Thursday, October 7, 2010

Risk Mangement

Let us say that a stock has met all your fundamental and technical criteria and has made it to your watch list. You stalk the stock for weeks waiting for the markets to give you the go ahead and for the stock to form one of your setups. Then finally the moment you've been waiting for arrives, all the stars align etc... . You know from your experience, research and/or back testing that this scenario has a high probability of doing what you expect it to and so now you are anxious to enter trade.

You identify the entry price and stop loss and calculate your position size. Since the stock is just about perfect and in the perfect group too and will be breaking into all time highs if it triggers your entry and the markets moving higher then you assume that the risk/reward is excellent. Not so fast!

What I have been doing unconsciously since I first started traded (picked it up from a Dr. Alexander Elder book I believe) was to use the average width of the preceding 10 to 20 days price action to determine the risk/reward of the trade. What I did in my head when trying to rank trades, that appeared to have about the same odds of success, by risk/reward was to compare the difference between my entry and stop loss to the average width of the past 20 days or so price bars. Then if the difference was just about equal to or less than the average width then I would take the trade and further I would give preference to trades which had a difference that was significantly less than the average width. Since the trades have the same odds of success over a series of trades with a difference that is significantly less will have a higher expectancy since expectancy = "odds x average risk/reward".

On my good trades, I generally manage to capture at least a 3 - 5 day swing and in the best trade a 2 - 3 week trend. So the logic behind the above concept is that if I only take trades where my risk is less than the average width of a price bar and I get a 3 price bar move in my favor (which as I outlined above is what I generally get on my good trades) then automatically I have already made 3x my risk on the trade! Even further, let us suppose that the risk was about half the average width of the price bars over the last 20 days (this would be the case with entry off a narrow range candle stick like a spinning top) then a 3 bar move in my favor is a profit roughly 6x my risk on the trade!

So what happens to me often is that I tend to pass on the initial set up in some stocks but take ones later down because of better risk/reward or pass on a high quality set up in one stock and take a slightly lower quality set up in another. So although I would catch a smaller part of the move in the first instance or even a smaller move in the instance of the lower quality set up assuming the same exits in all scenarios, the preferred trades would actually have a greater impact on my account than the ones passed on because I normalize my trades by units of portfolio risk or "R" as Dr. Van Tharp refers to them.

This concepts helps to improve your returns by pushing your portfolio into trades with a higher expectancy for people who use the percent risk model to determine position size and manage risk. So over a series of say 100 trades this could make a dramatic difference to your absolute returns.

Hope you find this concept useful.

Good luck and good trading!

No comments:

Post a Comment