A recent article published by Bloomberg seem to give credit to the oft-repeated notion that gold (GLD) (PHYS) is an essential part of everybody's portfolio. Specifically it seems that gold helps a portfolio to outperform in the long run by helping to stabilise losses during tail-risk events that negatively impact assets values.
There are many reasons to allocate a portion of a portfolio in gold, including the constant fiat currency debasement by all governments around the world. In this article we will concentrate on gold’s utility as a protection against market crashes or adverse geopolitical crises.
In August 2017, Ray Dalio of Bridgewater Associates LP, the biggest hedge fund in the world, recommended investors hold 5 to 10% of their portfolios in gold to hedge against political risks. According to Dalio (interview and transcript here):
Gold serves a purpose. It is first of all, a diversifier against other assets. You know, we have this risk on, risk off thing. We also have a monetary system. The Bretton Woods monetary system began after World War II, and it had the dollar as the world's reserve currency. There's a risk there. There's a lot of dollar denominated debt and so on. If somebody felt they didn't want to hold that, and so you could have exposures to that.
So it's a diversifying asset that is sensible, and that's the main reason to have gold in the portfolio, five to 10%.
Cameron Crise, a macro strategist who writes Bloomberg’s Macro Man column, conducted interesting research on this important topic: "Is gold really an effective hedge in periods of risk"?
In his research, Crise made a regression to search for evidence of a statistical relationship between risk aversion and the gold price. The CBOE Volatility Index, the VIX (VIXY), was used as a proxy for market risk aversion. Crise ran a series of multifactor regressions to determine whether or not equity volatility is statistically significant as an explanatory variable for gold.
Below you can see a multifactor regression that shows that the VIX, while not the most important driver, was still highly significant. Data used in this regression covers the period from 1990 to 2015. The regression included as independent variables inflation, real US yields, Central Banks purchases, a basket of industrial metals, the VIX, the US dollar and the ISM manufacturing index (see table below).
Results of the regression (Bloomberg).
All variables were significant, except the ISM Manufacturing data. The VIX, using the t-statistic as gauge of the importance of the explanatory variables, while not the most important explanatory variable, was significant. The t-statistic is the ratio of the departure of the estimated value of a parameter from its hypothesised value to its standard error.
Counter-intuitively, gold’s relationship with inflation over the period has been sharply negative, suggesting that the metal actually dropped when inflation increased. By using this regression, it appears that gold's reputation as an inflation hedge is somewhat exaggerated. Results would likely be rather different if a longer period of time (and especially the 1970s) were included in the analysis. As expected, gold has an inverse relationship with the US dollar price (using the DXY Currency basket) and is very sensitive to real interest rates (as opposed to nominal interest rates). Below you can see a chart plotting real interest rates and the gold price since 1971.
Negative Interest Rates and Gold price (Incrementum AG, In Gold We Trust 2017).
According to the results of the regression, the gold price is also highly correlated to Central Bank purchases (or sales) of gold in the open market. This is not a surprise as these sales and purchases are often announced well in advance and can sometimes represent turning points as in the case of the infamous Brown's Bottom, the sale of UK gold reserves pursued by HM Treasury over the period between 1999 and 2002, when gold prices were at their lowest in 20 years. The last takeaway from this regression is that there is a positive correlation between the price of gold and the price of industrial metals.
In order to dig deeper and assess gold's importance in providing stability in a portfolio, Crise identified 10 notable episodes of risk aversion over the past three decades and, in order to include the 1987 stock market crash and the Iraq-Kuwait war, extended the observation period to 30 years. The duration of each crisis episode was defined using the peak-to-valley move in the S&P 500 index around each event.
Below you can see the performance of US stocks, Treasuries, and gold during these episodes. Gold performed, as expected, as a risk-aversion hedge.
S&P 500, US Treasuries and Gold during 10 crisis episodes in the last 30 years (Bloomberg).
As the definition used the market peak-to-valley move in the S&P 500 as an indicator of a crisis, the equity market performance was obviously poor during these crises, with an average loss of 19.6% per episode and negative in each instance. Treasuries returned an average of 3.4% and performed positively on seven occasions. Gold, meanwhile, rose by 6.9% on average per episode, with gains in 8 of the 10 periods.
It is important to underline that this methodology doesn’t capture the price action that occurs after an equity trough, so it may not be the full picture. It is however interesting to see that in 8 out of 10 cases, gold actually went up when the stock market crashed.
Next, and this is the most important takeaway of this article, we want to know if a portfolio with 10% gold would outperform a typical balanced portfolio. What we want to know is if and how including gold in a portfolio over long periods of time could change the risk and returns of a portfolio.
In order to do this, Crise constructed two sample portfolios: a 60/40 asset mix calculated using the S&P 500 total return index and the Bloomberg Barclays US Treasury Total Return Index and another portfolio with an asset mix of 55/35/10. This second portfolio reduced the stock and bond weightings by 5% apiece and integrated 10% of gold.
In the chart below you can appreciate how the 55/35/10 portfolio outperforms the 60/40 portfolio by about 55 basis points per year even if this comes with a slightly higher volatility, as risk-adjusted returns were virtually identical for both portfolios.
The two portfolios' performance (Bloomberg).
The superior performance of a portfolio including gold brings us to conclude that the notion of gold as a hedge against serious risk aversion episodes is true. The statistics appear to show a relationship between volatility episodes and gold gains. Indeed, the anecdotal historical evidence supports the hypothesis that gold offers solid protection from asset market dislocations in recent times.
I hence recommend allocating 5-10% of your portfolio in gold.
Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.
Additional disclosure: I own physical gold.