Skip to content

Quick Take: Percent Exposure Donchian Channel Method

August 10, 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.

I have always been fascinated by Donchian Channels as they were the core method used by the legendary turtle traders to achieve their success.  Channels are simply bands containing the “n” day high bounded by the “n” day low . They can be seen visually on www.stockcharts.com  using their sharp charts by selecting the “price channel.”  I noticed that the larger the channel that the stock is currently in, the more likely the trend was to continue. In my own work on ranking stocks, i include a % distance from 52-wk high as part of the momentum factor which does in fact add a fair amount of value to the overall score. Many systems using channels to trade a wide variety of futures and indexes are still robust, even during the new mean reversion era. 

What is missing from all of the research that i have ever seen is a percent exposure method to catching trends. Wouldn’t it be smart to increase leverage and exposure as the move becomes confirmed? Most models are either all in or all out, that is why it is difficult to design short-term trend systems. The frequently incorrect signals given by shorter trend signals are too costly using a full bet size. More importantly, signals generated by short-term models  that are counter to the longer term trend are more likely to be incorrect or have a poor profit factor.

The question is how do we use this information? First, lets describe  how to trade channels: Exits and entries are determined by new channel penetration. That is, a new 50-day high for example would initiate a new long position after a previous 50-day low had been triggered without a 50-day high prior to that. Once a new entry is triggered, you would exit at a channel low that is half of the original channel length. That is, if you entered on a 20-day high, you would exit on a 10-day low, and so-on.

I think there are four concepts that should be stated from my observations regarding channels:

1) the 200 day channel position dictates the main trend: you will have to trace back on the chart to find the first penetration from long to short or vice versa to figure out what the main trend is.

2) smaller channels (a minimum of 20 days-5o days) can be used to trade within the main trend with less risk using a 2 ATR initial stop to minimize potential losses.

3) when trading counter to the prevailing trend, whatever the minimum channel length is in #2 you should use double that length as a trigger— so in this case to enter countertrend in bear markets you would use a 40-day channel, and the same applies to bull markets.

4) when trading counter to the prevailing trend….bet only half as much as when you are with the prevailing trend.

Ok, that is the basic framework. Lets figure out how we can scale in and out of a trend using channels. It makes sense to bet less on small channels, and more on larger channels. Note that when trading countertrend, you would bet half of these amounts:

1) 20% might be bet on a new breakout that is a 20-day high within the prevailing trend

2) 30% more might be bet on a new breakout that is a 50-day high within the prevailing trend

3) 50% more might be bet on a new breakout that is a 100-day high within the prevailing trend

This strategy provides us with a safe mechanism to add and subtract exposure, even when going counter to the main trend. I haven’t tested this out yet, as i have a long and large list. For anyone new to the blog…….this isn’t my core research area (designing stock rankings). But if you do test it out, let me know. I would be pleased to present your results.

dv

3 Comments leave one →
  1. August 11, 2009 8:51 am

    Good stuff. I’ve done some testing on adding leverage when trading with the trend (one could just add exposure rather than leverage and achieve the same result) and it worked well over the model I tested.

    Could you be more specific about calculating the percentages? When you say 20%, are you meaning 20% of available equity? What about a percent risk model where we start at 1% and ratchet up to 2 or even 3% based on the size of the channel and the prevailing trend?

    • david varadi permalink*
      August 11, 2009 9:54 am

      hi, the percentages refer to % of equity, a percent risk model could be combined with this by averaging the two values. Or to be very conservative, and consistent with the logic in the article you would multiply the % risk model by the %exposure. I would base the % risk based on volatility using 2 ATRs as an initial stop. Its something worth testing using multiple scenarios. Good suggestion.

      dv

Trackbacks

  1. Percentile Channels: A New Twist On a Trend-Following Favorite | CSSA

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: