Every trader also wish to strike it big ("hoot tua tua") when come to winning. But the reality is that as we wedge too big a position, we will also faced with the proportionally high loss when the trade does not turn out to in our favour.

To further complicate the problem, market is full of uncertainty. It does not mean that if you lost 1 trade, your next trade will turn out to be profitable. It does not mean that if you lost 10 trades, your next trade will turn out to to be profitable. Your position sizing strategy has so make you survive these bad times.

The rational of position sizing is always to control risk, preserve your capital, yet try to grow it as much/fast as possible. These are all conflicting demand and it is always a balance, depending on individuals.

There are a few major methods to size your trading position:


1. Percentage of Portfolio Risked by Volatility

In this method, the volatility of the instrument is determined - using standard deviation or Average True Range (ATR), or moving average of ATR as seen in Turtle trading system - and the position size is calculated based on the predetermined percentage potential loss to the overall portfolio and the volatility (or multiples of ATR as in the case of turtle trading system).

Using this method you will end up with different number of lots in various trades but have the same volatility risk for each trade.


2. Percentage of Portfolio Risk by Fix Dollar Amount

As above, except that now instead of volatility, a fixed dollar amount is used (say S$500), and the same calculation is done. Now if you got stopped out, you know that you loss $500 for that trade.

Using this method, you will still end up with different number of lots in various trades. The draw back is that you are risking the same dollar amount for different reward. You see, volatility work both ways, it can trigger your stop, it can also move the price in your favour.


3. Percentage of Portfolio

This is a simple way, just divide your overall portfolio into a number of equal amount, then use that amount to buy a trading position. You will end up with equal position size each a fraction of the overall Portfolio size.

If I recall correctly, there is a term called "Optimal f" related to this. Optimal f is the faction that when one position is traded over a period of time using the same system, this particular fraction will give optimal (max) growth to the portfolio size. Off hand I don't have details as how it is derived, but I think it is related to the win/loss ratios of your trading system methodology.


Van Tharp in his book "Trade Your Way to Financial Freedom" did some stimulation and come up with the conclusion that the first 2 methods will growth your portfolio (much) faster. Please go and read that book before saying they are the best for you. There are other factors such as the expectancy of your trading system / methodology and draw down to consider.

There are yet modifications, combinations or variations to all the above method. If you look at the turtle trading rule, it reduce the norminal portfolio (size that use for calculation of position size) by 20% with the actual portfolio was drawn down by 10%. The point here is not to exhaust all variations, but open up some doors for you to explore. You have to take into consideration of your own circumstances and find one that suit you. For example, the volatility method is not so easy to calculate if you don't have your data at hand.

Even if you wish to dump the whole portfolio into just 1 stock every time, if it suit you and you can devise other part of trading to take advantage of this method, so be it. But this is very risky tactic you better know what you are doing.

Just a side note: Many online trading game participants tends to pump all their virtual money into 1 or very few counters (hope to) to achieved good result. This method will make the portfolio value fluctuate greatly, both way.

It is your money, you decide.

How much faster? I was asked.

Based on Van Tharp stimulation in the above book on Dow Jones stocks using a $1 mil portfolio:

1. "Fixed number of share per position" model made $32,567 over 5.5 years, or 0.58% compounded annually, max drawdown 0.75%

2. "Fixed dollar-amount model per position" model made $360,000 approx (figure given in the book was wrong, I gather this from the equity curve chart) over 5.5 years, or 5.75% compounded annually, max drawdown 7.13%

3. "Fixed percentage of portfolio per position" model made $232,121 over 5.5 years, or 3.86% compounded annually, max drawdown 3.72%

4. "Fixed percentage risked to portfolio per position" model made $1,840,493 over 5.5 years, or 20.92% compounded annually, max drawdown 14.14%

5. Volatility based model made $2,109,266 over 5.5 years, or 22.93% compounded annually, max drawdown 16.61%

Of course all these are just hypothetical testings, so don't just take the number as it is as the market, market situation, your trading model's expectancy etc all affect the results. But the point here is to highlight the (potential huge) effect to the final trading profit when you change your position sizing strategy.