05/01/2015 11:41 am ET Updated Jun 29, 2015

Aggressive HFT and Volatility

Co-authored with Steve Krawciw

Aggressive HFT has long been associated with volatility. Several academic studies hypothesized that higher aggressive HFT participation leads to higher volatility levels, and for good reasons, as explained below. The empirical evidence, however, has been hard to come by, until now. This note explains the empirical relationships between implied volatility of options on stocks comprising the S&P 500 and participation of aggressive HFT, as measured by the AbleMarkets Aggressive HFT Index. As the note shows, two prominent conclusions can be made about the aggressive HFTs in relation to volatility in which the aggressive HFTs are present:
1. Higher aggressive HFT leads to higher implied volatility one full day ahead in a stock comprising the S&P 500
2. Cross-sectionally, aggressive HFT has a persistently opposite effect: when aggressive HFT is noted in stock A to be higher than in stock B, implied volatility of options written on stock A is lower than the implied volatility of stocks written on stock B.
Both results are robust for the inclusion of other independent variables, including the intraday range volatility of stocks, the change in the intraday volatility, price levels and returns.

On average, and in the absence of all additional explanatory variables, aggressive HFT alone can explain as much as 30%-40% of implied volatility variation on a T+1 basis within a time series for any one security. These successful predictions have significant economic consequences: an ability to accurately predict future implied volatility and a priori enter into profitable trading and risk management positions. The T+1 predictive power of the aggressive HFT, however, can vary significantly from one security to the next. Accurate predictions of implied vol significantly improve numerous vol trading, hedging and even collateral pricing applications. In collateral pricing, a debt can be viewed as an option on equity and priced as such. High-accuracy predictions of implied vol result in highly accurate collateral pricing results.

If one takes a snapshot of implied volatilities across the components of the S&P500 at any given time, however, and ranks the 500 stocks by their AbleMarkets aggressive HFT index recorded on the previous trading day, the following dynamic emerges: stocks with higher aggressive HFT participation tend to have lower implied volatilities in a peer-to-peer comparison without taking any additional history into account. This phenomenon allows traders to successfully build volatility trading strategies in a statistical arbitrage framework by the buying (selling) straddles on equities with a sudden increase (decrease) in aggressive HFT activity, as measured by the AbleMarkets Aggressive HFT Index.

Interestingly enough, AbleMarkets Aggressive HFT Index is largely independent of most common metrics predictive of T+1 volatility, such as the volatility recorded on day T. Specifically, the statistical explanatory power of the AbleMarkets Aggressive HFT Index does not change much regardless of what additional variables are included in the prediction of the next day's implied vol. Therefore, the AbleMarkets Aggressive HFT Index can be used as an independent additional source of predictability and can be combined many other factors.

In summary, financial markets professionals can no longer afford to ignore aggressive HFT participation. AbleMarkets Aggressive HFT index is a unique product that tracks aggressive HFT participation based on proprietary technology developed and continuously refined over the past eight years. The AbleMarkets Aggressive HFT Index is a stable well-performing mechanism that measures aggressive HFT across most electronically traded financial instruments, including equities, futures, foreign exchange and others.

Steve Krawciw [kro:sew] is CEO of Irene Aldridge is Managing Director of Able Alpha Trading, LTD., and and author of High-Frequency Trading: A Practical Guide to Algorithmic Strategies and Trading Systems (2nd edition, Wiley, 2013).