Given the spread widening that has occurred in credit over the past 18 months, Nikko Asset Management Australia's credit team has been trying to determine when it makes sense to increase our credit exposure. We have generally been of the view that being risk-averse through the US Federal Reserve's (Fed's) hiking period (at least in the early stages) made the most sense as volatility would return to the market. Given that the Fed's January statement indicates that its outlook is a little more dovish than expected, the end of this risk-off period could be coming closer. We have analysed the potential outcomes for credit spreads under different scenarios, including a theoretical recession, and believe that the time for getting back into credit may be nigh.
Is US economic data indicating a recession?
Except for employment, the improvement in US economic data has been slowing, making it harder for the Fed to increase rates. Certain economic models are putting the probability of a recession at around 20% - according to Bloomberg's January survey and big US banks, such as Morgan Stanley and Bank of America1. Although a recession isn't our base case, some economic data is concerning and warrants being aware of downside risks.
Chart 1 indicates that corporate debt has been steadily climbing over the past few years, with corporate debt to GDP at cyclical highs, similar to 1990, 2000 and 2008.
Chart 1: Corporate debt to GDP
Chart 2 shows industrial production (IP)—when IP falls like this, it usually indicates that we are heading into a slower period of growth (see also 1990, 2001 and 2008). It is not a healthy sign for an economy when high debt is combined with falling IP (profits). In comparing current conditions with the past, we have focused on the 2000 period because this appears the most similar to the current situation. Both 1990 and 2008 involved banking crises ? 1990 followed the savings and loans crisis and 2008 was the GFC. Any recession/period of slower growth is typically harsher when banks are involved.
Chart 2: Industrial production vs. US GDP
However, this time it is unlikely that we will see a banking crisis because the banks have succeeded quite well in deleveraging since the onset of the GFC (see chart 3). Furthermore, the late 1990s and early 2000s were characterised by weak emerging market performance, which is similar to what is occurring today.
Chart 3: Financial sector debt to GDP ratio
Analysing how credit spreads should perform under different scenarios
Let's assume that if there is a recession that it would be similar to 2001—a corporate crisis rather than a banking crisis. We can use IP to try and estimate how credit spreads should behave. As chart 4 shows, IP fell to approximately -5% over six months in the early 2000s (a rising line indicates falling IP). If a similar pattern were to occur today, according to our estimates this would see credit spreads widen to around 400 bps (from around 350 bps currently).
Chart 4: BBB spreads vs. IP (inverted)
Sources: Bloomberg and Nikko AM
Using this scenario, we can estimate expected returns if credit spreads widen by a further 55 bps, bearing in mind that credit has already suffered through 2015. Table 1 shows Australian corporate spreads to Commonwealth government bonds by maturity. The second half of the table shows the estimated expected return in a recessionary environment assuming government bond yields are unchanged. In that scenario, 0-3 year credit maturities should still outperform, but 3-5 year maturities will be negatively affected due to the longer duration. The 5-10 year maturities don't make sense to own from a spread perspective unless you are forecasting compression.
Table 1: Australian corporate spreads to Commonwealth government bonds by maturity
|Expected spread if credit widens 55bps||1.61||2.11||1.78||Recessionary Environment||0-3||3-5||5-10|
Bloomberg and Nikko AM. Expected outperformance calculated as ((Current Spread - Expected Spread) x Duration) + Current Spread
However, in our view, a recession is currently a low probability outcome. It is probably equally likely that if IP continues falling, the Fed will either stop raising rates or provide further easing—negative rates or QE are not necessarily off the table if the US recovery falters.
In our analysis, we assumed a 50% chance that credit spreads stay flat at current levels, a 30% chance of recession and a 20% chance that spreads contract (by around 30 bps). Table 2 shows each of the expected returns weighted by their probability. In 70% of these scenarios (even when spreads are flat), then credit is going to outperform. The blue section at the bottom shows the weighted-average outcome.
If the poor data discussed above means that the Fed refrains from hiking interest rates to 1.25% in 2016, then it is much less likely that the US will experience a recession. This means that based on current spread levels, credit is becoming attractive.
Table 2: Returns weighted by probability for expected return under each scenario by maturity
Bloomberg and Nikko AM
In terms of downside risks, the above analysis would come unstuck in two instances. Firstly, if spreads were to experience a 2008-style 'blow-up', then the pain from credit will be much worse than our scenario analysis suggests as spreads widened over 300 bps during that period. However, we think that this is a remote possibility since the banking system is in better health than 2007 and has lower levels of leverage. The second risk is if a recession is actually more likely than the 30% probability assigned above. In this case, holding short-dated credit would still make some sense, but it would be wise to avoid longer maturities.
1 http://www.bloomberg.com/news/articles/2016-01-14/return-of-the-naysayers-u-s-recession-odds-highest-since-13 and http://www.ibtimes.com/probability-us-recession-2016-rises-20-percent-big-banks-say-2276421