EXPECTATION
How do you know if your system is going to make money in the long run? Does your system have an edge? Are you managing risk to avoid ruin? If you keep your system simple and robust to work with any market in your portfolio selection, you need to test it to check for not only profitability but the probability or odds of long term success. As you test, realize that you can test each system component separately however also know that they may be interdependent.
Expectation is basically estimating your reward for the risk you are taking to make sure the reward is greater than the risk. For basic expectation, you calculate it from these system results through backtesting and ongoing as you go live with your system:
- Winning percentage of your total trades. What percent were winners and what percent were losses?
- Size of profits and size of losses. Of your winning trades, what was the average winning amount? Of your losing trades, what was the average loss amount?
With those numbers, you calculate the expectation by multiplying the winning percentage by the average amount of the winning trades. You also multiply the losing percentage by the average amount of the losing trades. Then you subtract the losing number from the winning number.
For three examples, we will first use a system that has a 50% win rate with the average win of $250 and average loss of $150. For the second set of numbers we will use a system that has a 30% win rate with the average win of $1250 and average loss of $250. Lastly the third set of numbers will use a system that has a 60% win rate with the average win of $50 and average loss of $45.
Example 1: ( 50% * $250 ) - ( 50% * $150 ) = Expectation of $50 made per trade
Example 2: ( 30% * $1250 ) - ( 70% * $250 ) = Expectation of $200 made per trade
Example 3: ( 60% * $50 ) - ( 40% * $45 ) = Expectation of $12 made per trade
Even though the winning percentage of example two is much lower, the expectation is higher as the winners are much greater than the losers. Other things being equal, example 2 would make more money in the long run. When evaluating a system or backtesting a system, find the expectation or probability of winning to estimate how successful the system will be.
Bruce Babcock's Guide to Trading Systems and Curtis Faith's Way of the Turtle go through multiple perspectives of building and testing a Trading System.
COSTS
In addition to the raw per trade profits and losses, also account for other transaction costs. Brokerage commissions can have a significant impact on your profits, especially in a short term or high frequency system. While backtesting gives a good idea of results, you will most likely experience slippage in live trading. An example of slippage is when placing a buy order at $1, the order is filled at $1.02. Backtesting would not account for the extra 2 cents and even though our example is 2 cents, it can make a big difference when buying a large number of leveraged futures contracts.
OPPORTUNITY
Lastly for expectation, take into account the frequency of your trading.
Your system may have a small winning amount but if it trades a lot, it will still make you money in the long run.
We will change the examples from above to show how opportunity can change the profitability of the systems when the expectation is annualized.
If example one gets about 50 trades per year, the yearly expectation is $2500
If example two gets about 12 trades per year, the yearly expectation is $2400
If example three gets about 200 trades per year, the yearly expectation is $2400
Even though all three have very different starting expectations and winning percentages, they will end up making about the same now that we have added in yearly opportunity.
Do you have the discipline to follow your rules? Can you make it through the drawdowns without abandoning your system? The next page is about the psychology it takes to stay with a winning Trend Following system.