Forex Trader Mentor

Newsletter  # 13

1st March 2006

It seems a short time ago that I sent the last newsletter out, but it is actually almost 2 months ago! I have been very busy mentoring, so time has flown. To all those who responded to the survey - thanks. I am trying to analyse the interesting responses, but could not finish in time for this newsletter.

I have decided to dedicate this newsletter to describing my Win Lose Probability simulation, which is a piece of kit born of necessity. I hope to bring the next newsletter out shortly, where you will be able to read part 2 of the psychology of trading and hopefully, get the results of the survey.

Win Lose Probability Calculator/Simulator

I am a practical sort of guy. If I have a mathematical problem, instead of using some fancy algorithm, I build a practical model in Excel. Maybe there are neater ways of doing things, but experience shows me that my method works.

I had a problem that had to be solved. Successful consistent trading means to me that you have to be able to work out your probability of getting the desired result over an extended series of trades - say, over the course of a year.

To me, the key parameters are the win/lose ratio and the Risk Return Ratio (RRR). Together, they give a mathematical expectancy, naturally enough a positive mathematical expectancy for profitable trading.

When you are trading, you should be trading with an 'edge'; that is, a method or system that produce more winners than losers over an extended number of trades. It would be nice to believe that there are systems and methods that produce 60%, 70% or even 90% winners to losers. There are systems peddlers who claim just this. And people buy into it. Only, nothing is mentioned about the Risk Reward ratio. If it means that you are going to take 3 pips profit for a 30 point stop, you can still lose money. How do I know? I built a model.

It should be clear, that the higher the RRR the better for a given Win/Loss ratio. The problem is, like odds at the bookmakers, the higher the RRR, the less probable that the trades will achieve that ratio. For example, you would get many more trades hitting a 2:1 ratio (say, 30 pip profit for a 15 pip stop) than a 7:1 ratio (105 pips profit for a 15 pip stop). Conversely, anything less than 2:1 needs a high win/loss ratio.

Interestingly, to get a reasonable return for a fairly low risk, your 'edge' can be less than 50%. That's right, you can have only 1 in 3 trades make money and still be ahead.

The third factor is how much you are willing to allocate on each trade. What sort of stop are you going to set? If you had a win/loss ratio of 50% and a 1:1 RRR and risked 20% capital on each trade, you would not get very far. Try it: Toss a coin where you get 1 point for a heads and 1 for a tail and see how many tosses it takes a) before you get ahead and b) before you get wiped out. So we have to find an optimum amount of risk. If we only risked a very small amount per trade, we would not risk much - neither would we gain much!

What I wanted to know, was what is the 'edge' I need, the optimum risk/reward ratio and how much capital per trade I need to allocate to get a given result over a large numbers of trades. So I built a simulation, where using random number generation, could plug in the suitable parameters of Win/Loss ratio, RRR and capital allocated, run the simulation over a number of times and come out with a result that I could say was going to be 'normal' for most of the time.

What do I mean by 'normal'? I mean that by applying a statistical method, called a standard deviation, I can create a band which statistically would mean that 68% of the results from the simulated portfolios lay between.

This model was constructed in Excel and what I enter as the base data are my results to date. I apply the same capital to each trade and use the same system and rule set. This is very important. My performance to date can by no means be expected to continue in the future, but the more trades I have, the more reliable the data. I just extrapolate over a multitude of similar portfolios.

What am I looking for?

First of all, my observations have shown me that an 'index' number of .8 or higher is required to have a positive result. This is the RRR times the win/loss probability. For example, RRR 2 (:1) and win/loss 40% = 2 x .4 =.8.

Secondly, I want to see what range of profit I can expect. This is final profit, the maximum attained profit (the two are not the same!) and the maximum annualized return.

Thirdly, I want to know how much of my capital I will be putting at risk and what I can expect as a maximum drawdown. Basically, it is the approximate number of consecutive losers. By asserting this number, I can quickly see if a series of losers is in line with the prediction, or I really do have a bum system.

This now gives the basis for setting the parameters for my budgets and gives me a good idea of the risk and reward numbers I have to set in the TradeCraft system to start trading.

A little knowledge is a dangerous thing.

This is where I should be saying 'I have found it - the Holy Grail of Trading' - just get this system and I will guarantee you millions in profits!'. Only that probability is just that - probability. No one can foresee the future. If we could, there would be no market, because surely the market is the expression of people's expectation on where the rate should be in the future?

You can have a system that gives a PME (positive mathematical expectancy) of 1.5, meaning that you will earn 50 cents for every dollar risked. Only that market conditions may change and your system may no longer be valid. In addition, you may have some very bad luck and just be a victim of a fluke where everything goes wrong at the same time ('Murphy's Law').

So, the tool has to be used as a guide to the probability of certain events unfolding.

It does address one of the fundamental problems faced by novice traders, namely: how good is my system? What does it mean if I have 4 losing trades in a row? Should I ditch the system? With this tool, it is possible to monitor the performance of your system and extrapolate the results subject to the caveat mentioned.

The other major caveat is that you do the same thing exactly each time. That is, you use the same signal, the same amount of capital, the same rules of entry and exit and the same rules of Risk/Reward. Only then can you attempt to predict a result over a sustained period. But hey! Isn't that what we are trying to achieve? Consistent profits?

I hope that you have got the point that trading is a compromise between having a system that delivers a given number of winners to losers, the amplitude of the profit and the amount of capital risked. You cannot have a system that produces profits 75% of the time at a RRR of 7:1 and risking 10% of your capital per trade. It is just not going to happen! You would be making 13,000% per annum - each and every year - if you didn't get wiped out in year 1!

On the other hand, going with a system that gives profits 40% of the time for a RRR of 1.5:1 risking .5% of your capital is going to produce a measly income in the best case, a loss in the worst case.

So, you will have to find something in the middle- something profitable, low-risk and sustainable.

I am not making this tool available on the site. This is because I believe that there are lessons to be learned about risk management, trading systems and rules before being in possession of such a powerful tool. So if you want to get the advantages, I suggest that you take some mentoring so that you can fit it into your style of trading

Examples

See what happens if you risk too much per trade? Traders often inadvertently do this by trading with too high leverage. The 'high' drawdown is close to 100% and the minimal annualized return average is -163%. You would be risking too much to achieve the potential 450%-540% profit.

The same thing, but risking 1% capital per trade. The highest possible drawdown is 11%, with a potential average reward of 50-60%). Note that you

have to test your system under live conditions (not back testing) to see if it will produce the win/loss ratio and the RRR. This requires doing some 20 or more trades to get a sufficient 'population', and is an ongoing process.

Only small changes are required to dramatically change the profitability. There is a probability that the return will be some 157% over an annual period. The overall capital risk is some 4%. Not a bad ratio!.

Conclusion

Finding a system is only part of the story. When the peddlers say their system is right 70% of the time, they are withholding important information - namely, what is the Risk Reward Ratio and what is the capital risk per trade. The next part of the story is to apply the system according to YOUR rules. The third part is to analyse the results and create a business plan from these results BUT constantly monitor your results to see if their are any changes in the whole set up that would lead to a deterioration in profitability over a longer period.

This tool is not going to bring you instant success - you still have to have a trading system that gives a certain number of winners and at the same time, shows you enough reward for your risk. What it does do is take the guesswork out of how good (or bad) your system/rules/money management is and will be in the future.

If you are going to trade forex and put your precious capital at risk, it is worth knowing your chances of success!