Mean reversion provides tool for trading FX volatility

“A simple mean reversion model can provide effective signals for option strategies even when trading costs are included,” she says. In a study released on Monday, Commerzbank shows that prior to the financial crisis implied FX volatility tended to revert to a long-term centre of gravity that appears to be well approximated by the sample mean.

Commerzbank finds that the reversion period is best measured in weeks or months rather than days, and that it varies across currency pairs. The study is based on 16 currencies and considers ten G10 crosses and 6 crosses for emerging market currencies versus the US dollar.

The best tenor for investors to benefit from the phenomenon is three months, according to Commerzbank. James explains that is because short-dated options have limited sensitivity to changes in implied volatility, with prices more sensitive to moves in the spot rate. “Conversely, at the long end of the term structure option prices are driven to a far greater degree by changes in implied volatility than by changes in the underlying spot rate.” she says.

Commerzbank estimated the half life of moves in three-month implied volatility away from their average for each currency pair. That showed the average mean reversion half life was 72 days prior to the financial crisis.

The study shows a strategy that triggers a long volatility position when implied volatility moves 1.5 standard deviations below its average – and conversely triggers a short position when implied volatility moves 1.5 standard deviations above its average – produces a good track record in forecasting directional trends in implied volatility.

“Tests of a simple model suggest that mean reversion signals can successfully forecast directional changes in implied volatility developments and that in many cases these signals can be used to generate successful option strategies,” says James.

Volatility spikes

James says despite offering great potential, volatility mean reversion strategies are vulnerable to sharp increases in implied volatilities, such as that experienced after the collapse of Lehman Brothers in 2008.

In that case, the mean reversion period was extended as turmoil swept across financial markets.

“Very rapid rises in volatility may initially manifest themselves as strong moves away from the recent average that trigger short volatility positions,” says James

She says a primary concern when devising mechanisms to avoid that risk is that rapid recoveries from extreme levels should not be missed.

“If one loses out by being short implied volatility into a sharp rise of risk aversion, the last thing one wants is to be stopped out of the market for subsequent declines in volatility that could be exploited to recover some of the initial losses.”

James therefore imposed a short-term stop-loss signal on her model that prevents investment in a mean reversion strategy for two days should the cumulative rise in volatility over the two previous days exceed three percentage points.

The result of applying that filter for GBPUSD three-month implied volatility, as can be seen in the figure below, reduced the size of the drawdown witnessed in 2008, but did not interfere with the normal operation of the reversion strategy.

"Given that implied volatilities and option prices are directly related, a greater understanding of the future development of implied volatility is desirable for traders wishing to identify overpriced or underpriced options,” says James.

“Meanwhile for those taking a more systematic approach, option strategies can be implemented to gain from, for example, a predicted rise in volatility.”