A study in diversification

Diversification can be a funny thing. It’s one of those things that sound so boring and mainstream that many people simply ignore them. Just something boring that the establishment wants us to do but really has no actual use. Kind of like taxes and seat belts.

But diversification can actually be of use. It’s especially important for trend following strategies where the return distribution over time for any individual market can have very fat tails.

I’m going to do a quick demo to show just how important the diversification can be. Let’s use the same simple 12 months momentum model that we used twice before. I put together a diverse universe of round 75 instruments, fairly evenly split on the five broad sectors. These sectors being agricultural commodities, equities, currencies, non-agricultural commodities and rates.

This model, as you might remember, only looks at whether the current price is higher or lower than a year ago. It goes long if the current price is higher than a year ago and short if it’s lower. That simple. We ignore everything in between. The demo here would work just as well with any other kind of trend following model as well.

Position sizes are equal risk sized, using a simple ATR based approach. Below I’ll show the results first for running this model on all markets at the same time, like most would do in reality. The model will trade all 75 markets across all asset classes all the time. The risk allocation per market is set to 10 basis points, using the classic ATR formula [(0.001 * AccountValue) / (ATR50 * PointValue * fxRate )].

Here are the results for the whole set.

• Annualized Return: 25%
• Max Drawdown: 27%

Now what would happen if we run this same simulation on each sector, one at a time? To get a fair comparison we’ll have to crank up the position sizes to match the same portfolio risk. The alternative would be to leave sizes the same, but since we have much fewer instruments we’d naturally end up with very low numbers. If we want to compare these investment universes as alternative strategies, they should be given the same portfolio risk allocation. We have five sectors with about equal amount of markets, so I’ll just round things off a bit and set position sizes to 50 points, five times the risk per market.

Here are the results.

• Agricultural –  Annualized Return: 14.5%, Max Drawdown: -66.2%
• Equities –  Annualized Return: 22.5%, Max Drawdown: -65.2%
• Currencies –  Annualized Return: 12.4%, Max Drawdown: -46.8%
• Non-Agricultural –  Annualized Return: 21.5%, Max Drawdown: -56.2%
• Rates –  Annualized Return: 26.9%, Max Drawdown: -57.6%

None of these look very appealing. They all show good profits, but at far too high risk.

Take a look at the chart below, showing the performance for each individual sector when traded by itself, as well as a fat black line showing the performance for the diversified portfolio.

Now you might point out that the rates sector had higher performance by itself, but would you really want to have that almost 60% drawdown from 2003, with six years recovery time?

The agriculture had an incredible run from a bottom in early 2006 and the returns two years after that were really spectacular. But then they had a sharp pullback again.

The point is that the markets is like a box of chocolate. You never know what you’re gonna get. This is why you need to follow a diverse set of markets.

For smaller portfolios, sub million dollar, it can be difficult to trade as many markets as I used here. Difficult being a bit of an understatement of course. If you’re outside of the US, you can try CFDs, but they’re banned stateside.

That opens for an interesting question: What’s the minimum number of markets you would need to get decent diversification in a trend following model? Well, let me get back to you that one another time…

 

On the topic of multiple markets… Did I mention that I’m now making my internal weekly futures research report available for public subscription? This is a buy-side research report, with none of the fluff that’s the hallmark of sell side letters. I do this report at my asset management shop to make sure everyone’s always aware of what’s going on across all futures markets that we trade. If you trade futures, perhaps this can help you and give you an insight into how futures hedge funds approach the global markets.

Sign up for a free month trial here.