This is the third in a series of articles. In the article from June 2005, we examined the roles of liquidity and volatility and how these two parameters affect our selection of stocks and markets to trade. In the article from October 2005, we looked at the three cornerstones of successful trading. This issue focuses on the "management" aspect of trading.

The Key to Survival: Position Size
Risk management. Money management. Trade size. Position size. Optimal f. Kelly Criterion. Collectively, these phrases determine a trader's ultimate fate. Once we know what we want to buy or sell and have a method for determining when to do it, the only thing left to figure out is how much. Without a reasonable estimate of the appropriate position size, traders experience disproportionate profits and losses; that is, you lose too much when a trade goes against you while failing to make as much as you should when the trade goes your way. Add this to the less than random win / loss ratio experienced by the majority of traders, and we have a real problem here, Houston!  

If a method loses more times than it wins, the few winners must make more dollars than all of the losers and then some, or else the trader is destined to become a statistic. Nearly everyone has a visceral understanding of this concept, but in reality, few implement valid strategies to avoid what is affectionately known as the risk of ruin.

The Risk of Ruin
In trading, there is a well-defined line between aggressive and insane risk-taking. For a thorough explanation, I highly recommend Fortune's Formula by William Poundstone. By the time you arrive at page 232 and see the graph that crystallizes the concept, you will understand why, regardless of skill, traders who ignore the question of "how much" are virtually guaranteed to leave the game broke by virtue of overbetting.

Risk per Trade
Traders use many approaches to calculate position size. To me, an effective method must take into account a number of factors.

Let's go over the math.If I have $100,000 in my trading account, a one percent risk is $1,000.What this means is that, with any given trade, I am willing to lose $1,000 if the trade does not go my way.To many novice traders, these figures look conservative, but there is a good reason for this. What happens to the account if a trader makes 10 losing trades in a row? Let's take a look.

Starting balance: $100,000

Traders might be discouraged to find that it does not take long for even a $100,000 account to be depleted. While investing can be done on a shoestring because the power of compounding can accumulate a fortune over time, trading is a business, and it's just a fact that businesses require capital to operate. Each of us must decide how much we are willing to risk. The more we risk on each trade, the faster the account balance declines during a losing streak and, of course, the higher the risk of ruin.

The Stop Loss
Once we have decided what percentage to risk on each trade, we can calculate the number of shares to buy or sell. This requires knowing the stop loss location. In other words, the trader needs to decide the point at which he or she will admit defeat if the trade does not go his or her way. 

Stop losses are tricky. Most traders who use them err on the side of being too tight. They place their stops too close together. Many others don't use them at all and leave themselves unprotected. Others use fixed dollar stops that do not take volatility into account. The key to profitable trading is a well-placed stop loss, and much time should be spent on investigating this subject.

Where exactly should we give up on a trade? In theory, I believe the optimal point to leave a trade is when we can see the next entry. For example, if the market is in an uptrend, there comes a point where a pullback becomes big enough to warrant selling the position in case a major trend reversal is at hand. 

By leaving at this inflection point, we avoid a potential major trend reversal. If the uptrend shows us the money by reasserting itself after a period of weakness, we can re-enter the trade with a setup designed to buy pullbacks. This way, we always trade from a position of financial -- and mental -- strength.

An Example

I

This is a weekly chart of Apple Computer. My trading system colors the price bars blue for buy and pink for sell on the upper panel. There are also small blue and pink dots above and below the price bars that indicate stop loss points. On the lower pane, my position sizing tool does two things.

On this chart, I have applied the tool to automatically size positions on the buy side. When the system is waiting for a buy signal, it provides me with an estimate of the position size I should take if the system issues a buy signal.

In July 2003, the system bought 261 shares of AAPL upon a close above $10.336. My $1,000 risk is divided by the difference between the close and the stop loss below to arrive at the position size.

As the price of AAPL goes up, the position sizing tool takes into account the unrealized profit, adds this to the original account size and calculates the number of shares that should be held to keep risk at 1 percent. By the end of January 2005, the system is holding 159 shares. The position sizing tool tells us when to sell to keep risk as a percent on a constant basis.

In April 2005, the trading system issues a sell signal on close below $38.635. By that time, the system has gradually sold shares along the way, and what are left are 147 shares to sell on the signal.

APPL was weak in May and June 2005, and in July, it resumed its uptrend. On the close above $41.277, the system issued a buy signal. Assuming a $100,000 account with 1 percent risk, the position sizing tool tells us that we should buy 90 shares because AAPL is much higher than it was before, in terms of price and / or volatility. Again, as time goes by, the tool recalculates the position size, so we can sell off shares to maintain a constant 1 percent risk.

Sanity Check
You can see that position sizing is a function of many variables, and each of those must be thought out carefully in terms of concept and application.  Even though the result is scientific, we must also ensure that the actual market can accomodate the calculated position size. For example, if a stock price is low and its volatility is low, the calculated position size may be quite large. We need to perform a sanity check against the actual volume and number of trades to make sure that we can easily get in and out of the position as necessary. 

In our example, we know that millions of APPL shares are traded per week over the course of tens of thousands of transactions. It would take a split second to enter and exit the calculated position size for our risk, so we can rest easy that there is enough liquidity to transact our trades in a timely manner without too much slippage.

The position sizing tool will be added, in the near future, to the PowerTools Add-On Studies for eSignal subscribers upon release of eSignal 8.0 beta.