Over the past two decades, corporate pension sponsors have learned that managing asset portfolios against liabilities is not an easy task. Aggregate funded status for S&P 500 companies fell from a peak of 126% in 1999 into underfunded territory and rose briefly above full funding in 2007. But despite strong asset returns and meaningful contributions since the global financial crisis, the majority of plans remain underfunded. Milliman’s Pension Funding Index shows a funding level of 89.9% for year-end 2018.
Spurred by changes to funding and Financial Accounting Standards Board (FASB) rules, plans have gradually progressed from trying to outsprint their liabilities with portfolios full of risk assets (and suffering the pain of falling woefully behind), to keeping pace with liabilities and taking some small risk to slowly catch up or inch ahead with varied commitments to and implementations of liability-driven investing (LDI) strategies. The past two decades also have seen the rise of passive management, leading more than a few sponsors to focus primarily on management fees, using passive or near-passive long-credit mandates to try and keep pace with liabilities. However, unlike broad equity indices, pension liabilities are not composed of liquid instruments whose weights move in a prescribed rules-driven relationship to one another that would be more amenable to passive management.
The management of pension assets relative to liabilities is an inherently active process on (at least) two dimensions:
1) The pension sponsor must choose how much of the liabilities they seek to hedge on a number of dimensions.
2) They must choose which manager(s) to delegate portfolio responsibility to and understand how and whether the managers’ decisions are in the plan’s best interest.
Managing long-credit portfolios is one area where this is of paramount importance. Liabilities are measured using an AA quality curve for FASB disclosures and an A-AAA quality discount methodology for funding purposes. Over the years some plan sponsors have explored custom benchmarks using these credit restrictions. These strategies can suffer for a few reasons. First, there is a relatively limited opportunity set in the restricted asset space, especially for the AA credits, which leads to undesirable levels of issuer concentration in portfolios. Second, a credit-matched index cannot regularly anticipate downgrades of constituents from those upper echelons. When the downgrade occurs, those credits are removed from the discount methodology but remain in the asset portfolio, so the liabilities will generally rise in value (as one of the highest-yielding securities is stripped from the discount curve) as the portfolio declines, all else being equal. Over time this creates slippage between the assets and the liabilities, with the assets failing to keep pace – even if initially “perfectly” hedged.
To make up for the slippage, the plan sponsor needs to include some credit exposure below the quality level of their liability index to increase overall portfolio yield. BBBs now comprise a little over 50% of the long investment grade market cap and almost 60% of the issuers, according to Bloomberg Barclays. They add some diversification benefits to individual issuers by broadening the market universe in terms of both size as well as number of issuers.
However, as a number of pension plans learned in 2015–2016, attaining the additional spread can come with more volatility and tracking error than anticipated in quiet or good times. There are few surprises that upset a C-suite more than finding out their supposed hedge has not lived up to expectations. Some plan sponsors who had built portfolios across multiple credit managers were surprised to find LDI portfolios that did not hedge liabilities as closely as they had anticipated when BBB spreads widened versus higher-rated credits. Figure 1 illustrates the degree to which this has happened over the past three decades.
The blue line in Figure 1 tracks the relative underperformance and outperformance of the long BBB versus the long A-AAA index. The green line shows the relative option-adjusted spread (OAS) of the two indices. The beginning of each underperformance period is indicated by a blue dot with the number of months that the period lasted until breakeven between BBBs and A-AAA. The periods are summarized in more detail in Figure 2.
The graph and table display a few interesting features:
- While slightly more than half of the periods are six months or less, there have been a few periods that extended to almost three years.
- The magnitude of the underperformance can be quite substantial, reaching -10.8% in September 2002, -8.4% in December 2008, and -9.6% in February 2016.
- Underperformance periods usually correspond to spread (BBB versus A-AAA) widening, but not always. The 2005–2007 event was a period of underperformance despite this spread tightening, indicating relative starting yield spreads can matter quite a bit.
An astute observer will note the relationship between the two lines in Figure 1. Relative performance generally troughs at or near to when the difference in credit spread peaks. This suggests a dynamic strategy: Stay near a market weight of approximately 50% BBBs when they are fairly valued. Add when BBBs are cheap. Lighten up when they get richer. Unfortunately, we’ve seen that many investors wait too long to enter, as snapbacks can be quick. A couple questions worth asking as a pension fund manager:
- Are you (and your bond manager) monitoring the relative attractiveness or richness of going down in quality?
- What is your degree of exposure to individual credits or sectors within BBBs as spreads get rich or cheap?
While the average spread between long BBB and A-AAA is 67 basis points, the annual average realized outperformance of BBBs versus their higher-rated peers is about 64 basis points with an annualized volatility of 3.1%.1 Thus, the returns for taking on that additional risk are not particularly large on average or on a risk-adjusted basis, especially if one considers that BBBs are roughly 7.5 times more likely to be downgraded to high yield in any given year than A rated credits.2 Remember, holding credit is akin to selling out-of-the-money put options. The more one goes down in quality, the closer the sold options are to the money, and option sellers should be adequately compensated for the risk they bear. If the market is pricing that risk cheaply (i.e., credit spreads are too tight to adequately compensate for risk), then investors may wish to seek returns in places where risks are better priced, or temporarily reduce their risk appetite.
But as Figure 1 illustrates, there is wide variation around that average spread or return. The periods of greatest spread and most meaningful trailing outperformance correspond to those years of heightened downgrades, as seen in Figure 3. The busiest year for downgrades of Baa3 can be as much as five times the quietest year. Passively holding credits ignores the dynamism in the process, especially as one gets closer to the high yield threshold.
Significant value can potentially be added if one has the ability to play the long game through the cycles but can make assessments as to whether spreads offer adequate compensation for the risk at a given point in time. As our colleagues have noted, we are getting late into this cycle. Spreads also need to be weighed against changes in rating agency methodology. This creates a need to focus intra-rating as well as inter-rating. For example, ratings methodology for financials is arguably more conservative than it was pre-financial crisis, with a number of firms carrying lower credit ratings as the sector migrated from the AA/A range to A/BBB despite having meaningfully more equity capital and better liquidity profiles. On the other hand, the methodology for nonfinancials has given more focus to interest coverage in the current low-rate environment and less to total balance sheet leverage, with the average debt-to-EBITDA ratio for BBB rated nonfinancial corporates up over one turn from 2000. (See “Investment Grade Credit: Be Actively Beware of BBB Bonds.”)
An additional wild card is the growth and broader holdings of bond ETFs. In calm times they broaden the universe of credit holders, but their perceived liquidity and ease of trading, which makes them useful as a tactical asset allocation trading vehicle, can exacerbate the sharpness of sell-offs or snapbacks if enough users and model followers get spooked. Up-to-date insights into the behavior of rating agencies and market participants are crucial to understanding this piece of the credit markets.
What do we think about the current outlook? After some spread widening in the fourth quarter and renewed tightening in January and February, we see the BBB landscape as fairly valued to slightly rich in aggregate but do see opportunities in sectors that have experienced past traumas and are running less risk. Sectors and names that have increased debt in the good times to lever up their capital structure or expand through mergers and acquisitions require great scrutiny.
Waters in the credit ocean have been reasonably calm for the past two years as the tumult in energy and commodity names has subsided. However, investors in the space need to keep vigilant for rougher seas ahead. When the cycle turns, and all do eventually, the passive ship may be rocked and shaken, inducing credit-triggered nausea, while the nimble portfolio that studies the seas and charts its course ahead of time can pick a quieter path to buffer the storms – and opportunistically ride the wave of wider spreads when there are clearer seas ahead.1 Source: Bloomberg Barclays. Data on average option-adjusted spread from 31 May 1993 to 31 March 2019; data on annual average realized outperformance of BBBs versus higher-rated peers from 31 May 1993 to 31 March 2019. Performance of BBBs is measured by the Bloomberg Barclays US Long Corporate BBB Index; performance of higher-rated peers is measured by the Bloomberg Barclays US Long Corporate A-AAA Index.
2 Source: Moody’s data from 1 January 2000 to 31 December 2018. For our purposes, S&P’s BBB rating is equivalent to Moody’s Baa rating.