Stock/bond correlations have been steadily decreasing since peaking in 2015: What does it mean?

Key takeaways

• A feature of October’s 11.5% intra-month S&P 500® decline (ending in a 1-month return of –6.8%) was that bond yields barely moved.

• The 5-year correlation between stocks and bond yields peaked in 2015 and has come down steadily since. Will it flip to negative like it did during the mid-1960s?


• If so, investors may need to reconsider how to protect against downside equity risk.

• That doesn’t mean bonds no longer play an important role in a diversified portfolio, since they still provide income, diversification, and lower volatility.

One of the notable things about the S&P 500’s 11.5% intra-month decline in October is that it left investors with basically no place to hide, other than in cash or a defensive sector such as utilities. Growth/momentum exposure offered no reprieve, and cyclicals and interest rate-sensitive groups got clobbered.

So more or less everything got hit, which can be seen in market-breadth indicators that compare the number of companies advancing versus declining. According to data from Bloomberg, an 11.8% intra-period decline in January/February 2018 produced a modest 801 new 52.week lows (across all exchanges), but the almost identical 11.5% decline this October produced a crescendo of 1,916 new lows.

But most noteworthy, I think, was that bonds offered no safe haven either. Yes, the 10-year Treasury yield fell from its high of 3.23% to 3.07% for a few weeks (producing a roughly 1% return), but those gains proved short-lived, as the 10-year yield quickly moved back to 3.22%.

Bond yields up while stock prices down? That is certainly a twist to the market playbook that we have gotten so used to over the past two and a half decades. Since the late 1990s, most market shocks have carried with them fears of deflation, which meant that drawdowns in stock prices were generally associated with declines in yields.

This has, of course, been very helpful to most broadly diversified portfolios, which rely on bonds for a measure of downside protection.

But what if things are moving from a positive correlation between bond yields and equity prices to a negative one? I think that could be groundbreaking, to say the least, because it could suggest that we are shifting from a deflationary to an inflationary regime. That could influence expected risk premia and valuations across many or even most asset classes, and it could change what investors might want to have in their tool kits with regard to diversifying assets.

So, let’s see what we can learn from history. Taking a look at some very long-term data, we can see that the correlation between interest rates and stock prices is ever evolving. Exhibit 1 shows the rolling 5-year correlation between monthly changes in the inflationadjusted S&P 500 price index and monthly changes in the inflation-adjusted yield on long-term Treasuries.

Like a pendulum, the rolling correlation is always ebbing and flowing (between roughly +0.5 and –0.5), but it can persist, positively or negatively, for long stretches.

For me, the key inflection points are 1960 and 2015. In 1960, the 5-year correlation swung from persistently positive (stocks up, yields up) to persistently negative (stocks up, yields down), as the secular backdrop evolved from the deflationary mindset arising from the Great Depression to the inflationary mindset of the late 1960s and into the “stagflationary” ’70s.

Once the correlation swung to negative in 1960, it more or less stayed there until the late ’90s, by which time the deflationary mindset was returning—first with the “Asian Flu” in 1997 and the 1998 collapse of hedge fund Long- Term Capital Management, followed by the bursting of the dot-com bubble in 2000, and finally the 2008–2009 global financial crisis (GFC).

With the exception of early 2009—when yields were rebounding from their GFC lows while equity prices were still falling—the correlation had been persistently positive from the late ’90s until 2015. Since then, the 5-year correlation between the real S&P 500 price index and real bond yield has been steadily falling, and as of October 2018 it stands at roughly zero.

The next step in this study is to create an analog of these two periods, lining up the two inflection points of 1960 and 2015 to see what we might learn. So, we are now at “1964” in Exhibit 2, which compares the historical and recent trends of the S&P 500 nominal total return in the top panel and the two periods’ 5-year rolling correlations in the bottom. Exhibit 3 shows the S&P 500 real return and the 5-year annualized inflation rate for each time frame.

The analog is compelling, to say the least (with the usual caveats). For one, I find it interesting that there was a large tax cut in 1964 (the United States Revenue Act of 1964), similar to today’s. Also, we know that deficit spending started ramping up in the later years of the 1960s. Third, the ’60s followed a long period of ultra-low interest rates and even something akin to the U.S. Federal Reserve’s quantitative easing (via the Fed’s interest rate caps during and following World War II). Finally, stocks were well into a secular bull market during the ’60s that had begun (by my count) in 1949 and which ended in 1968. We won’t know for a long time whether the current post-GFC market regime is a secular bull or not, but if it is, then the parallels to the ’60s are even more compelling.

Pages ( 1 of 3 ): 1 23Next »