Spreads, Default rates, Recovery Rates, Interest Rates and Dispersion
By Hamlin Lovell, NordicInvestor
The allocators interviewed by NordicInvestor have different opinions about the outlook for high yield credit spreads, default rates, recovery rates and other variables. We interviewed Insight portfolio manager, Paul Benson, to gather some historical and analytical perspectives.
For some investors, switching from equities to high yield is perceived as “de-risking”. While for others, increasing credit ratings by moving from high yield to “fallen angels” or investment grade would be a de-risking move. Insight manages “efficient beta” strategies in all of these credit asset classes, discussed in this previous interview.
Whether corporate credit spreads are historically rich or cheap partly depends on the lookback period, and on what adjustments are made to compare like with like over time.
The last 14 years of central bank QE and asset purchases was clearly not a normal period for credit spreads. Equally, including the genesis of the US high yield market in the late 1980s may not be very useful as the high single digit to low double digit default rates of junk bonds has come down since their usage expanded. A sensible starting point could be the early 1990s, and headline US high yield spreads are now somewhat below the long-term averages since then.
Of course, those averages are inflated by crises such as TMT and Enron 2001-2002, GFC 2008-2009, energy 2015-2016 and COVID 2020. Absent any palpable crisis, spreads should arguably be lower.
However, the average spread alone is a crude metric because the high yield universe itself has changed over the past 25 years.
Credit ratings within high yield
“25 years ago, there were more CCC and single B rated bonds and fewer double B rated bonds. Since 2000, the overall quality of the high yield universe has improved as the BB share has grown from about 30% to 50%. Meanwhile, the opposite has occurred in investment grade where the share of triple BBB has grown at the expense of higher quality bonds such as AA and A. This probably reflects the incentives of CFOs who want to optimize the cost and availability of funding,” explains Benson.
A more granular breakdown, comparing spreads per credit rating notch is somewhat more meaningful, though this still ranks high yield spreads as slightly below long-term averages.
Leverage and cashflows
Yet drilling down into individual credit ratings still overlooks structural changes in corporate finance. “Fundamental balance sheet structures have changed. Companies with memories of recession and downgrades have become more conservative. Leverage ratios are lower and interest coverage is higher. Companies used COVID to refinance and extend maturities out to 2025 or 2026, so default probabilities are lower,” says Benson.
Default forecasts and projections
Some eye-catching default forecast headlines might not give a realistic impression of the probable investor experience, based on history.
Ratings agencies’ recent default forecasts of 4-5% are equal weighted by names. “Market capitalization weighted measures of defaults (of which par-weighted is the most conservative methodology) show an average of 1.4% per annum over the past 20 years,” says Benson. Additionally, the agencies’ definition of defaults includes distressed exchanges, which are not perceived as defaults per se by many benchmark providers.
In reality, the phrase “default forecast” may lend spurious precision to a messy and uncertain exercise that might more meaningfully be termed a “projection”, as in some areas of economic forecasting. “It is difficult to predict multi-year default rates on a bottom-up basis, as these events are few and far between. Further out, the unpredictability becomes more difficult due to super linear curvature. The probability of accuracy goes down by orders of magnitude the longer the forecast,” explains Benson.
Notwithstanding these uncertainties, it is likely that US high yield defaults will rise from the 0.69% rate seen in 2022. “This was almost zero so it cannot get much better and is more likely to get worse. Higher interest rates, more sticky inflation, supply chain issues and geopolitics all provide challenges for corporates around the world,” says Benson.
The Insight US high yield efficient beta strategy has historically avoided 30-50% of index defaults, including in 2022, albeit the small sample sizes make it hard to have high statistical confidence in the results.
Even if defaults do pick up, they are often quite ephemeral. History since the early 1990s suggests that it would probably require persistent economic and/or financial market challenges – and an absence of policymaker responses – to sustain a multi-year default rate of mid-single digits or more. “Even in the GFC, rolling 12-month default rates topped out at 5-6%. But once this stretches out to a 2-3 year horizon, annualized rates revert towards the long term averages of 1.5%,” observes Benson.
Recovery rates and covenants
Though the default sample size has recently been very small, recovery rates on high yield debt have been above historical averages. “The historical range has been 20% to over 50%. We have seen an improving trend with recent recovery rates above 40%,” points out Benson.
Arguably, it is in the leveraged loans market where weaker covenants, enabled by more customized term sheets, could be contributing to lower recovery rates – and rate rises could already be precipitating some distress, since floating rate borrowers are feeling the heat from higher interest rates almost immediately.
Liquidity and illiquidity premia
In addition to default risk premia, there is some debate over what if any level of illiquidity premia is needed for holding high yield corporate debt in different regions, and how this has changed over time. In the Nordic corporate bond markets there have been episodes when some, but not all, funds suspended valuation and dealing, for instance in March 2020. And individual high yield bonds can be illiquid and expensive to trade in all geographic areas. “But the US high yield market can be accessed through more liquid wrappers including ETFs and other funds availing of basket trading mechanisms used by ETFs” says Benson.
These techniques considerably reduce bid/offer spreads and transaction costs. Transaction costs in 2022 were broadly consistent with previous years, though Benson does caution that some binary market phases make it more difficult to measure the costs: “during periods of one way risk on or risk off market activity, it is hard to gauge true bid ask spreads if we only see one side of the universe”.
Macro, liquidity and risk appetite
The macroeconomic, liquidity and broader risk appetite is also relevant for high yield, especially where top-down investors trade it as an asset class.
There are fears that the move from QE to QT could be a headwind. “Quantitative tightening from central banks could have more impact in Europe given the size of the ECB programs. The US has no run-off on that side,” says Benson.
Rate rises beyond what is factored into forward curves are another risk factor for high yield. “The level of credit spreads provides some degree of cushion against terminal rates being one or two quarter points above current expectations,” argues Benson.
Many market participants in 2022 were surprised to see the VIX index of equity implied volatility declining, but the real action lay in the bond markets where the MOVE index was at elevated levels. “The decline in MOVE in late 2022 and early 2023 has helped to support high yield corporate debt,” claims Benson.
Fallen angels
Benson clearly has a sanguine outlook for high yield, but investors who are seeking some additional diversification within credit allocations, and/or are more pessimistic on defaults could consider the fallen angels segment: high yield issuers that were previously investment grade. Fallen angels’ market cap of USD 140 billion in the US and USD 180 billion globally (global corporates excluding EM) makes up about 10% of the US and global high yield universes of USD 1,400 and 1,800 billion respectively.
“Some 85% of fallen angels are BB rated and in contrast to original high yield issuers, they want to reclaim their IG status to get lower funding rates. Longer term, fallen angels rarely underperform high yield and when they do outperform, it can be by a large margin. We view fallen angels as almost like high yield but with a free call option,” says Benson.
The anomalies can arise from forced selling and forced buying. Once issuers lose their membership of the USD 6 trillion IG universe, many investors are forced to sell, and the names can become “orphaned”. They fall between two stools: un-investible by strict IG mandates but also not yielding enough to appeal to many pure HY investors. Once they regain IG status, it is often a good time to take profits.
Dispersion and alpha from factors
Whichever credit asset class investors prefer, the climate has lately become more conducive to alpha generation. Increased dispersion within credit markets improves the opportunity set for discretionary and systematic active, semi-active and efficient beta managers and strategies to generate alpha.
For example, Insight’s efficient beta range uses proprietary quantitative factors to generate alpha commensurate with fund or client risk budgets.
“We do not attempt any macro market timing, which is most difficult, and find sector allocation somewhat less difficult, but generate good information ratios from selection within peer groups. In 2020-2021 there was almost no cross-sectional dispersion but this has picked up in the latter half of 2022, which allows us to add alpha,” says Benson.
“Our quality signal has helped us to get a lower loss given default than indices. Our value signal has started to kick in and find winners and losers during the second half of 2022. We also built a slower moving momentum factor into our quality signal to avoid the large transaction costs of implementing a pure momentum trading signal,” he explains.