Gold prices have continued to be very weak and very volatile. Since gold stocks comprise a significant weight in the TSX, this has implications for portfolio construction, as well as risk management. Not only does this volatility impact the gold equities themselves, but risk measures of non-gold equities are also impacted materially. When a sizable portion of the market is experiencing a spike in volatility, risk measures can be unstable. In this article, we will investigate the changing risk of gold equities throughout 2013, the implications for non-gold equities, and some alternate methods for measuring risk.
In 2013, the rising ex-post beta of the TSX gold sector reached the highest level seen in the past 10 years. With beta nearly 3.5 times more volatile than the S&P/TSX Composite, it has clearly been one of the most volatile segments of the Canadian market. Not only is this extremely high, note the wide range over the past 10 years. The instability of gold beta adds an additional challenge for portfolio managers. In some periods, the TSX gold sector’s beta was well below 1 and actually turned negative for a brief time in 2009. In other periods, such as 2013, the gold stocks have been 2 to 3 times more volatile than the S&P/TSX Composite.
Rolling 3-month Beta (TSX Gold vs TSX) – click to view full image
In reality, we view multiple rolling time periods (i.e. one month, three months, six months, one year) in order to put the volatility in context. Portfolios with significant over- or under-weights in the gold sector will see their portfolio beta measures impacted significantly. With gold betas currently high, portfolios with zero weight in gold stocks will have portfolio betas biased below 1. Portfolios with a large overweight in gold stocks will have portfolio betas biased well above 1.
Portfolio risk measures must be adjusted when a sizable subset of the Canadian equity market becomes significantly more risky than average. Magnifying this challenge is the inherent volatility of the gold sector’s beta. As seen in the chart above, the gold sector’s beta could swing from well above 1 to below 1, if history is any guide. Therefore, any adjustments to risk can quickly become unacceptable, and distorting, if the subset in question returns to normal after reaching elevated risk levels.
What to do?
We can make adjustments to better understand a portfolio’s true risk characteristics excluding the gold phenomenon. One method is to measure a portfolio’s beta excluding the gold sector. Specifically, the gold equities are stripped out of both the portfolio and the benchmark in order to determine the portfolio’s non-gold beta. The results appear below for a fictitious portfolio. Imagine a portfolio that is underweight gold throughout the past two years. As the TSX gold beta rose from 0.5 to 3.5, notice the growing gap between the portfolio’s ex-post beta and the beta ex-gold.
Source: Greystone for illustrative purposes
It is also interesting to note how betas in another commodity sector, such as Energy, tracked through 2013. In fact, the trend was the opposite as the TSX energy sector beta fell significantly to well below 1. West Texas Intermediate (WTI) oil prices have historically been quite volatile at times, but have been fairly steady at around $90-$100/barrel in the past year. This has led to less volatility than expected in the energy sector (surprisingly less volatile than the TSX).
Watch for part two of this article in the near future, where we compare the volatility of gold and energy equities to their underlying commodities.