Last week something interesting happened to gold options. The short trading week saw the return of strong investor interest to the gold market, with net long positions on Comex increasing by 4.5 moz. This constitutes the most significant weekly inflow since the 6.4 moz rise during the first week of September, which at that time helped gold break the $1000/oz barrier. Yet, unlike in September, 1 month contracts initially did not experience a pick-up in volatility that usually accompanies decisive shifts in the spot market. In order to understand what this means, it is worth stepping back to consider some unique characteristics of gold’s historical – and implied volatility.
In theory, gold should not be volatile at all. Inventories and storage capacities are a non-issue for this “commodity”. Unlike other commodities, gold volatilities cannot be easily dissected into structural (long-dated prices) and cyclical (forward curve structure) components, as the metal is not mean-reverting and its global inventory is always plentiful. Unlike platinum, or oil, gold is geologically abundant in almost all major regions of the world. And, unlike silver, gold is usually produced and accounted for predominantly as the main output of mining operation. Consequently, it does not suffer from low elasticity of supply.
Secondly, for the purposes of the investment and jewelry markets as we know them today, the supply from mines, scrap and, if need be, central banks is sufficient to satisfy demand. Finally, gold storage is not a big deal. Ultimately, people carry this stuff on their bodies. This is probably why gold is notorious for being the least volatile of all commodities, bar livestock. Indeed, long-term comparisons show that commodities difficult to store – such as power – are the most volatile. The long-term annualized daily return volatility of power rates (say, US PJM West Hub) is over ten times gold’s.
But there is another unique characteristic of gold’s volatility. During bouts of strong momentum buying, as during the last quarter of 2009, gold’s volatility surges. And while, at that time, implied volatilities of the VIX index continued to fall in the broader market – almost in synch with the bullish sentiment in equities, gold’s volatility increased in line with spot price gains. It eventually peaked some two weeks after the metal notched its all-time high level in dollar terms.
According to World Gold Council, over the last 20 years, standard deviation of positive returns for gold is above 3% (compared to 2.3% for S&P500). Typically, in the broader equity market, negative returns are associated with higher volatility, yet gold’s standard deviation for negative returns is actually lower, 2.5%.
How does volatility correlate with gold prices? The last three years provide an interesting lesson. During the year separating the “subprime” fallout in Sep 2007 and the Lehman debacle in Sep 2008, the correlation was positive and high (0.8). It dropped precipitously, as gold prices reacted negatively to the potentially deflationary shock of the banking collapse. Gold volatilities shot up, in synch with other asset classes, and the correlation between gold prices and gold volatilities reverted to negative. The relation remained negative until late summer 2009, but since October, the correlation has turned positive again, although less decisively so than in the preceding period (currently at around 0.15). As a result of new inflows into the market (both Comex and ETFs), over the last week we have observed the lowest ever ratio of volatilities to gold prices. In other words, during the recent quarters it has paid off to be long gold, but short gold volatilities. Unfortunately, in order to have more confidence in this trade going forward, it would be helpful to have a slightly higher implied volatility than it is the case now (old adage goes that shorting volatility is best when the implied volatilities are significant, but historical volatilities are unimpressive).
Gold options, including options on gold equity products, such as ETFs, differ in volatility skew from the rest of the market. Implied volatility of calls and puts has a forward skew (volatility is higher for higher strike options). This is not only the case of gold, but also of most other commodities, many of which have experienced much stronger swings in long-dated prices than the yellow metal. The most liquid SPX options, on the other hand, have a reverse skew (higher implied volatility at lower strikes). Not surprisingly, the options on gold mining companies, and on GDX index, have the volatility skew more akin to the broader equity market.
In the near term, if gold continues to notch up regardless of USD/EUR direction, I would expect volatilities to pick up and the record price/vol ratio ease in the short term. If not, we are in uncharted territory. But shouldn’t we get used to it by now?
Tags: Comex gold, correlation between prices and volatility, gold ETF options, gold price volatility, volatility skew
Leave a reply