Skip to content

Quick Take: Long-Term Historical Relative Strength by Market (RSM)

August 20, 2009

Note: Quick Takes will be short articles introducing new ideas or concepts that haven’t been tested. It is designed to help inspire your own personal systems. If you wish to demonstrate the results of Quick Takes ideas, your results will be published on this site subject to verification.

 The first priority in system design is to find a strategy that works consistently over time. The second and often overlooked priority in my opinion should be in market selection. Now this is a topic that is very broad and covers such factors as liquidity and contract sizes as well as how well the personality of the specific market matches your own personality. While these subjects are certainly very important, what i am referring to for the purposes of this article is to select the market(s) that is(are) likely to be the most profitable for your system. For example, using the Donchian Channel strategy mentioned in the previous post, after analyzing the profitability by market I noticed that some markets are consistently more profitable than others. Notable examples are currencies and petro-commodities (oil, heating oil and gas). For mean-reversion based strategies using indicators like the DV2, the best markets include the  S&P500 and the real estate sector index (IYR).

The million dollar question is: ” how do we identify these markets beforehand?” Well the answer is to simply create a rolling historical analysis of the profitability across different market and incorporate that into your strategy. What I am talking about is essentially taking a look at the relative strength of the equity curves for various markets and adjusting with proper frequency. How do we do this? Here are two initial rules:

1) If you are using a short-term strategy such as the DV2, RSI2, or a short-term moving average, than you should take a look at a 1-3 year  historical relative strength by market using a risk-adjusted measure such as the DVR or the Sharpe Ratio. Every 2-3 months you should re-evaluate the historical relative strength by market. Select the top 2 or 3 markets based on their performance using the strategy.

2) If you are using a long-term strategy such as Donchian Channels, RSI 14, or long-term moving averages, than you should take a look at a 10-15 year  historical relative strength by market using a risk-adjusted measure such as the DVR or the Sharpe Ratio. Every year you should re-evaluate the historical relative strength by market. Select the top 2 or 3 markets based on their performance using the strategy

Based on my observations, certain markets seem to persist in exhibiting trending or mean-reverting behaviour. As an example, over 10 years ago when i first looked at trend-following strategies, the very same markets were the best performers at that time  (across a wide range of strategies) were virtually the same in the recent 10 year period. In our more detailed work in the stock market, certain stocks and sectors demonstrate remarkable persistence using momentum-based strategies. For almost 35 years, the technology sector has responded spectacularly well to these types of strategies. Fast-forward to 2009, and the tech sector is the best perfoming major index in this recent market rally. While markets of course do change, this is independent of using a historical relative strength strategy. The first priority in designing your systems is to create a robust model that also adapts to what is working in the current environment. For example, knowing that the current environment favors mean reversion strategies we would select the DV2 as our favorite indicator. Given that you have done this, now you would look at which markets it has worked the best on, and you would naturally proceed to the S&P500, the Japan index (EWJ) or the Real Estate sector (IYR). If you wanted to trade medium or longer term moving averages-which still work– you would definitely choose the NASDAQ, the Russell 2000, and the commodity index (GSG or DBC).

For the diligent readers out there, if you wish to test this in a structured way, i would be pleased to present your results.

3 Comments leave one →
  1. August 20, 2009 8:46 pm

    Hi David,

    You said, “For example, knowing that the current environment favors mean reversion strategies we would select the DV2 as our favorite indicator.” Do you have an indicator that tells you what the current environment favors? Or is that a matter of looking at different strategies and seeing which one is working better?

    Thanks.

    • david varadi permalink*
      August 20, 2009 9:01 pm

      hi, i actually will be discussing the method that i use to determine the best strategy or best parameter to use in the near future. needless to say, i use a statistical method that adapts constantly to changing market conditions. The lookback and required confidence in a strategy also varies depending on the type of strategy or number of parameters etc. Perhpas the quote about selecting the DV2 was a bit of a misnomer, since i was just giving an example of choosing a system that works well that a trader is familiar with. I wasn’t suggesting that this is the best indicator–although i have yet to test a better short-term indicator than the DVO (of which the DV2 is a subset designed for the SPY).

  2. Rod permalink
    January 15, 2010 4:36 am

    Great post. Very helpful. There has been a similar discussion on Tradingblox.com in the past. The agreement there was that you can’t predict, so dont filter your universe other than for liquidity and obvious factors such as size. It is interesting that you have found persistence in behavior.

Leave a Reply

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

WordPress.com Logo

You are commenting using your WordPress.com 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

%d bloggers like this: