Measuring Trading Performance (Part 2)
In my last article Measuring Trading Performance – Part 1, I talked about the importance of stop losses in our trading. This was also demonstrated using my risk management software – TradeRisk – where both Bob and Jane were exposed to the same trade. In this example Jane, with the use of stop losses, was only risking a fraction of her capital whereas Bob who was ‘betting on the stock only going up’ did not use trailing stop losses and was fully exposed. This time I’d like to talk about why being right is NOT the same as being profitable. It amazes me how many people are swayed by trading systems that claim +80% of profitable trades and place little consideration on what happens to the other 20% of the losing trades. One could be fooled into thinking that since there are less losing trades than winning ones, overall the portfolio must be a winner. Nothing could be further from the truth! OK, so how do we measure the performance of a trading system to ensure that overall we end up winners when trading it? Well there is this thing called expectation or expectancy. This is a standard term within statistics meaning ‘probability’ or ‘anticipation’; however the first time I have seen it in the context of trading was in a book titled Trade Your Way to Financial Freedom by Dr Van Tharp. The purpose of expectancy is to tell us how our portfolio has performed in the past compared to our initial risks taken. It is important to compare trading performance to the risks involved, instead of just looking at the overall return. This will give a fairer assessment on the trading system. Ask yourself: would you rather trade a system that yields 80% return at 40% risk or a system that yields 50% return at 20% risk? As the reward/risk profile of the second system is higher (5/2 compared to 8/4), I’d trade that system over the first anytime. Even if I didn’t mind risking 40% - and I do – I can always leverage the second system to risk 40% and I’d gain 100%.-
Getting back onto the concept of portfolio expectancy, it is calculated as follows:
- Say you have gone long on a trade at $10.00 with an initial stop of $8.00. Here, you are risking $2.00 per share (ie the difference between the entry price and the initial stop loss);
- If you have purchased 1,000 of these units (this could be Forex, stocks, CFDs etc.), you know you are risking $2,000;
- Let's call the initial risk 'R'
- Now, if you:
- Have made $2,000 on this trade your trade would be a 1R winner;
- Have made $4,000 on this trade it would be a 2R winner
- If you have lost $2,000 - ie your trade stopped at the initial stop loss value of $8.00 your trade would be a -1R winner (or a 1R loser), etc.
-
A few things ought to be evident from the above scenarios:
- None of your trades should be more than -1R winners - ie you should never ever exit the trade below (for long trades) the initial stop. Of course there is always the slippage and transaction fees, however you can also account for them when defining how much you want to risk;
- The higher 'R' winners you have, the more profitable your portfolio will become
- HAVING MORE WINNING TRADES THAN LOSING ONES DOES NOT GUARANTEE YOU WILL HAVE A WINNING PORTFOLIO.
-
For instance, if you have five trades with the following R distribution:
- -1R, -1R, -1R, -1R, 5R - where you have only 20% of winning trades or
- 1R, 1R, 1R, 1R, -5R where you have 80% of winning trades
Where: ΣR is the sum of all R values within the portfolio. So for example (a) you have an expectancy of
whereas forexample (b) your expectancy is -0.2 – that is you are trading a losing system despite of have 80% of your trades as winners. In TradeRisk you can find your R values on the far right column for each of your trades as shown below:
Your expectancy is shown on the portfolio summary page as depicted below:
The purpose of the expectancy formula is to tell you how much money you are making on the overall portfolio per dollar risked. With this formula you can evaluate any trading system and compare them to each other on the same basis. You no longer have to evaluate systems by comparing how accurate they are. As demonstrated above, that is totally irrelevant, just compare the systems’ expectancy and the one with the higher number will have you making more money per dollar risked.
So when somebody tells you their system is +80% ‘accurate’ and it is selling now for a ‘bargain’ $9,999.99, ask for its expectancy. If they cannot tell you or do not even know what that is, forget about it. Besides, ask yourself, if you had the goose with the golden egg, would you really want to sell it at any price?
Most long term trend following system are less than 50% accurate and good ones have an expectancy ranging between 0.3 – 0.4. This means that for every dollar you risk you will get the dollar back plus another 30 - 40 cents.
To give you an indication, my trading system after some 1,000 trades over three years is ‘only’ 45% accurate. Whilst that is not much to write home about, my winners are almost three times as big as my losers helping me achieve an expectancy of $0.35 at the same time with an ROI of over 70% p.a. for the same period. For a more detailed view on my trading performance, please visit www.banksiaprice.com.
Good luck with your trading.Arpad Marton – HiQTraders.com Read more articles from the Stock Investor Journal. Get data to assist you in managing your risk.