Using Volatility in Trading
The market is relatively quiet and taking a pause, or so it seems, so I thought I would leave the market alone and talk about a general concept which can be useful. One thing that isn’t well understood in the pursuit of profit is volatility, i.e. by how much a price moves, not solely whether it moves up or down, which is what most traders focus on.
One of the very few ways it is possible to make profitable predictions regarding whether the price is likely to go up or down is by observing whether the price has been going up or down in its recent history, i.e. whether there is a trend. This works, to some extent, and has been well documented. Can the same thing be achieved with volatility?
The simple answer is yes. The pioneering work of Benoit Mandelbrot identified a prevalent phenomenon in financial markets: volatility clustering. What this means is that when volatility is relatively low, it is more likely than not to remain low the next day, and the same follows then volatility is relatively high. Of course, traders have understood this phenomenon instinctively for decades before it was proven scientifically. Traders have tended to exploit this by watching for a sudden explosion of volatility in a quiet financial instrument – traditionally, a breakout – and getting in on the move and hanging on until the volatility really starts to die down again.
You might ask whether this phenomenon can be observed in Forex. Let’s conducts a test of two major currency pairs over the past 15 years: the EUR/USD and the GBP/USD, and firstly measure the autocorrelation of the daily price movements. A result of 1 would be a perfect positive correlation, with -1 being a perfectly negative correlation. The results were:
This suggests that whether the price moves up or down today can tell you almost nothing, by itself, about the direction it will move tomorrow. However, if we use absolute values and therefore simply measure whether the size of today’s price movement tells us anything about the probable size of tomorrow’s movement – whether up or down – we get some markedly different correlation coefficients for the two currency pairs using the same data and same period:
This data suggests that while the effect might be relatively small, it exists. How can it be used practically to help with trading?
When you see a sudden price movement that is relatively very large, such as we saw in GBP/USD yesterday, keep in mind that another large movement is more likely to happen today. This means that you should widen your “normal” stops considerably. It also means that if you are long, you should not panic at a strong pullback, and if you are short, you should be going for more than just a few pips.
Don’t forget that the “real” commission you pay by way of the spread goes down as volatility goes up.