Although US equities still have lot of lost ground to make up on long-dated Treasuries since the outbreak of coronavirus, we think that they will do so over time as the Fed engages in renewed financial repression.
- Although US equities still have a lot of lost ground to make up on long-dated Treasuries since the outbreak of coronavirus, we think that they will do so over time as the Fed engages in renewed financial repression.
- The underperformance of US equities relative to long-dated Treasuries began all the way back in October 2018, based on the ratio of returns indices for the S&P 500 and Treasuries with more than 20 years to maturity. Since then, it has reflected a very strong showing from bonds, not just the recent weakness from equities between 19th February and 23rd March (the recent peak and trough in the stock market index). Although the underperformance has started to unwind since 23rd March, as the stock market has perked up, the ratio is still about 30% less than it was at its peak just over a year and a half ago. (See Chart 1.)
- If we look at the period around the Global Financial Crisis (GFC), we see a similar pattern to the one since 19th February, spread out over a much longer period: the ratio of the indices fell more slowly from a peak on 13th July 2007 and recouped less lost ground in a year than it has in the last two months. (See Chart 2.)
- While the speed of the recovery in the S&P 500 this time around suggests that the index may not rise a lot more in the very near term, it eventually closed the gap on Treasuries in the years after the GFC. Admittedly, the ratio only returned to its pre-GFC peak towards the end of the 2010s. (See Chart 3.) But in that decade overall, the average annual nominal return from the S&P 500 was ~13.6%, compared to ~7.4% for >20yr Treasuries. We anticipate that the ratio will also rise further by the end of 2021, for two reasons.
- First, we think that the nominal returns from long-dated Treasuries will be paltry. In the past forty years, these returns have been strong because the nominal yields of long-dated Treasuries have ground lower with the Fed’s policy rate. (See Chart 4.) Our view is that this development has been justified by a secular drop in the equilibrium level of the real policy rate, as we discussed here, since inflation has failed to pick up in the meantime. (See Chart 5.) Nonetheless, the average nominal yield of >20yr Treasuries is now only 1.3%.
- Accordingly, this is the maximum nominal return that >20yr Treasuries will generate from here unless the yield falls to an even lower level. That could conceivably happen if the demand for US safe assets increases, or the Fed changes its tune and cuts its policy rate below zero. For example, in Switzerland, which is also considered a haven and where the policy rate has fallen to minus 0.75%, the yield of >10yr government bonds fell below minus 0.70% in August last year. If the yield of >20yr Treasuries dropped to that level from its current level, we estimate that they would generate a return of something in the region of 40%! Nonetheless, we don’t think that the Fed will go down this route, preferring alternative stimulus instead.
- Second, we anticipate that the stock market in the US will recover more. As we discussed here, it did so in the first eighteen months of all but three of the 14 economic recoveries since the late 1920s, even when it had rebounded significantly before the end of the previous downturn. The main exception was after the mild recession of 2001, when the dot com bubble continued to deflate until late 2002. (See Chart 6.)
- What is more, we suspect that the stock market in the US will benefit at the expense of Treasuries from an extended period of financial repression (see here), akin to what we saw after the beginning of the country’s involvement in WW2 (see here). Between 1942 and 1951, the Fed capped long-term nominal yields at around 2.5% under pressure from the Treasury. And even after the central bank then began to pursue a more independent monetary policy, the yield did not shoot up during the next decade and a half. The Fed’s anchoring of Treasury yields made the stock market, which had been weak before the repression, appealing.
- Indeed, the (cyclically-adjusted) earnings yield of the S&P 500 – the reciprocal of Shiller’s CAPE – fell from 11.8% in the spring of 1942, to 4.2% at the end of 1965. The ride was bumpy at first amid volatile inflation, military conflict (later in Korea as well), and mild recessions. Yet the trend in the earnings yield was down, especially later on, when economic growth picked up strongly and inflation stayed low. (See Chart 7.)
- Admittedly, the earnings yield of the S&P 500 is less now than it was in 1942. But so, too, is the yield of 10-year Treasuries. Indeed, the gap in their yields is currently quite large by the standards of the last twenty years. (See Chart 8.) So, we still think that the earnings yield has the scope to trend lower if the Fed keeps a lid on long-term Treasury yields. That seems possible, given that caps for short- and medium-term Treasury yields were mooted in the minutes of the FOMC’s latest meeting. In that scenario, any failure to prevent the onset of deflation would be a game-changer for equities, but we think that is quite unlikely.
Chart 1: US Equity & Bond Return Indices
Chart 2: US Equity & Bond Return Indices
Chart 3: US Equity & Bond Return Indices
Chart 4: 30-Year Treasury Yield
Chart 5: Fed’s Policy Rate
Chart 6: Change In S&P 500/Composite
Chart 7: US Treasury & Equity Yields & Inflation (%)
Chart 8: US Treasury & Equity Yields & Inflation (%)
Sources: Refinitiv, Bloomberg, NBER, Shiller, Fed, CE
John Higgins, Chief Markets Economist, firstname.lastname@example.org