Maximum intraday drawdown

This is the maximum reduction in your capital made over a number of trades from the highest capital (original plus profit) reached at any time. The measurement should be made by utilizing the extremes of the bars during the positions, rather than measuring from close to close. It is a reflection of the pain that you will suffer as you see your capital balance reduce over a series of losses, and this includes a spike lower during one day, even if the price rallies to the close.

The figure shown in the above system report is S$25,708.13. This represents 25.7% of our original capital of S$100,000. This is rather on the high side of what is acceptable. Do remember, however, that the system results are based on each trade being of the same size, in this case S$10G,000. If the string of losses that caused the cumulative loss of S$25,708.13 occurred at the start of the data series, then you will have only S$74,291.87 to place on the subsequent trade. This will, of course, affect the overall results of the system. On the other hand, it would be possible to introduce new capital to make up this loss, but this also changes the return on capital.

Again, at the end of the day, the amount of capital you are prepared to lose is quite personal. It is probably reasonable, however, to expect that most users of a system would be uncomfortable if the maximum draw down was contained within around 20% of capital.

There are many ratios that should affect your assessment of a system. The simple idea that net profit is the most important is not necessarily true. Many of the numbers that you will extract from a system might not be quite what you'd like to see, but other areas may compensate for a disadvantage in one respect. All values should be assessed in conjunction with each other. Each does have its own particular impact and significance.

For example, let's say you found a system that gave generally good results alt round - except for the fact that the net profit was low while the percentage of profitable trades was near 80%. Assuming that walk-forwrard testing supports the view that the results are stable, then this would suggest that you have discovered a stable system that has an infrequent entry. If you had odds of 80:20 in your favor, the risk appears to be worth taking. The consistency of trading signals is something you can address with a second or third system.

Do make sure, however, that whatever type of system you adopt, you do build in buffers. Optimized results, even when supported by walk-forward testing, tend to overstate profits and understate losses and maximum draw downs. Always expect a worse situation and be prepared for it to happen, rather than believe your system's results will always be the same. Work out how you will react and what action you would take if the worst case scenario were to materialize.

Keep the above in mind and your approach to systems should be safe.

Was this article helpful?

0 0

Post a comment