Building a market neutral strategy

Last week I released a post about the groundwork for a market neutral strategy. Since then I continued my research. With this post I want to share with you some of the interesting highlights and inspire you to kick-off your own research.

As I wrote at the end of the last post: my main focus will be on improving the overall % size per trade. My goal is to have at least 0.75% per trade, before commission and slippage.

So let me share with you some of the ideas I have and how those materialized:

Improve ranking

In last weeks research I looked at how stocks have been mean reverting using an intermediate to longer-term indicator, e.g. RSI(14) or RSI(10) or ROC(252). My assumption to test: the mean-reversion effect for stocks is stronger when they are oversold on a shorter as well as intermediate term time-frame.  So Let’s add a second dimension to the ranking algo. I created a combination ranking of RSI(10) and DV2 . DV2 is an indicator from David Varadi. DV2 is looking at the past two day move of High, Low, Close in relation to the past 252 days. Hence it’s a significantly better RSI(2).

The test is done on a survivorship free SP500 index (link) . Data not adjusted for cash dividends, no commission / slippage considered.

Absolute returns have significantly increased showing the effectiveness of the addition. Unfortunately return per trades remain about the same, still to low for me.

Prolong the trade

Average trade size remains an issue. An idea to further improve the strategy is to prolong the time in the trade. In above’s test a trade is entered when the stock is among the top 5 (short) or bottom 5 (long) according to the combination ranking.  So Let’s extend that rule to

- Enter Long: when among bottom 5 stocks, Exit Long: when NOT within bottom 50 anymore.

- Enter Short: when among top 5 stocks, Exit Long: when NOT within top 50 anymore.

As you can see the Avg. length per trade has changed from 2.4 to 3.2 days and so has the avg. return per trade from 0.32% to 0.49%.

Further improvements

Now you have to do your own homework. Here is some more food for thought:

  • All tests have been done with an equal position sizing approach: 10% of equity per stock. To trade the strategy in a true market neutral approach one would need to adjust the position size according to a stocks volatility.
  • Furthermore you might want to think about stocks that mean revert better than others.

I’m going to continue my research.

- Frank

About these ads


  1. I’ve always thought that a really good metric that would maximize the utilization of capital would have to take the time factor into direct account. I’m thinking that it would be better to optimize for avg. return per bar, when allocating capital by percentage of equity, or optimize for avg. units of profit per bar as measured by ATR, when allocating based on volatility. I use the latter in my system metrics. This kind of formula requires that there are more than ample trading options for that capital such that it should be in play in some trade as much as possible. In other words, the assumption is that I should take less profit but get out more quickly because there’s another fresh opportunity with just as good a historical advantage just sitting there that will go to waste if I continue to tie up my capital in the first trade trying to get another 20% of the move by doubling my holding time. So the trades would tend to be of shorter duration, but earn more per bar. Jump in, grab the meat of the profit in 2 or 3 bars, get out, lather, rinse, repeat.

    In my research, what I’ve found that this approach runs up against is that the shorter your trade duration (or likewise the shorter the bar interval), the more you are playing admidst the market’s noise, and the harder to duplicate the historical findings in real-time. But my assumption is that there is a “sweet spot” to be found through research that is as optimal as one can get in a constantly-changing sea of noise.

    Thanks for your posts in this series! They have inspired me to do more research in this area.


    • Hi Michael,

      I agree with you thoughts and observation. However one needs to keep in mind the type of strategy you are comparing with / against. I’ve seen lower (historical) returns per bar for trend-following compared to shorter-term swing / mean-reversion strategies. However, I don’t want to miss this element in my mix of trading strategies.


      • I agree! Strategy and time frame diversification is so very important today! It’s the #1 thing I’m working on tweaking in my overall approach right now. Too much correlation between instruments to overcome at a symbol selection level.

  2. One idea would be to lagg exit by a day if the trade is in the same direction as the long term trend. In an uptrend you would hold a long position an extra day to get more juice out of the trade.



Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s


Get every new post delivered to your Inbox.

Join 304 other followers

%d bloggers like this: