The Monster Under the Bed

Barely a moment goes by these days without a business news story claiming volatility is disturbingly low, or has “disappeared”, or even suggesting the paradoxical-sounding notion that “the anti-volatility trade” is “one of the market’s hottest”.1 Traders are apparently standing around, brows furrowed, tension building, watching prices do… nothing. Investors tend to think of volatility not as something that disappears, but goes into hiding like the imaginary monster under a child’s bed, waiting until the hapless youngster drifts off to sleep to suddenly reach out and grasp his leg. The longer the wait, the more startling the scare and the bigger the overreaction. One has only to think of the long calm periods before the chaos of 1987, 1998, and 2008.

As in the past, compensation for risk, particularly in the form of investment grade credit spreads, has accompanied volatility in its long descent. The good news is that dramatic spread compression have produced strong price returns, and accounts we manage have benefited accordingly. Over the period since spreads began to compress 18 months ago, all of our credit-oriented strategies have outperformed their December 2015 starting yields by one-and-a-half to two times. Unfortunately, as you will see below, there is less potential in today’s valuations to achieve these kinds of results. Reaching back into previous Strategy Updates, we will roll out our favorite quote from year-end 2015,2 but bring it up-to-date with a red editor’s pen:

Q2 2017 BBH Taxable Fixed Income Strategy Quarterly Update

This Strategy Update takes a closer look at corporate valuations, exploring what has happened after the other times they’ve been this rich. We also review the latest interest rate and Federal Reserve (Fed) activity – rate hikes, taper plans, and stress-tests.

Summary

  • Corporate bond valuations are more expensive than they’ve been since before the Financial Crisis. History suggests that from these levels, index corporate credit is likely to underperform U.S. Treasuries over the next few years. Fortunately, our portfolios don’t resemble the indexes.
  • Within the corporate credit universe we find ourselves increasingly focused on high quality, liquid, short credit and loans.
  • Asset-backed securities (ABS) and commercial mortgage-backed securities (CMBS) still offer some value, and we continue to maintain positions in this sector.
  • As a result of the declining opportunity set, spread duration (sensitivity to changes in credit spreads) in our representative core account has decreased from 4.3 years at the bottom of the cycle in February 2016, to about 3.2 years now.
  • The yield curve has flattened dramatically. We are using available reserves to reduce the rate “barbell” that has built up in our most attractive credit investments.
  • The Fed is likely to stop reinvesting its U.S. Treasury and mortgage-backed securities (MBS) holdings by the fourth quarter of this year. We own very little MBS, and welcome the modest rate and MBS spread pressure expected from tapering.
  • The Fed also released its latest stress tests for U.S. Banks, which is based on deep knowledge of the current state of loan portfolios. They reveal that the only sector where the Fed expects any significant erosion in credit quality is consumer lending. Bank loan quality is often seen as the “canary in the coalmine” for economic downturns, but the canary appears to be alive and well.

How Low Can It Go?

Investment grade (IG) and BB/B-rated high yield (“HY”) corporate option-adjusted spreads stand at 116 and 296 basis points,3 respectively. Spreads have been at such low levels only a few times over the past 20 years, and in most cases did not remain at their lows for very long. The four-year period between 2003 and 2007 stands out as an exception. Referring to our earlier comments about volatility, the severity of spread widening seems proportional to the time spent at the lows. Consider, for instance the 2008 crisis following four years of dismally tight spreads.

Q2 2017 Investment Grade and BB/B High Yield Spreads

These index spreads, however, are incomplete for two reasons:

  1. The IG credit market is both longer in duration and lower in credit quality than it was in the pre-crisis era. Issuers are carrying greater leverage. In addition, the sector mix has changed substantially, with increased Yankee (non-U.S.) and 144a issuance introducing more country and liquidity risk, as well as large changes in the composition and size of the financial sector. All of these changes suggest higher spreads are required.
  2. The index spread is an average, and gives no sense of the distribution of spreads across sectors, potentially masking sector-specific and individual opportunities.

Our valuation framework adjusts for spread duration, quality, and sub-sector, and therefore gives a consistent representation of opportunities over time. We identify adjusted spread levels at or above which individual bonds are very likely (based on history) to outperform U.S. Treasuries, in excess of an appropriate cushion, or “margin of safety”4. We call these “buys” or “in the buy zone"5. Correspondingly, the level at or below which the bond is likely to underperform a similar duration U.S. Treasury we call sells. Bonds where there is some excess return available, but with a shrinking margin of safety are “holds”. In every case, we calculate and store an expected return at the current spread levels. Analyzing the output of our valuation framework can therefore give an “Apples to Apples” view of compensation for risk in bonds. When we use the terms “buy”, “hold”, “sell”, and “expected excess return” in the paragraphs and charts that follow, we are referring to our internal valuation screening system, not making a recommendation.

The current environment presents a limited set of attractive securities. Only 8% of the IG index and 22% of the HY index are in their respective buy zones. Based on our framework, valuations have been this rich only a few times before. Plotted to the right are a series of monthly observations (2000-2017), showing the percent of bonds in the buy zone in our valuation framework at the beginning of each quarter (horizontal axis) compared to the excess return to credit over the next three years (vertical axis). Once the percentage of buys is under about 15%, you can see that credit excess returns over the next three years are largely negative. Today’s valuation levels are marked by the red line. From this level, an index exposure to corporate credit has underperformed U.S. Treasuries over the next three years in all but one of seven observations.

Q2 2017 Expected Returns at Different Valuation Levels

Furthermore, the opportunities available are generally not deep in the buy zone. In the histograms below, the vertical line separates buy-rated securities from holds and sells (to the left of the line). The horizontal dimension indicates level of potential excess returns according to our valuation framework. Anything to the right of the line constitutes a buy and the distance to the right indicates the magnitude of excess returns available. Note that today, as in the 2003-2007 period, the right tail is very thin. Recalling the time series of spreads presented above, history suggests that when spreads begin to widen, they do so quickly. Reversion to wider levels will provide a larger and stronger set of opportunities for sure, but the interim period may deliver some pain as well. Consider the widening episode between May 2014 and February 2016. IG corporate bonds underperformed Treasuries by over 5% cumulatively and HY corporates by over 9%. On the plus side, the size of the buy zone increased to 65% and the right tail gained substantial mass.

Q2 2017 Valuation Opportunities in the Investment Grade Universe

There’s another aspect of this valuation desert that is particularly troubling: At prior spread tights in 2011 and 2014, there have been some pockets of distress where we’ve been able to find good values, often driven by fear-based selling. Many of our largest performance contributors over the past months came from these sectors where an excess of pessimism or resistance took hold – sectors such as Real Estate Investment Trusts (REITs), CMBS, energy, banks, and consumer finance, oversold individual names like Transalta, and Freeport-McMoran, and priced-to-move new deals like Western Digital and Dell. So where might one find excess pessimism today? If you read the newspapers, you might think retail is a good candidate. Department stores are closing at a record clip and Amazon is busily trying to dominate even the grocery business. Shouldn’t this sector have a few names trading at attractive levels that are nonetheless durable businesses?

The scatter plot below plots each bond (a dot) in the leisure, retailers, and retail REITS sectors. The green dots are at buy levels in our valuation framework. Holds are blue and gold are sells. In this graph, the Y-axis is the potential excess return net of our margin of safety, and the X-axis is the duration of the bond.

The majority of those green dots are Macy’s bonds, along with bonds of two retail REITs, DDR Corp and CBL & Associates (we sold the latter a little while ago). In our view, Macy’s is already a high yield credit, thus the compensation is not so attractive, and we are unconvinced it is a durable credit. As you can see, the vast majority of the sector is not trading at a level that is likely to outperform U.S. Treasuries by much, if at all.

Q2 2017 Leisure Retailers and Retail REITs Sectors

There are still values in very short corporate bonds, generally around 13 month maturities. Spread tightening and big price gains are not available there, but at least these securities offer some excess return while we wait for more opportunities. All of you will have noticed an influx of short bank and highly-rated industrial paper in your portfolios. These purchases have kept corporate allocations up while the longer bonds hit our sell thresholds. These are usually larger issuers, giving us the ability to sell the short paper in the happy event that opportunities arise before they mature. In fact, many of them are banks, as shown in a scatter plot of the senior banking issuance on the right.

As a result of this change, we have reduced market value exposure to corporate credit in core bond accounts by about 1%-3%. However, spread duration (sensitivity to spread changes) is down 0.5-0.75 years over the last six months, and more than a year (4.3 to 3.2 years) since the bottom of the credit cycle in February of 2016. We are generally carrying less corporate credit risk than in our clients’ benchmarks.

Q2 2017 Senior Banking Issuance

Higher London Interbank Offering Rate (LIBOR) has helped, but Index Exchange-Traded Funds (ETFs) are probably the primary reason very short credit screens well. Paper under 13 months falls out of the major fixed income indices and is dumped, unceremoniously, at the end of each month by the big index ETFs. We are very pleased to take advantage of traders moving large amounts of money for non-fundamental reasons.

Structured Products Not As Overbought

ABS and CMBS have also been increasing in price, but their valuations are not nearly as stretched. Structured product is not trading at pre-crisis tights, and still offers a significant yield advantage to index corporate credit. These sectors neither widened as much as others in the 2014-2016 cycle, nor tightened as much as corporate bonds over the past year. As a result, our portfolios are maintaining their ABS allocations, although we have allocated away from a few overpriced sectors and issuers. New issue on-the-run spreads are tight in CMBS as well, due to strong commercial real estate fundamentals and positive investor tone towards the newly implemented “risk retention” requirement. In the second half of 2017 we expect to find investment opportunities in secondary market in the seasoned “pre risk-retention” conduit deals or single-asset, single-borrower (SASB) deals collateralized by property types that have fallen out of favor, such as retail department stores and malls.

We are often asked why the yields in structured product persist, even as corporates are so frothy. We believe the causes are two-fold: First, there is a smaller universe of knowledgeable buyers in structured product. Very large investment firms have a hard time getting meaningful positions, and large institutional buyers often still preclude structured product from their investment portfolios. Secondly, there are no large index-driven buyers of structured product such as ETFs. Index ETFs have become one of the most influential buyers of corporate securities, and have contributed to rich valuations across all subsectors of corporate credit. JP Morgan has just built a broader index of ABS, which may lead to increased ETF participation in the sector. Should it happen, it might be very positive for prices.

Q2 2017 Structured Fixed Income Valuation

The largest change in interest rates over the past three years has been at three years and under, where yields are up about 1%. As is clearly visible in the graph to the right, the U.S. Treasury yield curve has flattened dramatically, meaning there is even less value (in yield and roll-down potential) in investing in longer maturities. Longer maturities still provide a hedge against deflation and recession, but that is a relatively expensive hedge at the moment. On the other hand, near-term Fed hikes seem fully priced into the curve, reducing the risk of locking in rates in the two to five year part of the curve. Our accounts have been “barbelled” in short ABS and longer bonds where we found value. A barbell structure has enhanced returns into the flattening curve. As we increase reserves, however, we have been neutralizing that barbell and will look to take advantage of roll down in the steeper intermediate part of the yield curve.

Q2 2017 US Treasury Yield Curves

Taper Humdrum

The Fed owns $2.5 trillion of U.S. Treasuries and $1.8 trillion of agency MBS. All eyes have been on the Fed minutes as to when and how fast it will let this inventory decrease. At their recent June meeting, policymakers reaffirmed the approach to balance sheet normalization and agreed that reductions in the Fed’s securities holdings should be gradual and predictable, and accomplished primarily by phasing out reinvestments of principal received from those holdings. To that end, they will gradually raise the amount of maturing principal that is not reinvested. The Fed appears to view low inflation as a transitory phenomenon, and we expect tapering to begin in September. Our interpretation of the Fed’s remarks is that they could reduce their holdings by as much as $800 billion by September of 2019. Since the Fed will not be actively selling securities, we do not expect the taper to have a dramatic effect, but it will put some upward pressure on rates and MBS spreads.

Relative to the market’s reaction in 2013 (a full 1% on U.S. Treasury notes of five year maturity, and a 20 basis point jump in spreads), credit and U.S. Treasury markets are taking this eventuality in stride. In fact, the five-year U.S. Treasury yield is only about 10 basis points higher in yield now than at the end of the “taper tantrum” in 2013, when the Fed simply suggested they would stop buying. Spreads are about 40 basis points lower now than in the taper tantrum.

Stress Tests Less Stressful

As indicated above, our portfolios hold a large allocation of short bank paper. The Fed just released its latest round of stress tests for the major U.S. Banks. It was not a surprise that most banks passed with flying colors. Nonetheless, the news kicked off a large rally in bank stocks as banks lined up to return capital to shareholders and make investments with their newfound excess capital.

What is revealing about the stress tests is how regulators change their assumed losses in different categories. The regulators setting these levels have far more loan information than we do, with detailed inside access to charge-offs and delinquency trends at every major bank. For that reason, their changing assumptions give a sense of what they are seeing, fundamentally, in bank loan books. As illustrated in the accompanying table, stress assumptions for bank lending are becoming less and less stringent. In contrast, credit cards and consumer finance are seeing increases in stress assumptions, albeit not substantial. Commercial and industrial loans and real estate stress assumptions are flat. The Fed also sees loan losses trending downward overall.

The stress test assumptions suggest that regulators are not seeing a negative change in bank credit, and only sound a light note of caution in the consumer loan and credit card sectors. Recession levels of loan delinquency would be a surprise. We continue to have a positive credit view of many U.S., Canadian, and Australian banks.

Q2 2017 Bank Lending Stress Assumptions

Conclusion

While Europe’s economies are doing better, there are still uncertainties in emerging markets and petro-economies, China growth and demand remains a conundrum, and it still isn’t clear why inflation remains so low in the U.S. Finally, it is undeniable that the Trump Administration’s political approach is to shake things up, be combative about trade with major partners, and try out novel approaches to Syria, North Korea, and the Middle East, generally. Trump supporters and detractors alike should concede he represents a geopolitical tail risk. There seems to be no shortage of political and other risks to act as a catalyst for poor credit performance, higher volatility, and, of course, improving credit valuations.

Investors should be wary of this level of optimism: when nothing but good news is built into prices, all it takes for risk assets to badly underperform Treasuries (and become good value again) is a little bit of unanticipated bad news.

We remain, as ever, guided by our valuation work, looking for durable credits at attractive yields. Until the unanticipated bad news surfaces, we will be in a more difficult period for generating excess returns, especially given our shrinking exposure to corporate credit risk. However, thanks to lingering opportunities in structured product, our portfolios continue to earn above-benchmark yields even as they build up reserves. Our ability to respond quickly to valuation changes across sectors should help us continue to generate excellent performance. We will use valuations, not timing, to determine our positioning, as we believe that will produce the best long-term outcomes for our clients, and the investments we have made alongside our clients in our funds.

We look forward to meeting with you this quarter. We thank you for your business and for your patience as we wait together for the market to startle awake and overreact, sensing the presence of the monster under the bed.

Sincerely,

Andrew P. Hofer
Portfolio Co-Manager

Neil Hohmann, PhD
Portfolio Co-Manager

1 Business Insider, July 6, 2017, http://www.businessinsider.com/markets-hottest-trade-volatility-etf-short-vix-biggest-day-yet-2017-7.
2 Keen readers will note that this is actually the third time we’ve used this language. Repeating yourself is a hazard of value investing.
3 Basis point is a unit that is equal to 1/100th of 1% and is used to denote the change in price or yield of a financial instrument.
4 With respect to fixed income investments, a margin of safety exists when the additional yield offers, in BBH’s view, compensation for the potential credit, liquidity and inherent price volatility of that type of security and it is therefore more likely to outperform an equivalent maturity credit risk-free instrument over a 3-5 year horizon.
5 Please remember that this is only our valuation screen. A “buy” in our framework is not a recommendation, but merely makes a security a candidate for credit research. The next step is to make sure the credit is durable, rather than eroding and fragile.