Funny story to kick off “Data”, in a nerdy sort of way:
- When Harry Markowitz defended his dissertation on portfolio theory in front of the University of Chicago’s Economics Department in the early 1950s, no less a figure than Milton Friedman argued that the work was ineligible for a Ph.D. It was not “Economics”, he argued.
- Nonetheless, after a brief debate Markowitz received his doctorate.
- In his 1990 Economics Nobel Prize speech, awarded to Markowitz for the same ideas as his dissertation, he remarked that Friedman was in fact correct: “at the time I defended my dissertation, portfolio theory was not part of Economics. But now it is.”
The reason Markowitz’s portfolio theory now sits soundly inside the discipline of Economics is because it explains how rational investors allocate capital, specifically the idea that a return-maximizing/risk minimizing portfolio explicitly requires both fixed income and equity securities. As long as the return correlation between them is less than 1.0, investors will want businesses to split up their balance sheet into debt and equity. These same asset owners will also have a place for sovereign debt, both notionally riskless and risky, as long as the correlations to other assets is less than 1.0. On top of that, there is a place for other financial instruments – venture capital and private equity, for example – as long as they have sub-1.0 correlations.
The question now, of course, is whether Markowitz’s paradigm still holds true in a world of ultra-low interest rates and extraordinary fiscal/monetary policy actions. Simply put, is the classic 60% equity/40% debt portfolio – born directly from Markowitz’s work – dead, hibernating, or still alive?
To answer that question, we pulled the last 5 years of 30-day correlations between long-dated Treasuries, investment grade corporate bonds, and high yield corporates to the S&P 500. Two points here:
#1: The average correlation across this timeframe, which includes a wide array of market conditions, is as follows:
- +20-year Treasury (using the TLT ETF)/S&P 500: -0.36
- Investment grade corporate bonds (LQD)/S&P 500: -0.02
- High yield corporate bonds (HYG)/S&P 500: +0.69
Takeaway: not particularly surprising data, but worth noting that high yield bonds are lousy 60/40-style diversifiers while investment grade has historically been OK, and Treasuries work well.
#2: Stock/bond correlations have shifted materially in 2020 (charts below, our take here):
- Long dated Treasury/S&P 500 correlations did what you’d expect them to do.
They went strongly negative (-0.80) both at the onset of the COVID Crisis in February and again when it became clear the US economic comeback would be slow (July).
Since early August correlations have dropped to just -0.19.
- Investment grade (IG) corporate bonds/S&P 500 correlations went up and have stayed there ever since.
The average 30-day correlation between IG and the S&P 500 since the March 23rd lows for US stocks is 0.38, over one standard deviation (0.29) from the 5-year mean of 0.02 mentioned above.
- High yield (HY) corporate bond/S&P 500 correlations have actually behaved pretty normally through the initial COVID-related economic shock to now.
In late March, the HY/S&P 500 correlation spiked to a 5-year record high of 0.92 and it hit 0.91 in late June.
Since then, however, the correlations here have dropped to 0.61, below the 5-year average of 0.69 cited previously.
In summary, we see three important points from this data:
First, long dated Treasuries continue to show a track record of real portfolio diversification benefits. Yes, rates are low at 1.41% on the 30-year. But they were scarcely 2.0% in February and T-bonds still rallied strongly when capital markets turned “risk-off”.
Second, high yield corporates do not provide much diversification versus equities. In fairness, with a long-run correlation of 0.69, they never have. Forget for a moment about the Federal Reserve’s notional backstop in this market. The reality is that HY is basically equity with a better dividend/coupon.
Lastly and most importantly, investment grade corporate bonds appear to be a new reality where their correlations to equities are structurally higher than in the past. That makes sense. Corporate leverage levels are higher, the out-of-favor banking sector is the largest slice of this market, and a US economic recovery from the COVID Crisis remains slow.
The bottom line here: 60/40 is alive and well, but where you put the “40” matters a lot more now than when Milton Friedman wondered aloud if portfolio construction was relevant to Economics.
Markowitz Nobel Prize Speech: https://www.nobelprize.org/uploads/2018/06/markowitz-lecture.pdf