lose when that inevitable series of bad trades strikes. And the higher the K, the
shorter that series of bad trades needs to be to wipe you out.
The previous examples also assume that what held true in the past also will
hold true in the future. Most likely, though, that will not be the case. Extrapolating
a little from Murphy’s Law, a strategy will most likely not work as well in the
future as it has in the past or during testing. Therefore, one should always trade at
a K value considerably lower than what is calculated as ideal to maximize the equi-
ty. Although this might mean a less-than-optimal equity growth, it also means a
higher likelihood of not being thrown out of the game. In other words, you’re
decreasing the role of luck—good and bad—in your trading.
For example, with the help of the spreadsheet you can try different K values
higher or lower than the one calculated. If you do, you will find that trading at a
K value higher than the calculated K might result in a few test runs producing
higher returns, but also that more test runs will end up with a loss. Similarly, trad-
ing at a K value lower than the calculated K will result in a slower equity growth
for most test runs, but also in more runs that are profitable.
In the extreme example, if you risk everything you have on each trade, it will
only take one loser to wipe you out completely. Even if you don’t risk everything
on each trade, a series of bad trades, where you’ve risked too much, can still force
you out of the game. On the other hand, being too conservative won’t make the most
out of a good trading strategy. Generally speaking, the relationships are as follows:
Risking too little on each trade results in small losses and shallow drawdowns,
but the equally small profits will keep you in the drawdowns for a long time and
the overall equity growth will be slow. When risking too much on each trade, the
losses will be large and the drawdowns deep, but the large winners will help you
get out of them rather quickly and the overall equity growth will be fast. The para-
dox is that the larger the drawdowns you allow, the faster the equity growth. But the
larger the drawdowns allowed, the larger the likelihood to go bust. Also, the larger
the drawdowns, the more erratic your equity curve, the riskier the trading in gener-
al terms, as measured by the risk–reward ratio. Thus, we need to balance the steep-
ness of the equity curve with its smoothness so that the steeper the better and the
smoother the better. We do this by finding the proper amount to risk, which will
govern the size (both in magnitude and time) of the drawdowns. One way to do this
is to use the Kelly formula, and not trade past the calculated K value.
As already mentioned, the major flaw of the Kelly formula is that it assumes
two outcomes only—a winner of a certain magnitude and a loser of a certain mag-
nitude. Trading, with its virtually infinite number of potential outcomes per trade,
is not such a simple game, however. The Kelly formula should, therefore, be used
only for initial research and experimentation. For a better approximation of the
optimal trade size, one should determine what is popularly known as the optimal
f, which requires a slightly more complex calculation.294 PART 4 Money Management