- With elevated inflation, slowing growth from the Covid rebound and a historically tight labour market, stagflation concerns remain high among investors. And, with the Fed embarking on one of the steepest hiking cycles since the 1990s, there are also growing concerns over recession risk.
- To gauge the market-implied probability of a recession, we look at a broad range of leading and coincident market indicators. Currently they send a mixed message, with cyclicals vs. defensives valuations and the 2s10s yield curve more bearish, and other indicators less so.
- Current levels of market-implied recession probability of between 20% and 30% (based on leading indicators) have historically been followed by a recession 28% of times within 12m. Only when the recession probability was above 60% were there almost no instances where a recession did not follow within 12m.
- Market timing around recessions is notably difficult. Market-implied recession probabilities can help but investors need to balance the ‘time in the market’ with ‘timing the market’ as divesting too early can mean giving up positive equity returns. We remain OW equities in our asset allocation but would focus on managing risks with tail risk hedges.
Growing Risks But Too Early For Recession Obsession
With elevated inflation, slowing growth from the Covid rebound and a historically tight labour market , stagflation concerns remain high among investors – especially for Europe, amid the energy crisis due to the ongoing Russia/Ukraine war. With the Fed embarking on one of the steepest hiking cycles since the 1990s, there are growing concerns over recession risk: the US yield curve, the 2s10s spread, inverted at the end of March . This is usually an early recession warning: as we wrote before , since the late 1980s the time lag between US yield curve inversion and US recessions has been on average 20 months. That said, there are often wrong signals – in periods of high inflation the yield curve has inverted deeper before a recession and combining different segments of the yield curve points to a low probability of recession in the next 12m.
To gauge the market-implied probability of a recession, we screen a wide range of leading and coincident indicators and combine their signals into two measures of recession risk: the probability of entering a recession within 1 year, and that of already being in a recession. On average, both the leading and coincident indicators have picked up YTD but they are currently not pricing much risk (Exhibit 1 & Exhibit 2 ):
- For the coincident indicators, we use the S&P 500 1-year drawdown, the excess bond premium, USD HY credit spreads and the VIX, which all increased sharply in Q1 due to Fed tightening fears and the Russia/Ukraine war. However, they have declined again materially – the likelihood of an imminent recession is priced very low now at 6%, which increases the risk of disappointment in the event of a sharp growth slowdown.
- For the leading indicators, we use the 2s10s yield curve, the 12m forward implied change in the Fed funds rate, cyclicals vs. defensives P/E ratios and MBS spreads. Only the 2s10s yield curve and cyclicals vs. defensives P/E ratios are pointing to higher than normal recession risk – the others point to very little risk.
Current levels of market-implied recession probability of between 20% and 30% (based on leading indicators) have historically been followed by a recession 28% of times within 12m (Exhibit 3 ), which is broadly in line with our economists’ assessment of the risk of the US entering a recession over the next year . On the flip side, the very low probability implied by coincident indicators suggests that markets have faded imminent recession risk, particularly since the strong recovery in risky assets last month.
While the recession probability embedded in leading market variables has increased YTD, it still points to a relatively low level of risk. Historically, the market-implied recession probability had to be above 30% for a recession to occur 1 out of 2 times within 1 year, and above 40% for a recession to occur 2 out of 3 times within 1 year. Only if the recession probability was above 60% were there almost no instances where a recession didn’t follow within 12m, but such elevated levels were reached only before 1990.
Market timing around recessions is notably difficult. Equities are forward-looking, peaking a few months before a recession starts and troughing before it ends. But before the market sell-off starts, equities can still deliver positive returns: in the 1-year before a recession the distribution of equity returns is usually similar to the unconditional one.
Market-implied recession probabilities can help market timing – if leading indicators point to a higher probability of a recession, equity drawdown risk increases. Exhibit 4 shows that once the probability of a recession in the next 12 months increases above 20%, similar to now, the risk of corrections is somewhat higher than normal. However, the risk of larger and more frequent drawdowns increases in particular once that probability is above 40%.
Scaling down the risky asset exposure based on recession risk requires a balance between ‘timing the market’ and ‘time in the market’. Exhibit 5 shows the annualised total return vs. the S&P 500 of a stylized strategy that switches from equities to cash if the market-implied probability of a recession within the next 1 year is above a given threshold.
Switching to cash at lower recession probability levels decreases the drawdown risk of the portfolio at the cost of giving up positive equity returns – either ahead of the market correction or if the indicator is giving a false signal. Unsurprisingly, de-risking too early tends to lead to lower returns than never divesting from equities.
All in all, reducing equity exposures based on recession probabilities outperforms only if a deep bear market materialized, usually around actual recessions (Exhibit 6 ). We therefore remain OW equities in our asset allocation (and also are OW commodities & cash, UW bonds & credit): while we acknowledge the combination of geopolitical uncertainty and hawkish policy increases the risk of a correction even without a recession, we still think equities stand a better chance of beating inflation in the long run.
Given that timing the market can prove particularly challenging, we would focus on managing risks with tail risk hedges. Since mid-March, cross-asset implied volatility has reset, especially for risky assets – equity tail risk hedges appear again relatively attractive, even though equity implied vol still looks expensive in absolute terms. In addition, rates volatility appears high vs. implied equity volatility (Exhibit 7 While we think that uncertainty on rates remains high, especially in the front end ). , we would now have a bias for selling rates vs. equity vol. In the near term, with elevated inflation, large declines in rates are unlikely without much higher recession and growth risks, which might weigh more on equities – selling receivers on USD rates could help f inance equity downside hedges.