The first sign of active HFT involvement is that most of the low-hanging fruit would have been plucked. i.e., text-book arbitrage opportunities would not exist. In the futures market, the low-hanging fruit is the cash-futures basis. Equity futures have a ‘fair-value’:
Futures Price = Cash Price [1+r (x/360)] – Dividends; where x = days to expiration of the futures contract
So whenever the basis goes out of hand, its possible to pocket some risk-free profit.
- if Futures << Cash, go long futures, short cash. Or,
- if Futures >> Cash, go short futures, long cash.
Given the absurdity involved in shorting stocks, option 1) is not feasible for individual investors. But if HFT arbitrageurs are efficient, then scenario 2) should not exist. So I pulled up some charts and I have annotated days where HFT machines were probably in sleep mode. The filled candlesticks are the cash and the black boxes are the futures candlesticks. Ideally, the futures black boxes should coincide with the cash candlesticks.
Since August this year, there has been at least 5 instances where a risk-free profit could have been locked in by shorting futures and going long cash – if you had a low-latency trading machine.
I pulled up the charts of some of the other punter counters:
So what exactly is going on here? Is the cost of funding so high that the HFT guys let this one slide? Or is the liquidity so bad that this can’t be done in size, and hence not attractive to HFT? Will the real arbitrageur please stand up?