I need a better explanation for why the gold "spice" should be expected to improve portfolio performance. Without an explanation that will apply in the future, the historical data between these apples and oranges could be waved off as a coincidence.
Do you, really? I think Markowitz did a good job of explaining that. It doesn't improve performance, so much as efficiently reduce risk, which sometimes is more important.
The null hypothesis is that gold just happened to rise during a couple of key bear markets in the past without a relationship to stocks. Historical data alone are insufficient to reject this null hypothesis because the number of relevant bear markets is in the single digits. We need a mechanical understanding of how this theory is supposed to work before we say it'll happen again.
The first gold ETF was introduced in November 2004, and holding physical gold was
illegal prior to 1977. I don't think we can analyze gold as an investment prior to 1977 because it wasn't an investment.
That leaves us with 1980, 1987, 2000, 2007, and 2020 as our five relevant data points. So how did gold do those five times? From the ERE blog post which supported a gold allocation we obtain:
During the brief 2020 pandemic bear market, GLD rose about 27%.
Gold Performance During Bear Markets1980: -50.1%
1987: +7.3%
2000: +12%
2007: +18.5%
2020: +27%
--------------
avg: +2.94%
The interesting thing is that the one inflationary crisis (1980) was when gold as a hedge failed catastrophically, and the next four crises, when gold did well, were disinflationary. So much for the theory of gold as an inflation hedge or stock market hedge! It's something else.
Here's my attempt at a mechanical explanation: Could gold be less of a hedge against stocks going down, and more of a hedge against interest rates going down? In theory, holding gold incurs the opportunity cost of the interest one could obtain in a similarly volatile bond. As interest rates decrease and the opportunity cost goes down, gold becomes more attractive. As rates rise and the opportunity cost increases (as in 1980) gold becomes expensive to hold and people flee from it.
Yet gold has not consistently outperformed bonds during correction years. The following data are obtained from
NYU's historical returns dataset:
Baa Corporate Bond Performance During Bear Markets1980: -3.32%
1987: +2.81%
2000: +9.39%
2007: +4.84%
2020: +10.6%
------------------
avg: +4.86%
Based on this take, Baa corporate bonds were better to own during stock correction years than gold was. This finding is completely due to the 1980 wipeout, which co-occurred alongside Volker's aggressive rate hikes.
So maybe gold is a good hedge against another disinflationary / deflationary recession when rates are cut, but it will wipe you out in the event of another outbreak of inflation and rising rates? That's less of a hedge than a commodities / interest rate bet that could go either way. Will the next recession be triggered by inflation and rising rates, or will it be a more common disinflationary recession with falling rates? If you bet incorrectly, you'll only increase your stock losses.
Why accept any ding to your long-term portfolio performance to hold something that functions as a hedge in some bear market circumstances, and a money-loser in others? Why not just use a
costless collar or
protective put to keep returns within acceptable bounds? These are actual hedges tied to stock prices, instead of a commodity that might rise or fall depending upon the specifics of a future bear market.
Throwing more asset classes at the problem is not a solution. Instead we need to plan for inflationary or disinflationary bear markets and understand how assets will work in either outcome. Hedges need to be direct and supported by theory, not just what worked the last time or few times.
