What is mean reversion?

October has been a funny old month for the markets, swinging from strength to weakness, to strength and back again.

For intraday traders, these kinds of swings are no bad thing, giving us plenty of opportunities to pick up profits with each move.

However, talking to traders, I get a distinct feeling that people are looking for safer options.

Sure, it’s well documented that the public are sick and tired of City bankers taking unnecessary risks, and there’s a clear move away from casino-style banking.

But I also get this feeling from ordinary traders I speak to – people who are tired of feeling like the market gives with one hand, only to take away again with the other.

Sometimes it helps to take a completely new perspective on trading – I’m talking about taking a walk around the other side of the mountain … where we might just find an easier route to the top …

What is mean reversion?

With the ongoing Mean Reversion Profit Strategy (MRP Strategy) trial in full swing, I’ve been asked a few times “What is mean reversion?”

So, I thought now was as good a time as any to look at what mean reversion is – and whether it can work as a trading strategy?

First off, I should explain that the way MRP uses mean reversion is very different from the “norm”.

But before I go into that, I’ll explain what common-garden mean reversion does …

Mean reversion: just like Mama used to make

Mean reversion trading strategies are based on the idea that price highs and lows are temporary glitches and that a price will tend to go back to its long-term average (or “mean”).

So, our mean-reversion trader will calculate the average price over time … figure out a high-low range …

Then, when the price has deviated to the low of that range, he’ll buy, or if it deviates to the high of that range, he’ll sell.

It sounds very simple.

And the premise is a sound one – most price movement is a reversion to some mean. And it’s exactly what everyone’s shouting about when they are predicting any kind of asset bubble.

The use of pivot points is a form of mean reversion – based on the principle that the price will move back towards the central point for the previous day’s trading.

But the problem comes with knowing which mean the price is reverting to, and when.

So, the first problem we run into is how to calculate our “mean”. How far back should we go? Do we allow price extremes to affect our mean? (A freak price spike could make our average price significantly different.)

And the next problem we have is time scale … When is our price going to revert to the average? Prices can trend strongly for days … weeks … years …

And this gets to the heart of why ordinary investors run into problems with old-fashioned mean reversion – it can take so long for a price to revert to its long-term average, that most of us will run out of capital while we wait!

Sure, big investment houses can afford to wait for it to happen – but for ordinary Joes like us – it’s just not a realistic way to trade.

Time for some long words …

Now, the form of mean reversion that the beta testers on MRP have been using – is very different. This, too, is used by big City investment houses, but there’s no reason why ordinary folk like us can’t tap into it too.

This crosses over with something called “quantitative statistical arbitrage” – and whilst I love using all those long words (because boy do they make me sound clever!), I have to admit that the “science bit” of this is way over my head.

The gist of statistical arbitrage is that we take a pair of stocks or other instruments that traditionally move together. Then we place trades based on a “mean reversion” of the price difference between the two – i.e. that the price difference will fluctuate around an “average”.

It’s considerably safer than just sitting there hoping that the price of some stock you bought “cheap” will eventually come back to its long-term average. However, it’s long been overlooked by ordinary folk as something that’s a bit too technical.

Fortunately, I don’t have to understand the maths to see that this way of trading can work (for major hedge funds, and for our beta testers – albeit on a slightly smaller scale!)

You can catch up on how the beta test is going here.

Related posts:

Trend following VS mean reversion (and how to combine them) How to use the Commodity Channel Index (CCI) indicator