Saturday, February 28, 2015

Calmar ratio

As regular readers of the blog will know, I plot a number of performance metrics to keep me on track. These can provide clues as to the robustness of the system or alter me to a possible  change or deterioration in performance. These are shown here.

One metric I use to keep track of performance is the Calmar Ratio. This is a quick and dirty metric, measuring returns against drawdown risk and is calculated as follows:

Annualised annual return of last three years/maximum drawdown. 

As the stats recorded on this blog are for the period slightly less than three full years, the formula has been adjusted as required.

For some background information, here is an old article I came across written by Terry Young, who developed the Calmar ratio. Peter Brandt also gives his opinion on Calmar here.

As of this week's close, I have calculated this metric based on both closed trades only, as well as all trades:
 

Some points to note here:

The maximum drawdown incurred in this period was 30.67%. This was the period from June 2014 to early January 2015 where I encountered a long run of losing trades, which are shown in all their glory here. A drawdown of this magnitude will be too much for some people. Trend following methods can suffer periods of non-performance, but if the approach is based on sound principles, these periods have to be endured from time to time. For a comparison, if you look here, you will be able to look at some of the most successful trend followers and fund managers performance, with their maximum drawdowns also shown. In some of these cases, those drawdowns have been bigger than 50%!

It is a lot easier to ride out a drawdown when you have some past performance to rely on. Someone who went into such a drawdown immediately upon starting to trade a new method will have a completely different opinion.

I use fixed fractional position sizing, risking 2% of cash equity per trade - again, some people will say that is overly aggressive. As equity grows, this is an area which may need to be reviewed because of potential position sizing issues. If that happens, then both returns and drawdowns will reduce, but their relationship will remain in balance.

I have a win rate of 1 in 3, but my wins are much bigger than my losses, with a profit factor just over 5, so that gives me the positive expectancy. Some people seek comfort in a high win rate, but a win rate of between 30% - 40% is typical of a trend following approach.

This type of Calmar reading can only be achieved if you are very diligent in cutting your losses as quickly as possible, and letting your profitable trades run as far as possible.

No comments:

Post a comment