Many years ago, when I was in graduate school, I remember carrying on a loud argument with someone in the college cafeteria, which ended when I said, “I don’t believe in empirical evidence!” Later that evening, as I thought about the conversation, I realised that the research I was doing towards my PhD was all about collecting, collating and analysing empirical evidence.
This conflict was certainly one of the seeds that led me to drop my research project, but as I moved on to other things — and, I might add, as I graduated from academics to reality — I learned that, whatever you feel about it philosophically, you can’t hide from empirical evidence; you use huge amounts of it virtually every moment of your life.
And, of course, you can — sometimes — use empirical evidence to make money from markets.
Back in March 2009, I had noticed that the volatility of copper had risen substantially (8 per cent) above that of nickel. This was clearly an anomaly — the volatility of a material is an intrinsic property and on a long-term average basis, nickel volatility was nearly 15 per cent higher than that of copper. Again, every time the spread had turned positive, it reversed within a few months. Clearly, it was a good empirical bet to sell the spread, and, sure enough, in two months, it had returned to 16 per cent, which, with only modest leverage (5X), provided a wonderful annualised return of 85 per cent.
Unfortunately, that was a very exotic trade, extremely difficult to execute since the LME does not trade volatility independently (of options). I explored different ideas, but the huge volatility of both underlying assets made the basis risk much too high. So, it was just a good idea, but no cigar.
Today, however, there is another volatility-spread anomaly, which is much easier to express and could also result in some nice gains.
Since the middle of May this year, the volatility of the Sensex has been unusually holding a few per cent below that of the Dow. Now, emerging economy equity markets, like India, are usually more volatile — and sometimes much more volatile — than developed economy equity markets. Indeed, since the start of liberalisation of the Indian economy (in 1991), the volatility of the Sensex has been higher than that of the Dow for as many as 83 per cent of total trading days.
In fact, the spread (between Sensex and Dow volatility) has been negative only twice before. The first time it went negative was in July 2002, and it stayed negative for over a year (till August 2003). This was likely because the Indian market was in the earliest stage of development, when, amongst other limitations, foreign investment into equities was severely constrained. In other words, it hadn’t really become a market yet — indeed, the four-week average of gross FII inflows crossed $500 million for the first time ever only in September 2003. As it turns out, if you had bought the Sensex (or the Dow) on each day that the spread was negative and held it for a year, you would have had an average return of 55 per cent (17 per cent)!
The second time the spread became negative was during the recent global crisis. Sensex volumes fell below Dow volumes in November 2008 and the spread stayed negative till March 2009. The reason for this deviation from the norm was quite obvious — with the epicentre of the crisis in the US, it was hardly surprising that the volatility of the Dow rose much more sharply than that of other markets (including India). Again, if you had bought either market on each day when the spread was negative and held for a year, your return would have been 83 per cent and 29 per cent, respectively. Incidentally, the default return since July 2001 — where you bought the index every day and held for a year — would have been 26 per cent and 2 per cent, respectively.
Now, here we are again — the volatility spread is negative, and, indeed, has reached its lowest level since 2004. So, does this mean that both markets are going to rally by substantially more than the default return of 26 per cent and 2 per cent, respectively, over the next 12 months?
Well, empirical evidence — admittedly a little thin — does point in that direction, even though it is hard to understand what the structural reason for this anomaly is. However, common sense suggests that there may be more causality for the Sensex than for the Dow.
So, depending on your belief level and your trading capabilities, you could a) go long Nifty futures; b) go long Nifty VIX and short US VIX, this would, of course, require an off-shore play; c) buy out-of-the-money Nifty calls.
And, send me a cheque in August 2011.
