15 January, 2015

Engineering a Synthetic Volatility Index – Part 1

I employ volatility analysis extensively in the hedge fund I manage both at the individual stock level and at the index level. Both types of analysis are essential if you hope to trade successfully, so in today's blog I'm going to discuss each of them. Index volatility will however form the focus of our discussion: I'll share a highly effective technique that I've developed to create a synthetic volatility index for any of the global market indices. This is a very useful tool because it allows one to quickly and simply glean an understanding of broader market volatility in any global market regardless of whether the respective market has an associated volatility index. We'll kick-off today's discussion with single stock volatility.

Understanding volatility inherent in individual stocks helps one to define one's expectations of the severity of future price swings, which in turn enables one to implement prudent risk management. For example, one may decide to allocate smaller portions of one's capital to highly volatile stocks, while maintaining higher capital allocations to stock's with low volatility. One could apply different rule sets to stock's that exhibit high or low volatility. The possibilities are endless. The point is: having an understanding of single stock volatility is of outmost importance for effective risk management, particularly in short-term trading.

But there's another approach to analysing volatility which is no less important: index volatility analysis. Index volatility provides one with a view of broader market risk. High levels of index volatility are typically associated with uncertain market conditions that present additional risk to trader's portfolios. Therefore, index volatility can be used as an input into high level risk-on/off decision making. In fact, this type of analysis is an integral part of my risk management strategy in the fund I manage. In the USA I study the VIX, but many market indices don't have an associated volatility index. So how do we analyse index volatility in those instances? We can't, so we'll have to build our own custom algorithm. The good news is that I've done all the hard work for you. I've developed a simple and highly effective method to build a synthetic volatility index for any market index. But before we move on, let's first examine the VIX.

The VIX (CBOE Volatility Index) is a well-known volatility index which represents the market's expectation of 30-day volatility in the S&P 500. It's a widely used measure of market risk and is often referred to as the "investor fear gauge". The calculation of the index is quite complex, involving implied volatilities from a range of S&P 500 index options. Due to this, constructing a similar measure of volatility for any other market index is near impossible, especially for emerging indices with non-existent options markets. This led me to develop and quantify a custom approach that 1) when applied to the S&P 500 mirrors the VIX index as closely as possible and 2) relies solely on price as an input so that it can be applied to any market index.

After some elbow grease I came up with a solution that when applied to the S&P 500 has a correlation of +0.92 with the VIX. In other words, it's almost perfectly correlated with the VIX and could therefore be effectively employed in favour of the VIX to measure volatility in the S&P 500. But more importantly, we can apply the indicator to any market index.

Next week I'll share the logic of my Synthetic Volatility Index with you and provide the Metastock code so that you can use it too.