticked down. This negative move was offset by an acceleration in Eurozone PMI and a decline in bullish sentiment toward the dollar (Chart 6, bottom two panels). Less bullish dollar sentiment is a signal that the global recovery is becoming more synchronized which means that U.S. Treasury yields must rise more quickly for a given level of global growth.4 Returning to the U.S. growth outlook specifically, a recent BCA Special Report 5 showed that cyclical spending as a percent of overall GDP is an excellent leading indicator of economic downturns (Chart 7). Cyclical spending has been relatively firm as a percent of GDP during the past couple of years, and would have been stronger if not for stagnant residential investment (Chart 7, panel 3) and contracting non-residential investment in equipment & software (Chart 7, bottom panel). However, leading indicators suggest that both of these factors should shift from being sources of disappointment to sources of strength in the coming months. Chart 610-Year Treasury Fair Value Model 10-Year Treasury Fair Value Model 10-Year Treasury Fair Value Model Chart 7Cyclical Spending Is Firm... Cyclical Spending Is Firm... Cyclical Spending Is Firm... Chart 8 shows the year-over-year change in each of the three cyclical components of GDP as a percent of overall growth alongside a reliable leading indicator. Consumer confidence suggests that consumer spending on durables will remain firm (Chart 8, panel 1). Our composite indicator of New Orders surveys also points to a rebound in nonresidential investment on equipment & software (Chart 8, panel 2). In prior reports we observed that nonresidential investment was held back by the 2014 oil price shock and should recover now that oil prices have found a floor.6 Also, any potential benefit from a more favorable tax and regulatory environment under the new federal government would only increase the upside for capex. Residential investment as a percent of GDP also rolled over last year, but homebuilder confidence has been trending sharply higher during the past few months (Chart 8, bottom panel). Home construction will be strong this year, despite the recent increase in mortgage rates. As was recently observed by our U.S. Investment Strategy service,7 the constraint on housing demand since the financial crisis has not come from un-affordable monthly mortgage payments. In fact, we calculate that even if mortgage rates rise by another 200 bps from current levels, the mortgage payment as a percent of income for the median household would still be below its long-run average (Chart 9). Chart 8...And Likely To Increase ...And Likely To Increase ...And Likely To Increase Chart 9Higher Rates Won't Kill Housing Higher Rates Won't Kill Housing Higher Rates Won't Kill Housing Rather, the constraint on housing demand has come from insufficient savings on the part of potential first time homebuyers relative to required down payments. This constraint can only subside as household savings increase and mortgage lending standards ease, two trends that are ongoing. Finally, housing supply is approaching historically low levels relative to demand (Chart 9, bottom panel) even including the "shadow inventory" from foreclosed properties which has now mostly vanished in any case. With supply at such depressed levels and demand likely to remain firm, it is no wonder that homebuilders are feeling more confident. Bottom Line: U.S. economic growth will remain solidly above-trend this year, helped along by renewed strength in both residential and non-residential investment. Above-trend growth will ensure that inflation remains in its current gradual uptrend. Ryan Swift, Vice President U.S. Bond Strategy rswift@bcaresearch.com 1 Please see Geopolitical Strategy Weekly Report, "Donald Trump Is Who We Thought He Was", dated March 8, 2017, available at gps.bcaresearch.com 2 Please see U.S. Bond Strategy Weekly Report, "Inflation: More Fire Than Ice, But Don't Sound The Alarm", dated January 24, 2017, available at usbs.bcaresearch.com 3 A composite of 104 economic indicators designed to capture the probability of PCE inflation exceeding 2.5% during the subsequent 12 month period. https://research.stlouisfed.org/publications/economic-synopses/2015/11/06/introducing-the-st-louis-fed-price-pressures-measure 4 A more detailed explanation of the inverse relationship between dollar sentiment and Treasury yields can be found in the U.S. Bond Strategy Weekly Report, "Dollar Watching: Another Update", dated January 31, 2017, available at usbs.bcaresearch.com 5 Please see BCA Special Report, "Beware The 2019 Trump Recession", dated March 7, 2017, available at usbs.bcaresearch.com 6 Please see U.S. Bond Strategy Special Report, "Seven Fixed Income Themes For 2017", dated December 20, 2016, available at usbs.bcaresearch.com 7 Please see U.S. Investment Strategy Special Report, "U.S. Housing: What Comes Next?", dated March 27, 2017, available at usis.bcaresearch.com Fixed Income Sector Performance Recommended Portfolio Specification

Highlights U.S. equity valuations are historically high, based on a variety of metrics. However, relative to competing assets and global equity peers, U.S. stock valuations are not an extreme. For U.S.-based investors, our upbeat view on the U.S. dollar implies that efforts to diversify globally may come up short. The Fed will allow its Agency bond and MBS portfolios to run off starting in 2018, but it is much more uncertain whether it will ever cease the reinvestment of its Treasury holdings. Feature The S&P 500 remains near record highs, despite a modest setback last week. And the only period when stocks were more expensive was during the halcyon days of the dot-com bubble. Have stock prices outpaced fundamentals, and if so, how much of a risk does this present over the cyclical horizon? And should U.S. investors look further afield than domestic markets for a relative deal? On a historical basis, it is hard to argue that U.S. equities are anything other than expensive. A preferred valuation metric is the cyclically adjusted P/E ratio (CAPE), see Chart 1. Based on this metric, stocks are expensive, trading at their highest valuation outside of the dot-com bubble. However, valuing equities is a complicated issue, and the CAPE is not without its weaknesses. Examining a broad array of valuation indicators provides a slightly different message; U.S. stocks are expensive in absolute terms based on historic relationships, but are less stretched relative to both other asset classes and other equity indexes. Expensive, But... Our BCA valuation index captures the message from a broad range of metrics in one gauge (Chart 2). The valuation index was constructed using 11 different measures in an attempt to approach valuation from multiple angles. Decomposing the index into its three major components - earnings, balance sheet metrics and yield - show that stocks prices are well into expensive territory in absolute terms based on traditional fundamentals: Chart 1(Part I) U.S. Stocks Are Expensive ##br##Relative To History (Part I) U.S. Stocks Are Expensive Relative To History (Part I) U.S. Stocks Are Expensive Relative To History Chart 2(Part II) U.S. Stocks Are Expensive Relative ##br##To History (Part II) U.S. Stocks Are Expensive Relative To History (Part II) U.S. Stocks Are Expensive Relative To History Earnings Group: There are five inputs to the earnings component of our valuation indicator, including trailing price/earnings ratio, price/sales, market cap as a share of GDP. The second panel of Chart 2 shows that the aggregate of the Earnings Group indicators sits at historical highs, excluding the tech bubble. Balance Sheet Metrics: This component includes measure of the market value of equities relative to corporate net worth, both using market value (replacement cost) and historical cost. This measure of valuation has the same profile as the Earnings Group. Yield Group: The yield group compares the price of stocks to interest rates, nominal and real, government and corporate. Of the three groups, it is this Yield Group that gives a less expensive reading on equities (bottom panel of Chart 2). Overall, the Valuation Indicator is already well into "overvalued territory". There is only one episode since 1970 when the indicator has reached a significantly more extreme reading (the dot-com bubble). ...Not So On A Relative Basis Stocks are expensive on an absolute basis, but are far more appealing in relative terms. The current earnings yield on stocks is well above the real corporate bond yield and corporate bond spreads are historically very tight, despite the erosion in balance sheet health (our corporate health monitor has been deteriorating for several months). And compared to housing returns, stocks look downright cheap (Chart 3). Within the U.S., we expect stocks to be the biggest beneficiary of investment flows in the next year or two, in part because equity market value is the most appealing. Meanwhile, relative to global peers, U.S. equities valuations have been climbing since 2009 (Charts 4 and Chart 5). This eight-year rise in valuations now leaves U.S. P/Es at the higher end of the historical range relative to G10 ex-U.S. equities. U.S. stocks are especially expensive relative to Japanese equities. In any case, standard valuation measures have always been lower in Japan, with the exception of price-forward earnings. As our Bank Credit Analyst monthly publication points out, Japanese companies generally have a much higher interest coverage ratio compared to Corporate America. Nonetheless, they tend to come up short in terms of profitability. Operating margins in the U.S. have typically been double that of Japan. Japan's return-on-equity (RoE) has been dismal because of low levels of corporate leverage and loads of low-yielding cash sitting on balance sheets. Nonetheless, the valuation gap is at an extreme, with Japanese equities appearing to be a screaming value relative to U.S. stocks. Chart 3Stocks Look Less Expensive Relative To Competing Assets Stocks Look Less Expensive Relative To Competing Assets Stocks Look Less Expensive Relative To Competing Assets Chart 4(Part I) U.S. Outperformance Phase Can Continue (Part I) U.S. Outperformance Phase Can Continue (Part I) U.S. Outperformance Phase Can Continue Chart 5(Part II) U.S. Outperformance Phase Can Continue (Part II) U.S. Outperformance Phase Can Continue (Part II) U.S. Outperformance Phase Can Continue A similar, albeit less extreme, valuation case can be made for European stocks relative to the U.S. Eurozone stocks have also almost always traded at a discount to U.S. equities and this continues to be the case. Stocks have gotten even more expensive, more quickly, in the U.S. over the past year. But relative valuations are not near historic extremes. Tack on the fact that BCA's view is that the dollar will continue to appreciate over the next six-twelve months. For U.S.-based investors, the coming rise in the domestic currency implies that efforts to diversify globally may come up short, despite better value in major foreign markets. It is important to note that BCA does not view valuation measures as market timing tools. They are only useful at extremes. The bottom line is that U.S. equities are certainly far from cheap, but are not so expensive in relative terms to warrant an allocation change on this basis. We believe that equity returns should outperform Treasuries, cash and high-quality corporate bonds over the next 1-2 years as the bond bear market plays out. The Fed's Balance Sheet: What's Next? Recently we have received a number of client questions about the Fed's balance sheet and how it will evolve during the next few years. In response, we reprint below work from our U.S. Bond Strategy team, who recently addressed the topic in detail. The Fed's Stated Plan The most up-to-date guidance we have received about the Fed's balance sheet plans comes from Janet Yellen's recent Congressional testimony: "The FOMC has annunciated that its longer run goal is to shrink our balance sheet to levels consistent with the efficient and effective implementation of monetary policy. And while our system evolves and I can't put a number on that, I would anticipate a balance sheet that's substantially smaller than at the current time. In addition, we would like our balance sheet to again be primarily Treasury securities, whereas as you pointed out, we have substantial holdings of mortgage-backed securities." From this, and similar statements from other Fed officials, we conclude that the Fed will allow its balance sheet to shrink once the fed funds rate is somewhere in the range of 1% to 1.5%. Surveys also show that the median primary dealer expects the Fed will change its balance sheet policy when the target fed funds rate is 1.38%. As such, and under reasonable assumptions for the pace of rate hikes, we think it is very likely that the Fed will start to let its balance sheet shrink sometime in 2018. MBS First, Treasuries Maybe Later Yellen's statement to Congress also makes clear that the Fed would be more comfortable with a balance sheet that consists entirely of Treasury securities. For this reason, the central bank will start by simply ceasing the reinvestment of its Agency bond and MBS portfolios. At least initially, the Fed will continue to reinvest the proceeds from its maturing Treasury portfolio. Yellen also left open the possibility that reinvestment could be "tapered" rather than just halted altogether. While this is possible, and in fact 70% of primary dealers think that reinvestments will be phased out over time while only 14% think they will be ceased all at once, it seems to us like a needless complication. We expect that reinvestments of Agency bonds and MBS will end all at once sometime in 2018. As for the Fed's holdings of Treasury securities, it is much less clear whether the Fed will allow these balances to run down. In a Report in 2014,1 we describe in detail the differences between the Fed's pre-crisis mode of operation, when bank reserves were scarce, and the Fed's current mode of operation with large bank reserve balances. As of now, the Fed has stated that it intends to eventually drain bank reserves from the system and return to its pre-crisis mode of operation, but there are several possible advantages to running a system with an outsized Fed balance sheet and large bank reserve balances. None other than Ben Bernanke pointed out a few of those reasons in a blog post last fall.2 In our view, the most compelling is that regulatory changes have increased private sector demand for safe, short-maturity, liquid assets in recent years. If the Treasury department is unwilling to supply T-bills in sufficient numbers, then the Fed can supply safe, short-maturity, liquid assets to the market by purchasing long-maturity Treasury securities and replacing them with bank reserves. Chart 6Reserves Can Be Drained Fairly Quickly Reserves Can Be Drained Fairly Quickly Reserves Can Be Drained Fairly Quickly Of course, we take the Fed at its word when it says that it would like to eventually drain excess bank reserves from the system. But even in that case, the steady growth of currency in circulation means that bank reserves will decline over time even if the Fed keeps the asset side of its balance sheet flat. For example, Chart 6 shows what would happen to bank reserves if the amount of currency in circulation grows at a conservative 5% per year pace, and if the Fed decides to allow its Agency bond and MBS portfolios to run off at the beginning of next year while keeping its Treasury portfolio flat. We assume that MBS runs off the Fed's balance sheet at a pace of $15 billion per month, slightly below the recent pace of MBS reinvestment. During the past three years, the Fed has reinvested between $20bn and $40bn MBS each month with an average monthly reinvestment of $32bn. In this scenario, outstanding bank reserves would decline to zero by the end of 2025. At that point the Fed would have to start adding to its Treasury holdings just to keep pace with the amount of currency in circulation. Bottom Line: While it is very likely that the Fed will allow its Agency bond and MBS portfolios to run off starting in 2018, it is much more uncertain whether it will ever cease the reinvestment of its Treasury holdings. If the Fed does allow its Treasury holdings to run down as well, it will have to start buying Treasuries again before 2025. Lenka Martinek, Vice President U.S. Investment Strategy lenka@bcaresearch.com 1 Please see U.S. Bond Strategy Special Report "Cleaning Up After The 100-Year Flood", dated June 10, 2014, available at usbs.bcaresearch.com. 2 https://www.brookings.edu/blog/ben-bernanke/2016/09/02/should-the-fed-keep-its-balance-sheet-large/

Highlights Chart 1Keep A Close Eye On Financial Conditions Keep A Close Eye On Financial Conditions Keep A Close Eye On Financial Conditions The market's rate hike expectations moved sharply higher during the past two weeks as a string of Fed speeches, including one by Chair Yellen, all but confirmed a March rate hike. The market is now priced for 75 basis points of hikes during the next 12 months, compared to 50 bps at the end of January. At least so far, broad indicators of financial conditions have not tightened in response to this re-rating of the Fed (Chart 1). However, there are some preliminary indications that the reflation trade is fraying at the edges. The trade-weighted dollar has appreciated +0.2% since the end of January, the 2/10 Treasury slope has flattened 9 bps and the 10-year TIPS breakeven inflation rate has declined 1 bp. The Fed is currently testing the markets with hawkish rhetoric but, with inflation and TIPS breakevens still below target, will ultimately support the reflation trade if it comes under threat. In this environment investors with 6-12 month investment horizons should maintain below-benchmark duration, remain overweight spread product and continue to position for a steeper curve and wider TIPS breakevens. Feature Chart 2Investment Grade Market Overview Investment Grade Market Overview Investment Grade Market Overview Investment Grade: Overweight Investment grade corporate bonds outperformed the duration-equivalent Treasury index by 48 basis points in February. The index option-adjusted spread tightened 6 bps on the month and, at 112 bps, it remains well below its historical average (134 bps). Our research1 shows that when core PCE inflation is between 1.5% and 2%2 investment grade corporate bonds produce an average monthly excess return of close to zero. A 90% confidence interval places monthly excess returns between -19 bps and +17 bps with inflation in this range and excess returns do not turn decisively negative until core PCE is above 2%. Given the Fed's desire to nurture a continued recovery in inflation, we are not worried about significant spread widening until inflation is sustainably above 2%. In the meantime we expect corporate bond excess returns to be low, but positive. While supportive monetary policy should ensure excess returns consistent with carry, investors should not bank on further spread compression as corporate spreads have already discounted a substantial improvement in leverage (Chart 2). Energy related sectors still appear cheap after adjusting for differences in credit rating and duration (Table 3), and our commodity strategists expect oil prices to remain firm even in the face of a stronger U.S. dollar. This week we upgrade the Wireless and Packaging sectors from underweight to neutral and downgrade the Consumer Cyclical Services sector from neutral to underweight. The former two sectors now appear cheap on our model, while the latter has become expensive. Table 3ACorporate Sector Relative Valuation And Recommended Allocation* How Much Can Markets Take? How Much Can Markets Take? Table 3BCorporate Sector Risk Vs. Reward* How Much Can Markets Take? How Much Can Markets Take? High-Yield: Neutral Chart 3High-Yield Market Overview High-Yield Market Overview High-Yield Market Overview High-Yield outperformed the duration-equivalent Treasury index by 112 basis points in February. The index option-adjusted spread tightened 25 bps on the month and, at 349 bps, it is currently 170 bps below its historical average. One of our key investment themes3 for this year is that the uptrend in defaults is likely to reverse (Chart 3), mostly due to recovery in the energy sector. However, still-poor corporate health and tightening monetary policy will lead to a resumption of the uptrend in 2018 and beyond. Despite the positive outlook for defaults, we retain only a neutral allocation to High-Yield because of very tight valuations. The index option-adjusted spread is now within a hair of the average level of 340 bps that prevailed during the 2004 - 2006 Fed tightening cycle, when indicators of corporate balance sheet health were in much better shape. In fact, the index spread is now only 116 bps wider than its all-time low of 233 bps, reached in 2007. Our preferred measure of High-Yield valuation is the default-adjusted spread - the average spread of the junk index less our forecast of 12-month default losses. At present, the default-adjusted spread is 142 bps. Historically, a default-adjusted spread between 100 bps and 150 bps is consistent with positive excess returns during the subsequent 12 months 64% of the time. It is only when the default-adjusted spread falls below 100 bps that positive excess returns become unlikely. Junk has provided positive excess returns over a 12-month horizon only 13% of the time when the starting default-adjusted spread is between 50 bps and 100 bps. MBS: Underweight Chart 4MBS Market Overview MBS Market Overview MBS Market Overview Mortgage-Backed Securities outperformed the duration-equivalent Treasury index by 3 basis points in February. The conventional 30-year MBS yield fell 5 bps on the month, driven by a 7 bps decline in the rate component. The compensation for prepayment risk (option cost) increased by 1 bp, as did the option-adjusted spread. MBS spreads remain extremely tight relative both to history and Aaa-rated credit, although they have begun to widen somewhat relative to credit in recent weeks (Chart 4). More distressing is that the nominal MBS spread appears too tight relative to interest rate volatility (bottom panel). As we noted in a recent report,4 the long-run trend in interest rate volatility tends to be driven by uncertainty about the macroeconomic and political environment. In fact, rate volatility can be modeled using forecaster disagreement about GDP growth and T-bill rates. While the Fed's policy of forward guidance and a fed funds rate pinned at zero limited the amount of forecaster disagreement in recent years, this disagreement will re-emerge the further the fed funds rate moves off its lower bound. Another medium-term risk for MBS comes from the Fed ending the reinvestment of its MBS portfolio. As we described in a recent Special Report,5 the Fed is likely to allow its MBS portfolio to shrink at some point in 2018, putting further upward pressure on MBS spreads. Government Related: Underweight Chart 5Government-Related Market Overview Government-Related Market Overview Government-Related Market Overview The government-related index outperformed the duration-equivalent Treasury index by 30 basis points in February, bringing year-to-date excess returns up to +51 bps. The high-beta Sovereign and Foreign Agency sectors outperformed the Treasury benchmark by 90 bps and 59 bps, respectively. Meanwhile, the low-beta Domestic Agency and Supranational sectors each outperformed by 4 bps. Local Authorities returned 24 bps in excess of duration-matched Treasuries. Sovereigns have outperformed Baa-rated corporate bonds year-to-date, a trend consistent with the rise in commodity prices and a trade-weighted dollar that has weakened by 1.5% (Chart 5). However, the dollar has started to appreciate in recent weeks and probably has further upside in the medium-term, especially if the Fed maintains its hawkish posture. Historically, it has been very rare for Sovereigns to outperform U.S. corporate bonds when the dollar is appreciating. After adjusting for credit rating and duration, the Foreign Agency and Local Authority sectors continue to appear cheap relative to U.S. corporate credit. In contrast, Sovereigns, Supranationals and Domestic Agencies all appear expensive. We continue to recommend overweight allocations to Foreign Agencies and Local Authorities, alongside underweight allocations to the rest of the government-related index. In a television interview last month Treasury Secretary Steven Mnuchin confirmed that GSE reform is still a priority for the new administration but that tax reform is much higher on the agenda. This means that agency spreads will likely remain insulated from any "reform risk" until next year at the earliest. Municipal Bonds: Neutral Chart 6Municipal Market Overview Municipal Market Overview Municipal Market Overview Municipal bonds outperformed the duration-equivalent Treasury index by 13 basis points in February (before adjusting for the tax advantage). The average Municipal / Treasury (M/T) yield ratio has fallen 4% since the end of January and remains firmly anchored below its post-crisis average. The decline in the average M/T yield ratio was concentrated in short maturities, while ratios at the long-end of the curve actually rose (Chart 6). Accelerating fund flows and falling issuance will continue to support yield ratios in the near term. In fact, our tactical yield ratio model - based on issuance, fund flows and ratings migration - shows that yield ratios are presently very close to fair value. Although the average M/T yield ratio still appears expensive if we include the global economic policy uncertainty index as an additional explanatory variable.6 One risk to Munis is that yield ratios have already discounted a substantial reduction in state and local government net borrowing in Q1 (panel 3). While we expect this improvement will materialize in the next few quarters, net borrowing is biased upward beyond this year based on the lagged relationship between corporate sector and state and local government health.7 Treasury Curve: Favor 5-Year Bullet Over 2/10 Barbell Chart 7Treasury Yield Curve Overview Treasury Yield Curve Overview Treasury Yield Curve Overview The Treasury curve has bear-flattened since the end of January as the market revised its Fed rate hike expectations sharply higher. Both the 2/10 and 5/30 Treasury slopes have flattened by 9 basis points since January 31. As such, our recommended position long the 5-year bullet and short the 2/10 barbell - designed to profit from a steeper yield curve - has returned -26 bps since the end of January, although it has returned close to 0 bps since it was initiated on December 20.8 As was stated on the front page of this report, the Fed's increasingly hawkish rhetoric has already caused the uptrend in TIPS breakevens to pause and the nominal Treasury slope to flatten (Chart 7). With inflation still below target these trends are not sustainable from the point of view of Fed policymakers. If the trend of decreasing TIPS breakevens and a flattening curve persists, we would expect the Fed to back away from its hawkish rhetoric. This dynamic will support a steeper yield curve at least until core PCE inflation is back to the Fed's 2% target and long-dated TIPS breakevens are anchored in a range between 2.4% and 2.5% (a range that is typically consistent with core PCE inflation at 2%). The persistent attractiveness of the 5-year bullet relative to the rest of the curve makes a position long the 5-year bullet and short a duration-matched 2/10 barbell the most attractive way to position for a steeper yield curve (panel 3). The carry buffer in the 5-year helps mitigate some of the risk of curve flattening. TIPS: Overweight Chart 8TIPS Market Overview TIPS Market Overview TIPS Market Overview TIPS underperformed the duration-equivalent Treasury index by 18 basis points in February. The 10-year TIPS breakeven rate declined 3 bps on the month and, at 2.04%, it remains well below its pre-crisis trading range of 2.4% to 2.5%. While the catalyst for the recent softening in TIPS outperformance seems to be the hawkish re-rating of Fed rate hike expectations, the uptrend in TIPS breakevens was probably due for a pause in any case. Breakevens had become stretched relative to our TIPS Financial Model - based on the dollar, oil prices and the stock-to-bond total return ratio (Chart 8). Diffusion indexes for both PCE and CPI inflation have also shifted into negative territory, suggesting that realized inflation readings will soften during the next couple of months. On a cyclical horizon, however, the Fed will be keen to allow breakevens to rise toward levels more consistent with its inflation target and will quickly adopt a more dovish stance if breakevens fall significantly. This "Fed put" should remain in place at least until core PCE inflation is firmly anchored around 2% and long-dated TIPS breakevens return to a range between 2.4% and 2.5%. As we detailed in a recent report,9 while accelerating wage growth will ensure that inflation remains in a long-run uptrend, the impact from wages will be mitigated by deflating import prices meaning that the uptrend will be slow. We continue to expect that year-over-year core PCE inflation will not attain the Fed's 2% target until the end of this year. ABS: Maximum Overweight Chart 9ABS Market Overview ABS Market Overview ABS Market Overview Asset-Backed Securities performed in-line with the duration-equivalent Treasury index in February. Aaa-rated issues underperformed the Treasury benchmark by 2 basis points, while non-Aaa issues outperformed by 12 bps. The index option-adjusted spread for Aaa-rated ABS widened 3 bps on the month. At 50 bps, the spread remains well below its average pre-crisis level. Banks are now tightening lending standards on both auto loans and credit cards (Chart 9). While we do not think this will have much of an impact on consumer spending,10 it is usually an indication that there is growing concern about ABS collateral credit quality. While credit card charge-offs remain well below their pre-crisis levels, net losses on auto loans have in fact started to trend higher (bottom panel). We continue to recommend Aaa-rated credit cards over Aaa-rated auto loans, despite the spread advantage in autos. We will closely monitor the evolving credit quality situation, but for now continue to view consumer ABS as a very attractive alternative to other short-duration Aaa-rated spread product such as MBS and Agency bonds. The main reason being the sizeable spread advantage that has persisted in ABS for some time. At present, Aaa-rated consumer ABS offer an option-adjusted spread of 50 bps, compared to 31 bps for 30-year conventional Agency MBS and 18 bps for Agency bonds. Non-Agency CMBS: Underweight Chart 10CMBS Market Overview CMBS Market Overview CMBS Market Overview Non-agency Commercial Mortgage-Backed Securities underperformed the duration-equivalent Treasury index by 34 basis points in February. The index option-adjusted spread for non-agency Aaa-rated CMBS widened 7 bps on the month, but remains below its average pre-crisis level (Chart 10). Rising CMBS delinquency rates and tightening commercial real estate lending standards make us cautious on non-agency CMBS. This caution has only intensified now that spreads are firmly entrenched below their pre-crisis average. Agency CMBS: Overweight Agency CMBS underperformed the duration-equivalent Treasury index by 20 basis points in February. The index option-adjusted spread for Agency CMBS widened 5 bps on the month, and currently sits at 53 bps. The spread offered on Agency CMBS is similar to what is offered by Aaa-rated consumer ABS (50 bps) and greater than what is offered by conventional 30-year MBS (31 bps) for a similar amount of spread volatility. We continue to recommend an overweight position in Agency CMBS. Treasury Valuation Chart 11Treasury Fair Value Models Treasury Fair Value Models Treasury Fair Value Models The current reading from our 2-factor Treasury model (which is based on Global PMI and dollar sentiment) places fair value for the 10-year Treasury yield at 2.42% (Chart 11). Our 3-factor version of the model, which also incorporates the Global Economic Policy Uncertainty Index, places fair value at 2.21%. The lower fair value results from the large spike in the uncertainty index last November, which has only been partially unwound (bottom panel). Large spikes in uncertainty that do not coincide with deterioration in other economic indicators tend to mean revert fairly quickly. So we would be inclined to view the fair value reading from our 2-factor model as more indicative of true fair value at the moment. For further details on our Treasury models please refer to the U.S. Bond Strategy Weekly Report, "The Message From Our Treasury Model", dated October 11, 2016, available at usbs.bcaresearch.com At the time of publication the 10-year Treasury yield was 2.49%. Ryan Swift, Vice President U.S. Bond Strategy rswift@bcaresearch.com 1 Please see U.S. Bond Strategy Weekly Report, "Inflation: More Fire Than Ice, But Don't Sound The Alarm", dated January 24, 2016, available at usbs.bcaresearch.com 2 Year-over-year core PCE inflation is currently 1.74%. 3 Please see U.S. Bond Strategy Special Report, "Seven Fixed Income Themes For 2017", dated December 20, 2016, available at usbs.bcaresearch.com 4 Please see U.S. Bond Strategy Weekly Report, "The Road To Higher Vol Is Paved With Uncertainty", dated February 14, 2017, available at usbs.bcaresearch.com 5 Please see U.S. Bond Strategy / Global Fixed Income Strategy Special Report, "The Way Forward For The Fed's Balance Sheet", dated February 28, 2017, available at usbs.bcaresearch.com 6 For further details on the model please see U.S. Bond Strategy Special Report, "Seven Fixed Income Themes For 2017", dated December 20, 2016, available at usbs.bcaresearch.com 7 For further details on the linkage between corporate sector health and state & local government health please see U.S. Bond Strategy Special Report, "Trading The Municipal Credit Cycle", dated October 18, 2016, available at usbs.bcaresearch.com 8 Please see U.S. Bond Strategy Special Report, "Seven Fixed Income Themes For 2017", dated December 20, 2016, available at usbs.bcaresearch.com 9 Please see U.S. Bond Strategy Weekly Report, "Inflation: More Fire Than Ice, But Don't Sound The Alarm", dated January 24, 2017, available at usbs.bcaresearch.com 10 Please see U.S. Bond Strategy Weekly Report, "The Odds Of March", dated February 21, 2017, available at usbs.bcaresearch.com Fixed Income Sector Performance Recommended Portfolio Specification Corporate Sector Relative Valuation And Recommended Allocation Total Return Comparison: 7-Year Bullet Versus 2-20 Barbell (6-Month Investment Horizon) Current Recommendation

Feature Recently we have received a number of client questions about the Fed's balance sheet. When will the Fed start to shrink its balance sheet (if at all)? If the Fed does decide to shrink its balance sheet, how long will that process take? How will the Fed control interest rates in the years ahead? And most importantly, how will these decisions impact financial markets? To answer these questions, this week we are sending you a Special Report titled "Cleaning Up After The 100-Year Flood" that was first published on June 10, 2014. This report explains how monetary policy is conducted at an operational level, and also how the dramatic expansion of the Fed's balance sheet forced the Fed to modify its approach. But first, we have some additional thoughts on how the Fed's balance sheet is likely to evolve during the next few years. The Fed's Stated Plan The most up-to-date guidance we have received about the Fed's balance sheet plans comes from Janet Yellen's recent Congressional testimony: The FOMC has annunciated that its longer run goal is to shrink our balance sheet to levels consistent with the efficient and effective implementation of monetary policy. And while our system evolves and I can't put a number on that, I would anticipate a balance sheet that's substantially smaller than at the current time. In addition, we would like our balance sheet to again be primarily Treasury securities, whereas as you pointed out, we have substantial holdings of mortgage-backed securities. From this, and similar statements from other Fed officials, we conclude that the Fed will allow its balance sheet to shrink once the fed funds rate is somewhere in the range of 1% to 1.5%.1 Surveys also show that the median primary dealer expects the Fed will change its balance sheet policy when the target fed funds rate is 1.38%. As such, and under reasonable assumptions for the pace of rate hikes, we think it is very likely that the Fed will start to let its balance sheet shrink sometime in 2018. MBS First, Treasuries Maybe Later Yellen's statement to Congress also makes clear that the Fed would be more comfortable with a balance sheet that consists entirely of Treasury securities. For this reason, the central bank will start by simply ceasing the reinvestment of its Agency bond and MBS portfolios. At least initially, the Fed will continue to reinvest the proceeds from its maturing Treasury portfolio. Yellen also left open the possibility that reinvestment could be "tapered" rather than just halted altogether. While this is possible, and in fact 70% of primary dealers think that reinvestments will be phased out over time while only 14% think they will be ceased all at once, it seems to us like a needless complication. We expect that reinvestments of Agency bonds and MBS will end all at once sometime in 2018. As for the Fed's holdings of Treasury securities, it is much less clear whether the Fed will allow these balances to run down. The accompanying Special Report describes in detail the differences between the Fed's pre-crisis mode of operation, when bank reserves were scarce, and the Fed's current mode of operation with large bank reserve balances. As of now, the Fed has stated that it intends to eventually drain bank reserves from the system and return to its pre-crisis mode of operation, but there are several possible advantages to running a system with an outsized Fed balance sheet and large bank reserve balances. Chart 1Reserves Can Be Drained Fairly Quickly Reserves Can Be Drained Fairly Quickly Reserves Can Be Drained Fairly Quickly None other than Ben Bernanke pointed out a few of those reasons in a blog post last fall.2 In our view, the most compelling is that regulatory changes have increased private sector demand for safe, short-maturity, liquid assets in recent years. If the Treasury department is unwilling to supply T-bills in sufficient numbers, then the Fed can supply safe, short-maturity, liquid assets to the market by purchasing long-maturity Treasury securities and replacing them with bank reserves. Of course, we take the Fed at its word when it says that it would like to eventually drain excess bank reserves from the system. But even in that case, the steady growth of currency in circulation means that bank reserves will decline over time even if the Fed keeps the asset side of its balance sheet flat. For example, Chart 1 shows what would happen to bank reserves if the amount of currency in circulation grows at a conservative 5% per year pace, and if the Fed decides to allow its Agency bond and MBS portfolios to run off at the beginning of next year while keeping its Treasury portfolio flat. We assume that MBS runs off the Fed's balance sheet at a pace of $15 billion per month, slightly below the recent pace of MBS reinvestment. During the past three years, the Fed has reinvested between $20bn and $40bn MBS each month with an average monthly reinvestment of $32bn. In this scenario, outstanding bank reserves would decline to zero by the end of 2025. At that point the Fed would have to start adding to its Treasury holdings just to keep pace with the amount of currency in circulation. Bottom Line: While it is very likely that the Fed will allow its Agency bond and MBS portfolios to run off starting in 2018, it is much more uncertain whether it will ever cease the reinvestment of its Treasury holdings. If the Fed does allow its Treasury holdings to run down as well, it will have to start buying Treasuries again before 2025. Investment Implications Treasuries As our U.S. Bond Strategy service has written several times,3 considered in isolation it is unlikely that any decision by the Fed to allow its Treasury holdings to run off will have much of on an impact on the Treasury curve. To see why, we need to consider the process by which the Fed currently rolls over maturing Treasury securities at auction. At the moment, balances of matured Treasury securities are added to upcoming note/bond auctions as non-competitive bids. In other words, as Treasury securities mature the Fed buys an equal amount at upcoming Treasury auctions. If the Fed were to cease this reinvestment, that amount would need to be added to the competitive portion of the auctions and would greatly increase the gross issuance of Treasury debt to the public. For a sense of scale, we calculate that Treasury issuance to the public would need to increase by $426bn in 2018 and $378bn in 2019 if the Fed were to cease the reinvestment of its portfolio at the end of this year (Chart 2). However, the fact that this process is intermediated by the Treasury department means we also have to consider potential changes to fiscal policy and the U.S. government's financing mix. For instance, since running down the Fed's Treasury portfolio would also reduce the amount of bank reserves in the system, it is very likely that the Treasury department would seek to increase issuance of T-bills to compensate for the banking sector's loss of safe, short-maturity liquid assets. At present, bill supply as a percent of total Treasury debt is near a multi-decade low (Chart 3) and any increase in T-bill issuance would limit the impact of Fed balance sheet run-off on long-dated Treasury yields. Chart 2Fed Runoff Will Increase Issuance To Public... Fed Runoff Will Increase Issuance To Public... Fed Runoff Will Increase Issuance To Public... Chart 3... But Mostly Through T-Bills ... But Mostly Through T-Bills ... But Mostly Through T-Bills Bottom Line: When forecasting Treasury issuance and any potential impact on yields we must consider both the Fed's balance sheet and fiscal policy together. In our view, whatever the government's financing requirement in the years ahead, a considerable portion will be met through increased T-bill issuance, limiting the impact on long-dated Treasury yields. Mortgage-Backed Securities As our U.S. Bond Strategy service has recently written,4 the unwinding of the Fed's MBS portfolio poses a considerable threat to MBS spreads for two reasons. First, the transfer of MBS from the Fed to the private sector will put upward pressure on implied volatility. While private investors often hedge their MBS positions by purchasing volatility, the Fed has no incentive to do so. It follows that by removing a large stock of MBS from private hands the Fed has also removed a large source of demand for volatility. When this supply is re-introduced into the market, demand for volatility will increase pressuring MBS spreads wider (Chart 4). The second reason relates more directly to the supply and demand balance for MBS. In years when net MBS issuance (adjusted for Fed purchases) has been negative, excess MBS returns have tended to be positive (Chart 5). Further, while negative net MBS issuance (adjusted for Fed purchases) has been the norm since Fed asset purchases began in 2009 (Chart 6), this state of affairs will change once the Fed starts to unwind its MBS portfolio. Chart 4MBS Spreads Are##br## Linked To Vol MBS Spreads Are Linked To Vol MBS Spreads Are Linked To Vol Chart 5Annual MBS Excess Returns##br## Vs. Net Supply Since 1989 The Way Forward For The Fed's Balance Sheet The Way Forward For The Fed's Balance Sheet Chart 6Adjusted Net Issuance Will ##br##Turn Positive In 2018 Adjusted Net Issuance Will Turn Positive In 2018 Adjusted Net Issuance Will Turn Positive In 2018 Bottom Line: The unwinding of the Fed's MBS portfolio will pressure MBS spreads wider through increased supply and increased demand for volatility. Note: Please see the U.S. Bond Strategy Special Report, titled "Cleaning Up After The 100-Year Flood", dated June 10, 2014 available at usbs.bcaresearch.com Ryan Swift, Vice President U.S. Bond Strategy rswift@bcaresearch.com 1 In a Q&A from June 2015 New York Fed President William Dudley floated 1% to 1.5% as a potentially reasonable range for the fed funds rate before the Fed considers changing its balance sheet policy. https://mninews.marketnews.com/content/feds-dudley-qa-markets-should-not-be-surprised-liftoff 2 https://www.brookings.edu/blog/ben-bernanke/2016/09/02/should-the-fed-keep-its-balance-sheet-large/ 3 Please see U.S. Bond Strategy Weekly Report, "Is It Time To Cut Duration?", dated January 17, 2017. And also U.S. Bond Strategy Weekly Report, "Currencies: The Tail Wagging The Dog", dated August 18, 2015. Both available at usbs.bcaresearch.com 4 Please see U.S. Bond Strategy Weekly Report, "The Road To Higher Vol Is Paved With Uncertainty", dated February 14, 2017, available at usbs.bcaresearch.com Cleaning Up After The 100-Year Flood In this Special Report we consider how the dramatic expansion of the Fed's balance sheet has influenced the conduct of monetary policy from an operational perspective. Massive reserve balances have made the federal funds market largely irrelevant, but the Fed's new overnight reverse repo facility will allow it to tighten policy when the time comes. The Fed will likely provide new guidelines for its exit strategy before the end of 2014. We anticipate how these guidelines will be modified to reflect the challenges of implementing monetary policy with a large balance sheet. Large reserve balances do not pose an inflation threat, but they do have implications for the state of banking sector regulation and the policy tools at the Fed's disposal. Chart 1What Are Implications Of Fed's Epic Intervention? The Way Forward For The Fed's Balance Sheet The Way Forward For The Fed's Balance Sheet Feature The Federal Reserve resorted to a number of very aggressive and extraordinary monetary policy measures to deal with the failure of Lehman Brothers, the subsequent financial crisis and Great Recession. The result has been a flood of liquidity that has supported asset prices and spurred the recovery, yet has left the central bank balance sheet exponentially larger than at any time in its 100-year history (Chart 1). In formulating its exit strategy the Fed will finally be forced to grapple publicly with the aftereffects of its dramatic intervention in financial markets, which has complicated how monetary policy is implemented at an operational level. This Special Report is divided into three sections. In the first section, Before The Storm, we provide some background on the process of money creation and explain how the Fed implemented monetary policy prior to the Great Recession. In the second section, The Flood Waters Rise, we consider how monetary policy is implemented today in light of the dramatic expansion of the Fed's balance sheet. In the third section, Building On Higher Ground, we examine the way forward for the Fed, describe how the exit is likely to be managed and discuss the potential problems with this approach. 1. Before The Storm Prior to the financial crisis, the Fed expressed the stance of monetary policy via a target for the federal funds rate. The federal funds rate is the rate at which banks borrow and lend reserves to each other in the overnight market. The Fed conducted its day-to-day operations with the goal of supplying all the reserves demanded by the banking sector, while steering the fed funds rate toward its target. To understand how this was accomplished, we first require some background on the money creation process. Money Creation: More Than Just A Printing Press Contrary to what many believe, the process of money creation does not begin at the Federal Reserve. Rather, it begins in the banking system at the point of loan origination and ends at the Fed (Figure 1). The process is set in motion when a bank makes a new loan. This loan creates a new asset on the aggregate balance sheet for the banking system. The newly created money typically ends up as a deposit, either at the same bank or elsewhere in the banking system.1 The increase to the asset side of the banking sector's balance sheet is offset by an equal increase on the liability side. Only then does the Fed enter the picture. In a fractional reserve system, banks must hold reserves equal to a proportion of their deposits. Therefore, in the pre-QE world illustrated in Figure 1, any increase in deposits also creates demand for more reserves. Crucially, only the Fed is able to supply the banking sector with the needed reserves. The Fed will increase the supply of reserves either by purchasing Treasury securities or lending money in the repo market. This increase on the asset side of the Fed's balance sheet is balanced by an increase in reserves on the liability side. The creation of new bank reserves is the final step in the money creation process. Figure 1The Money Creation Process The Way Forward For The Fed's Balance Sheet The Way Forward For The Fed's Balance Sheet Chart 2QE Has Not Encouraged Lending The Way Forward For The Fed's Balance Sheet The Way Forward For The Fed's Balance Sheet This is not to suggest the Fed is powerless to control the rate of money creation. On the contrary, Fed policy is the most influential determinant of the rate of money growth. One important distinction, however, is that the Fed exerts control over the pace of money creation because it controls the overnight interest rate. The interest rate, in turn, is the most important driver of bank lending. Changes in the level of bank reserves not associated with changes in interest rates, QE for example, have no effect on credit growth (Chart 2). The Pre-Crisis Fed Funds Market How then, prior to the Great Recession, was the Fed able to maintain the fed funds rate at its target, while still satisfying the banking system's demand for reserves? It accomplished this task by maintaining what it calls a "structural deficiency" in the supply of bank reserves. In practice this means the Fed was careful to supply no more than the quantity of reserves demanded, so that each day it would typically add to the reserve supply to accommodate the newly created demand. As a practical matter, the Fed increases the supply of reserves by either buying securities or lending money in the repo market. Both of these transactions enter the Fed's balance sheet as an asset, which must be offset by a liability, in this case an increase in bank reserves. The Fed can also reduce the supply of reserves by either selling securities or borrowing in the repo market (using the securities it owns as collateral, deemed a reverse repo from the point of view of the borrower). Although due to the "structural deficiency" in the reserve market, the Fed would typically transact to increase reserve balances. Chart 3The Pre-Crisis 'Channel System' The Way Forward For The Fed's Balance Sheet The Way Forward For The Fed's Balance Sheet If, for example, the Fed wanted to increase the fed funds rate. It would be slow to accommodate the increase in demanded reserves throughout the day. Banks in need of reserves to meet intra-day payment processing needs would bid up the fed funds rate towards the new target. Effective communication of the target fed funds rate also aided this process. Since the target for the fed funds rate was known in advance, and the banking sector believed the Fed would supply all necessary reserves at that target rate, most federal funds transactions tended to occur at rates very close to the target. By the end of the day, however, the Fed must always supply the exact amount of reserves demanded by the banking sector if it wants to maintain control of the fed funds rate. If the Fed were to supply more reserves than the banking system required, banks would try to lend the unwanted excess reserves in the fed funds market, driving the federal funds rate lower to the interest rate paid on excess reserves (IOER), which was zero prior to 2008. Or, consider the opposite case where the Fed supplies too few reserves. In this instance banks would clamor to borrow reserves to meet their regulatory requirement. This incremental demand would drive the federal funds rate higher to the Fed's discount window lending rate, which is always available for banks to access in times of severe stress. The IOER and discount window rate thus created a channel for the fed funds rate (Chart 3), within which the Fed could nudge the rate toward target by being either too quick, or too slow to accommodate increases in demanded reserves throughout the day. Bottom Line: A "structural deficiency" in reserve balances prior to 2008 allowed the Fed to conduct monetary policy by setting a target for the fed funds rate. Also, it is the level of interest rates, and not the level of reserves, that determines the rate of money creation in the economy. 2. The Flood Waters Rise The Federal Reserve began large scale asset purchases (quantitative easing) in late 2008, dramatically increasing the asset side of its balance sheet and consequently the supply of bank reserves (Figure 2). Suddenly, the banking system found itself with far more reserves than it demanded. Predictably, trading in the fed funds market dried up and the fed funds rate was driven toward its lower bound, the IOER.2 The Fed's target for the federal funds rate quickly became irrelevant. Figure 2Illustrative Post-Crisis Balance Sheets For The Fed And The Aggregatve Banking System*: ##br##An Explosion In Excess Reserves The Way Forward For The Fed's Balance Sheet The Way Forward For The Fed's Balance Sheet In the presence of excess bank reserves, the Fed needs a mechanism to control the lower bound of overnight lending rates. In theory, the IOER could serve as a floor beneath the fed funds rate because banks should not be willing to lend reserves at a rate lower than what can be earned at the Fed. Yet the fed funds rate has consistently traded below the IOER since 2008 (Chart 4). The reason for this violation is that the IOER is only available to depository institutions with reserve accounts at the Fed. Other suppliers of short maturity funds, mostly the GSEs, are still willing to transact at lower rates (Chart 5). Chart 4In Need ##br##Of A Floor The Way Forward For The Fed's Balance Sheet The Way Forward For The Fed's Balance Sheet Chart 5Fed Funds Market Smaller, ##br## And Dominated By GSEs The Way Forward For The Fed's Balance Sheet The Way Forward For The Fed's Balance Sheet Chart 6Reverse Repo Facility Is New Floor On ##br##Rates Money Markets Under The Microscope The Way Forward For The Fed's Balance Sheet The Way Forward For The Fed's Balance Sheet A new facility is required, that is capable of absorbing all of the supply of overnight funds including that emanating from outside the traditional banking system. The Fed answered this requirement by creating a fixed rate overnight reverse repo (RRP) facility. Once fully implemented, the Fed will stand ready to borrow overnight in unlimited amounts, at a rate that it chooses (i.e. is set independently of market forces). By making this facility available to a larger set of counterparties than the IOER, including money market funds and the GSEs, the Fed now has a "hard floor" on rates that it will be able to use to raise interest rates when the time comes. Even though it is still in a testing phase, the RRP already appears to be acting as a floor for overnight rates (Chart 6). Bottom Line: The stockpile of excess reserves created by the Fed's large scale asset purchase program has made the federal funds market largely irrelevant. The Fed is now only able to implement monetary policy by placing a floor under short-term interest rates, the RRP. 3. Building On Higher Ground From an operational perspective, there are two possible ways forward for the Fed as it prepares to lift rates. One option would be to return to the pre-crisis method of operation described in the first section. To do this, the Fed would first have to drain all excess reserves from the banking system by either selling securities, or deploying some of the tools on the liability side of its balance sheet, such as term deposits.3 This would re-launch the federal funds market and the Fed could return to setting policy in its traditional manner, by targeting the fed funds rate. Unfortunately, there are simply too many excess reserves in the system to make this a viable strategy, at least for the next several years. Moreover, the pace of asset sales required to drain excess reserves in a timely manner would lead to large spikes in the Treasury term premium. Instead, the Fed will almost certainly choose to maintain large reserve balances and operate monetary policy by lifting the floor RRP rate. Specifically, the Fed will set the RRP rate equal to (or slightly below) the IOER. It will then hike rates by increasing both in tandem. The Fed may still choose to set a target for the fed funds rate at a level somewhat above the RRP to maintain consistency in its communications, but this rate will be meaningless. We outline the likely sequence of the Fed's exit strategy in the following Box. Box The Exit Strategy Revisited The Fed first articulated the likely sequence of the exit strategy in the minutes to the June 2011 FOMC meeting.4 That sequence was as follows: Cease reinvestment of principal on securities holdings. Modify forward guidance on the path of the federal funds rate, and initiate reserve draining operations (e.g. term deposits). Begin raising the target federal funds rate. Begin sales of securities holdings, with a goal of returning the balance sheet to a more traditional size within two to three years. The above sequence suggests the Fed was planning to first drain excess reserves and then conduct monetary policy operations in the fed funds market, as it did prior to QE. This strategy has now been abandoned, and we expect to receive a modified exit sequence before the end of the year. The revised sequence will be consistent with the implementation of policy using a floor system, with large excess reserves, and could look something like: Modify forward guidance on the path of interest rates (including IOER, RRP and fed funds). Begin raising interest rates. First by raising the RRP rate to slightly below the IOER, and then by raising both rates in tandem. A few months after rate hikes begin; cease reinvestment of principal on Agency and Agency MBS holdings. Much later; cease reinvestment of principal on Treasury securities. The Fed will probably cease reinvestment of its MBS holdings prior to its Treasury holdings, and will then let its MBS holdings run-off passively to zero. The Fed will also probably let some of its Treasury holdings run-off passively, but could decide to maintain a permanently larger balance sheet, depending on the success of the RRP and floor system. In the next few years, as its balance sheet begins to shrink through passive run-off, the Fed may decide to drain the remaining excess reserves and return to its traditional operating method as outlined in Section1 above. Either way, U.S. monetary policy will operate under a "floor system", using the RRP rate, for at least the next few years. This new method of operation comes with several potential drawbacks, which we address below. Excess Reserves Are "Dry Powder" For The Banking System Chart 7Drivers Of Bank Lending The Way Forward For The Fed's Balance Sheet The Way Forward For The Fed's Balance Sheet Many have speculated that banks have been choosing to sit on large excess reserve balances. The thinking is that eventually the economy will reach a turning point and banks will decide to convert their excess reserves to loans en masse, leading to a surge in bank lending, and eventually, inflation. This implies that the presence of large excess reserve balances would force the Fed to hike rates more quickly than in their absence. This concern stems from a misunderstanding of the money creation process described above. The Fed could fall "behind the curve" and normalize policy too slowly, which could ultimately lead to higher inflation. However, this would simply be a consequence of keeping interest rates too low for too long. The presence of excess reserves does not in itself create a desire to lend and thus poses no additional inflation risk. For one thing, the banking system in aggregate is powerless to reduce the amount of excess reserves without the Fed also taking action to reduce the supply. As shown in Figures 1 and 2, the supply of excess reserves is determined solely on the Fed's balance sheet. There is no danger of excess reserves "leaking out" into the economy. More importantly, however, is that the process of money creation begins with the origination of a loan and ends when the Fed increases the supply of reserves. The catalyst for the process, the amount of bank lending, is determined by (Chart 7): loan demand banks' perceived profitability from additional lending banks' concerns about taking too much risk on the balance sheet, putting their viability at risk regulatory requirements concerning capital and liquidity The Fed exerts control over these four factors through its interest rate policy, but not through changes in the balance of excess reserves. Prior to 2008, the lack of excess reserves did not act to constrain bank lending, rather the Fed chose to encourage or discourage lending by decreasing or increasing the interest rate. Similarly, the large excess reserve balances since 2008 have not provided an incentive to lend. Excess Reserves and Bank Regulation One implication of the Fed having sole control over the supply of bank reserves is that, through QE, it has effectively forced reserves onto bank balance sheets. These reserves obviously factor into banks' calculations concerning required regulatory ratios for liquidity and capital. Liquidity Coverage Ratio Chart 8Fed Purchases Pushed ##br##Term Premium Lower The Way Forward For The Fed's Balance Sheet The Way Forward For The Fed's Balance Sheet Under the proposed liquidity coverage ratio (LCR), banks must maintain a balance of high-quality liquid assets (such as bank reserves and Treasury securities) equal to their expected net cash requirement during a 30-day period. In theory, should the Fed ever decide to reduce the supply of excess reserves, banks could have trouble meeting the LCR requirement. In removing reserves, however, the Fed would also be selling securities. Banks falling short of their LCR requirement would be natural buyers for the securities offloaded by the Fed. Thus, the Fed's operating decisions will probably not exert any influence over banks' ability to meet their liquidity requirements. The LCR, however, does have implications for the equilibrium level of the Treasury term premium. Much as the Fed's Treasury purchases pressured the term premium lower (Chart 8), any future Treasury sales could be expected to unwind this effect. Even so, bank demand for those same Treasury securities would mitigate some of the upside for the term premium. Supplementary Leverage Ratio Large U.S. banks face a supplementary leverage ratio (SLR) which requires them to hold capital equal to at least 5% of total assets, not risk-weighted. In other words, large excess reserves force banks to hold more capital, which could have an adverse economic impact. Banks falling short of the SLR can either raise capital, or reduce assets. If they are either unable or unwilling to raise capital, then the large balance of excess reserves thrust upon them by the Fed could in theory crowd out bank lending. In other words, if the banking sector refuses to increase capital, then the onus falls on the asset side of the balance sheet to adjust to SLR standards. Since the banking sector in aggregate is unable to reduce reserve balances, any desired contraction in total assets could conceivably translate into an incentive to reduce the pace of bank lending. As currently proposed the SLR does not appear to be too big a hurdle for the largest U.S. banks. It is very likely they will be able to meet the requirement through retained earnings and new equity issuance. Nevertheless, it still provides a potential drag on bank lending that would not exist under the traditional model of monetary policy operations. Collateral Shortage Chart 9RRP Alleviates Collateral Shortage The Way Forward For The Fed's Balance Sheet The Way Forward For The Fed's Balance Sheet One side effect of the Fed's large scale asset purchases is that they have removed a lot of high-quality collateral from the financial system. Chart 9 shows that during periods when the Fed is adding to its balance sheet, the amount of collateral in the tri-party repo market declines. As the supply of collateral dwindles, repo rates are also pressured lower. The problem is that once repo rates approach the zero lower bound, counterparties have an increasing incentive to fail on delivery of repo contracts. Given the widespread use of repo financing, persistent repo fails have the potential to undermine liquidity in financial markets. Thankfully, the Fed's new tool for controlling the overnight interest rate, the RRP facility, solves this problem. In a reverse repo transaction, a counterparty purchases securities from the Fed with the understanding that it will sell them back the next day, earning the RRP rate in the process. This means the private sector once again gains access to collateral that had been cordoned off on the Fed's balance sheet. This should have the effect of keeping the repo rate above the floor set by the RRP, and well above zero. Repo fails have already levelled off and should begin to decline once the RRP facility is fully implemented. Financial Stability Concerns We have seen that monetary policy operates under a floor system when there are large reserve balances. One complication is that the U.S. is operating with two different floors, the IOER and the RRP. Due to its availability to a wider selection of counterparties, the RRP is the true floor on rates. From a monetary policy perspective, the easiest way forward is to set both rates at the same level and hike them in tandem. However, in a recent speech5 New York Fed President Dudley pointed out that from a financial stability perspective an RRP rate equal to the IOER could result in money flowing out of institutions eligible to receive IOER and into the less regulated shadow banking sector. It is therefore probable the Fed will choose to maintain the RRP at a level slightly below the IOER as rate hikes commence. We maintain focus on the RRP as the true floor on rates. President Dudley also made the case that the Fed's RRP facility could have positive implications for financial stability. He observed that with a Fed-backed short-term safe asset now more widely available, it could crowd out the creation of money-like liquid assets by the private sector. Those privately created liquid assets, such as commercial paper, are more prone to fire sales during times of stress. In a recent paper,6 John Cochrane agreed forcefully with this sentiment. Due to its potential for eliminating privately created money-like liquid assets, Cochrane referred to a monetary policy regime operating with large reserve balances as "a very desirable configuration of monetary affairs." The downside of the Fed providing a short-term safe asset is that it could encourage runs into the RRP during times of crisis. President Dudley rightly concludes that this is more of a technical hurdle that could be managed using caps on usage of the RRP facility. Bottom Line: The Fed will be able to operate monetary policy with large reserve balances, using the RRP as a floor on interest rates for several years while it decides by how much to run down its balance sheet and whether it should revert to its traditional fed funds rate target. Investors should remember that large reserve balances, by themselves, do not pose an inflation risk. Whether or not inflation becomes a problem will depend on the Fed's foresight to raise interest rates in a timely manner. Ryan Swift, Vice President U.S. Bond Strategy rswift@bcaresearch.com 1 Alternatively it could be held in cash. This would be reflected on the banking sector's balance sheet as an increase in loans and a decrease in reserves, and on the Fed's balance sheet as an increase in currency in circulation and a decrease in reserves. 2 The Fed began paying interest on excess reserve balances on October 6, 2008. 3 The Fed's current Term Deposit Facility (TDF) temporarily drains reserves from the banking system by receiving funds from the banking sector for a period of 7 days, paying 26 basis points of interest. The early stages of the Fed exit strategy will rely more heavily on the overnight RRP facility rather than the TDF. But term deposits could be deployed once the Fed's balance sheet has returned closer to its traditional size, and the Fed decides it wants to drain the remaining excess reserves and return to its pre-crisis method of operation. 4 http://www.federalreserve.gov/monetarypolicy/files/fomcminutes20110622… 5 "The Economic Outlook and Implications for Monetary Policy" available at http://www.newyorkfed.org/newsevents/speeches/2014/dud140520.html 6 Cochrane, John H. Monetary Policy with Interest on Reserves. Available at http://media.hoover.org/sites/default/files/documents/2014CochraneMonet…

Highlights Rate Volatility: Forecast disagreement about GDP growth and T-bill rates will increase over the course of the year. This, alongside elevated policy uncertainty, will translate into higher interest rate volatility. Treasury Yields: Higher rate volatility should cause the term premium in the Treasury curve to increase at the margin. However, this impact could be offset if rate volatility and equity volatility rise in concert. An increase in equity vol would encourage flight-to-safety flows into bonds. MBS: Higher interest rate volatility and the unwinding of the Fed's mortgage portfolio will lead to wider MBS spreads during the next two years. Feature Low interest rate volatility has been a constant feature of the investing landscape during the past few years. In fact, you need to go back to the 1970s to find another period when interest rate volatility was consistently at or below its current level (Chart 1). Not surprisingly, the implied volatility priced into Treasury options is also as low as it has been during the past 30 years, with the exception of the period just prior to the financial crisis in 2007 (Chart 2). Chart 1Yield Volatility: Lowest Since The 70s Yield Volatility: Lowest Since The 70s Yield Volatility: Lowest Since The 70s Chart 2Implied And Realized Yield Volatility Move Together Implied And Realized Yield Volatility Move Together Implied And Realized Yield Volatility Move Together This begs the question of whether the current low-vol environment can be sustained, or whether overly complacent investors are in for a shock. At the very least, we believe that rate volatility has already passed its cyclical trough and will start to move up this year. Investors should prepare themselves for higher volatility. In this week's report we examine the key macro drivers of interest rate volatility and discuss the implications of rising vol for both Treasury yields, and crucially, mortgage-backed securities. Macro Uncertainty & Rate Volatility Chart 3Macro Drivers Of Rate Volatility Macro Drivers Of Rate Volatility Macro Drivers Of Rate Volatility In a Special Report published in 2014,1 we posited that the long-term trends in volatility across all asset classes are largely driven by common macroeconomic factors. Specifically, investor uncertainty regarding the outlook for economic growth and monetary policy. A 2004 paper by Alexander David and Pietro Veronesi2 provides some theoretical justification for this view, as the authors observed that investors tend to overreact to new information when macro uncertainty is high, and underreact when uncertainty is low. To test the linkage between interest rate volatility and macro uncertainty we consider three measures of uncertainty. The first two measures, shown alongside the MOVE index of implied Treasury volatility in Chart 3, are measures of GDP growth and T-bill rate forecast dispersion. We measure dispersion - the disagreement among forecasters - by looking at individual forecasts of GDP growth and T-bill rates and calculating the difference between the 75th and 25th percentiles. The series shown in Chart 3 are equal-weighted averages of the forecast dispersion calculated for five different time horizons, ranging from the current quarter to four quarters ahead. As can be seen in the top two panels of Chart 3, implied interest rate volatility is higher when the disagreement among forecasters is greater, consistent with our thesis. The third measure of uncertainty we consider is the Global Economic Policy Uncertainty Index created by Baker, Bloom and Davis.3 This index tracks uncertainty about the macro environment by counting the number of mentions of certain key words in major global newspapers. Elevated readings from this index have also coincided with high rate volatility in the past (Chart 3, bottom panel). GDP Growth Forecast Dispersion Chart 4Forecast Dispersion & Corporate Lending Forecast Dispersion & Corporate Lending Forecast Dispersion & Corporate Lending Disagreement among GDP growth forecasts reached an all-time low in the fourth quarter of 2016, but has since recovered to slightly more typical levels. Historically, we have found that C&I lending standards and corporate sector balance sheet health correlate most closely with GDP growth forecast dispersion (Chart 4) and both measures suggest that forecast dispersion is biased upward. T-Bill Rate Forecast Dispersion T-bill rate forecast dispersion was abnormally low between 2011 and 2014 for two reasons. The first reason is quite simply the zero-lower-bound on interest rates. A short rate bounded at zero necessarily trimmed the distribution of possible T-bill rate forecasts, since forecasters logically assumed that further interest rate cuts were not possible. This impact will gradually dissipate the further the fed funds rate moves off zero. Chart 5Fed Says March Meeting Is Live Fed Says March Meeting Is Live Fed Says March Meeting Is Live The second reason for extremely low T-bill rate forecast dispersion was the Fed's forward guidance. During this timeframe the Fed was actively trying to convince the public that interest rates would remain low. The most obvious example being the "Evans Rule", where the Fed promised not to lift interest rates at least until the unemployment rate had fallen below a specific threshold. This activist forward guidance limited the range of conceivable T-bill rate forecasts and crushed interest rate volatility. Nowadays, the Fed is engaged in a different sort of forward guidance, trying to convince markets that every FOMC meeting is live and that rate hikes could occur at any moment. Essentially, the Fed is trying to inject volatility into the rates market. Just a few weeks ago, when asked about the low probability markets are assigning to a March rate hike (Chart 5), San Francisco Fed President John Williams replied flatly: "I don't agree. All our meetings are live." Global Economic Policy Uncertainty We have written a lot about the policy uncertainty index in recent reports,4 focusing specifically on how it has diverged from its historical relationships with many asset prices. At the very least, we expect that sustained elevated policy uncertainty will place upward pressure on asset price volatility at the margin. Bottom Line: Forecast disagreement about GDP growth and T-bill rates will increase over the course of the year. This, alongside elevated policy uncertainty, will translate into higher interest rate volatility. Rate Volatility & Treasury Yields Long-dated nominal Treasury yields can be decomposed in a few different ways. In recent reports we have focused on the decomposition of the nominal 10-year Treasury yield into its real and inflation components. By identifying different macro drivers for each component we concluded that nominal Treasury yields will increase this year, driven by a rising inflation component and relatively stable real yields.5 Alternatively, we can think of the nominal 10-year Treasury yield as consisting of an expectations component equal to the market's expected path of short rates over the next ten years, and a term premium that reflects all of the other market imbalances and uncertainties associated with taking duration risk. This second approach is complicated by the fact that it requires a model of ex-ante interest rate expectations and every commonly used model is fraught with its own unique difficulties.6 Setting that aside, if we use the Kim & Wright (2005)7 estimate of the 10-year term premium we observe an expectations component that generally tracks the fed funds rate and a term premium component that is correlated with implied Treasury volatility (Chart 6), although the latter correlation is less than perfect. This decomposition also suggests that nominal Treasury yields should rise. The Fed is much more likely to hike rates than cut them and we have concluded that rate volatility is likely to trend higher from current depressed levels. However, the relationship between rate volatility and the term premium is complicated. The main reason for the complicated relationship between interest rate volatility and the term premium is the fact that elevated interest rate volatility also tends to be correlated with high equity volatility (Chart 7). So while higher rate volatility puts upward pressure on the term premium, the associated increase in equity volatility tends to raise investor risk aversion and increase the perceived value of bonds as a hedge against equity positions. This mitigates some (or often all) of the impact of rising rate volatility on the term premium. Chart 6Which Way For The ##br##Term Premium? Which Way For The Term Premium? Which Way For The Term Premium? Chart 7MOVE & VIX Have Opposing##br## Impacts On Bond Yields MOVE & VIX Have Opposing Impacts On Bond Yields MOVE & VIX Have Opposing Impacts On Bond Yields Bottom Line: Higher rate volatility should cause the term premium in the Treasury curve to increase at the margin. However, this impact could be offset if rate volatility and equity volatility rise in concert. An increase in equity vol would encourage flight-to-safety flows into bonds. Rate Volatility & MBS The relationship between rate volatility and MBS is much more straightforward than for Treasury yields. We observe a tight correlation between nominal MBS spreads and the MOVE implied volatility index (Chart 8). Chart 8 suggests that, even in the near-term, MBS spreads are too low for current levels of rate vol. The relationship between MBS spreads and rate volatility is easily explained. The defining characteristic of a negatively convex asset, such as MBS, is that its duration is positively correlated with the level of interest rates (Chart 9). This correlation leads to increased losses when yields rise and lower gains when yields fall. It's not surprising that negatively convex assets perform best in low volatility environments. Chart 8MBS Spreads Are Linked To Vol MBS Spreads Are Linked To Vol MBS Spreads Are Linked To Vol Chart 9MBS Duration Moves With Yields MBS Duration Moves With Yields MBS Duration Moves With Yields We maintain an underweight allocation to MBS given that spreads are already low and that the volatility environment is poised to become less favorable. Further, if the Fed continues along its planned normalization path it is likely to cease the reinvestment of its MBS portfolio at some point in 2018. There are two reasons why this poses a risk for MBS. The first reason is that the unwinding of the Fed's MBS portfolio is likely to place upward pressure on implied volatility. While private investors often hedge their MBS positions by purchasing volatility, the Fed has no incentive to do so. It follows that by removing a large stock of MBS from private hands the Fed has also removed a large source of demand for volatility. When this supply is re-introduced into the market, demand for volatility is likely to increase. The second reason relates more directly to the supply and demand balance for MBS. In years when net MBS issuance (adjusted for Fed purchases) has been negative, excess MBS returns have tended to be positive (Chart 10). Further, while negative net MBS issuance (adjusted for Fed purchases) has been the norm since Fed asset purchases began in 2009 (Chart 11), this state of affairs will change once the Fed starts to unwind its MBS portfolio. Chart 10Annual MBS Excess Returns ##br## Vs. Net Supply Since 1989 The Road To Higher Vol Is Paved With Uncertainty The Road To Higher Vol Is Paved With Uncertainty Chart 11Net Issuance Will Turn##br## Positive In 2018 Net Issuance Will Turn Positive In 2018 Net Issuance Will Turn Positive In 2018 During the past three years the Fed has been buying between $20bn and $40bn MBS per month, just to keep its balance sheet stable. Net new MBS issuance will not be strong enough to overcome this hurdle in 2017, but net MBS issuance (adjusted for Fed purchases) will swing quickly into positive territory in 2018 if the Fed decides to let its MBS portfolio run down. Bottom Line: Higher interest rate volatility and the unwinding of the Fed's mortgage portfolio will lead to wider MBS spreads during the next two years. Ryan Swift, Vice President U.S. Bond Strategy rswift@bcaresearch.com 1 Please see U.S. Bond Strategy / Global Fixed Income Strategy Special Report, "Volatility, Uncertainty And Government Bond Yields", dated May 13, 2014, available at usbs.bcaresearch.com 2 "Inflation and earnings uncertainty and volatility forecasts", Alexander David and Pietro Veronesi, Manuscript, Graduate School of Business, University of Chicago (2004). 3 Please see www.policyuncertainty.com for further details. 4 Please see Theme # 4 in U.S. Bond Strategy Special Report, "Seven Fixed Income Themes For 2017", dated December 20, 2016, available at usbs.bcaresearch.com 5 Please see U.S. Bond Strategy Weekly Report, "Is It Time To Cut Duration?", dated January 17, 2017, available at usbs.bcaresearch.com 6 Please see U.S. Bond Strategy Weekly Report, "Bond Volatility - The Unwelcome Guest That Will Not Leave", dated June 16, 2015, available at usbs.bcaresearch.com 7 Don H. Kim and Jonathan H. Wright, "An Arbitrage-Free Term Structure Model and the Recent Behavior of Long-Term Yields and Distant-Horizon Forward Rates", FEDS 2005-33. https://www.federalreserve.gov/econresdata/feds/2005/index.htm Fixed Income Sector Performance Recommended Portfolio Specification