A familiar face at STA monthly meetings, on Tuesday 12th September (the first chance members had to catch up after the summer holiday) it was Trevor Neil’s turn to stand at the podium. Bravely admitting to 40 years’ experience trading the markets he showed us a photo of his younger self in classic 1970’s garb, complete with kipper tie; this week’s dapper outfit of classic navy suit, smart tie and classic black lace-ups was a distinct improvement.
Trevor was first introduced to the Vix pop (as in burst, not Dad or corks) by Phillip Kahler in October 2015. A man whose professional profile lists algo trading, consulting and sales as his speciality the system is surprisingly neat, simple, and easy to programme. It tries to capitalise on contrarian trading, buying stocks when fear is at its highest.
This was the mantra of the first Baron Rothschild: ‘Buy when there’s blood on the street, even if it’s your own’; Warren Buffet too is a fan: ‘Be greedy when others are fearful’. Trevor installed it on his PC, which is called Wilma (wife of Fred Flintstone), to back-test this automated trading system. Making notes of his thinking when buy signals were generated (this is a long-only system), he has discovered that it’s better to do as Wilma says rather than add layers of complexity and doubt to the decisions. ‘Over optimising is setting yourself up to fail’.
He strongly believes that ways of measuring risk are poor and under estimate the pitfalls. That there is an over reliance on historical data and that the volatility of volatility itself is high – but does mean revert. First plot a daily chart of the value of Vix, the CBOE measure of 30 day implied volatility over a selection of strike prices. Observe roughly at what level spikes have peaked in the past.
To this add standard Bollinger bands (using a simple 20-day moving average and 2 standard deviations) for both daily and weekly data on this same chart. Watch for daily price moves above both of the upper bands and for days when the daily Bollinger trades above the weekly one. Either of these generate a buy signal at the following day’s open so that ‘getting a bad execution is good as the market is roaring higher. You know if you’re wrong if the market doesn’t flick up quickly. This is a quick response system’. Then use a 3 per cent trailing stop – which can later be reduced to 1 per cent. One could also reduce the size of the standard deviation to improve results.
Caveat: the system generates very few trades.