One of the hardest tasks in education now is to encourage students to develop critical thinking skills when information sources have fragmented, and debate and argument seem to have degenerated into simply restating positions with increasing vehemence. When confronted with a statement, we push them to ask: “how likely is this to be true?” and then, self-critically, to reflect “why do I think that?” The second question is really key: do I think it is true because it fits with my beliefs or because I’ve heard other people say it? Or is there a clear and rigorous evidence base supporting it? That is of still greater importance as increasingly we live within echo chambers that feed and nurture our prior conceptions. If you are thinking “well, I don’t” remember that the commercial property industry is a classic echo chamber encouraging group think.
It was in this light that I read the piece “New Year, Oldest Asset” by the Undercover Investor on the Property Chronicle. It’s definitely worth a read. In it, the author states that gold, as an investment “has out-performed the S&P 500 since the turn of the century,” that the commodity had “come out on top of markets as a whole” and made a strong case for its properties as an insurance in times of uncertainty. True to the principles I teach, my initial reaction was “really?” and then I checked myself. For much of my professional and academic life, I’ve had to confront embedded beliefs, one of which is, indeed, that gold is a great asset in portfolios. Given that many of those other beliefs are not firmly grounded in fact, the nerd in me demanded that I went to find out.
First things first: if one accepts the clearly false premise that the new century started in 2000 (there’s no year zero: the first ten years end at the end of year ten, the first hundred years end at the end of year 100, the second thousand years finish at the end of year 2000) then between 2000 and 2023 the dollar market price of gold on the London market increased at a compound rate of 8.8% while the S&P500 (allowing for reinvestment of dividends) grew at 8.2%. Not much in it, but it’s a win. Gold prices were also slightly less volatile over the same period, so that looks like a better risk-adjusted return. As this is the Property Chronicle, we should note that private commercial real estate – here measured by NCREIF – delivered 8.1% total return. The raw index suggests that property returns were less volatile than gold, but it is an appraisal index and so subject to valuation smoothing: applying a standard de-smoothing adjustment suggests similar volatility to gold. Not much to call, then.
It is, though, more complex, more nuanced. We do need an investment vehicle here. It is straightforward to invest passively in an S&P500 tracker fund, and transaction and management costs are very low. It is much harder to invest in commercial real estate (and avoid tracking error). What about gold? If we are going to hold the physical item in some form, then we need to think about carry costs, which will reduce our return. If we are going to use, say, an ETF or a gold-linked derivative, we will have greater volatility and tracking error from contango effects. Maybe it is not quite so clear-cut.
Furthermore, parceling up time is, in essence, an arbitrary partitioning of a continuum. What’s so special about the turn of the century (particularly when it was celebrated a year early)? So I took successive ten-year periods, starting in 1973 (the data go back longer, but aren’t helpful as there was still dollar to gold convertibility until the Nixon era). The S&P generated higher ten-year returns than gold in 63% of those ten-year periods. If we reduce the holding period to five years, the equity market wins more than two thirds of the time. The story is similar for real estate, albeit for a slightly shorter time span: gold only beats NCREIF around a third of the time for a ten-year holding period and only a little over a quarter of the time for a five-year investment horizon. In long time horizon, over the last fifty years, the S&P500 generated total returns close to double those of gold, with no real volatility penalty. We should also note that both the S&P and NCREIF generate income in the form of dividends and net rent: while we are accounting for these in the returns, many investors will need that cashflow.
What of the insurance story? Gold definitely seems to generate better (annual) returns in times of economic weakness (although the equity market tends to recover rapidly from these). More valuably, change in the gold price looks to be negatively correlated with the equity market – consistent with the asset switching story the Undercover Investor tells. This diversification benefit seems to hold over different time periods, too. Thus gold looks like it would genuinely diversify annual risk within the portfolio. So too does real estate (even correcting for the appraisal smoothing) but not to the same extent. Given that the return story is perhaps a little less compelling, then how valuable is dampening the short-term volatility at portfolio level?
