cutting the losses short and letting the profits run, to create profitable trading peri-
ods rather than profitable trades.
The rolling time-window analysis shows no losing three-year trading
sequence and that the latest sequence, with an annualized return of 10.5 percent,
actually is the worst one over the entire testing period—which occurred during the
current bear market. Note that in Figures 28.1 and 28.2, the current drawdown is
not the worst one and, despite the fact that this is a long-only strategy, it has coped
really well with the current bear market.
The worst drawdown happened sometime in late summer 2001, but since
then we also have had a new equity high, which indicates that the strategy is capa-
ble of making money even during the most adverse times. Granted, Figure 28.1
shows that the equity growth has come to a halt since mid 2000, and more or less
just fluctuated between 8 and 10 million, which probably would have scared off
many investors, but compared to a buy-and-hold strategy, there is no doubt that
this strategy can hold its own.
The robustness of the strategy also shows in the period preceding the latest
bear market during which the maximum drawdown surpassed the 10-percent level
only twice. I would most definitely trust this strategy in the future, but probably
modify it so that it would trade the short side as well, or combine it with a strate-
gy that does. This would most likely decrease the average annual return in a bull
market, but because it also would increase it in a bear market, it is likely that the
overall return over a longer period of time remain approximately the same.
Also remember that both the return and the drawdown are functions of how
much we risk per trade. By decreasing the risk, both the return and the drawdown
numbers will be lower. Therefore, it also is fully possible to first analyze and then
trade the strategy so that you lower the risk per trade when it is evident the behav-
ior of the market has changed into unfamiliar ground for the strategy.
Note that the time in the market per symac is approximately 25 percent,
which means that, theoretically, we need four markets to be in a trade all the time.
With the fictive fat 3 percent and the MMP on average 15 percent away from the
entry price, Figure 26.1 tells us we can be in only five trades simultaneously on
average. If we multiply the theoretical number of markets to track to allow us to
be in trade at all times by the number of markets we can be in simultaneously, we
get 20.
Because this strategy monitors 58 markets, the strategy probably has to skip
a lot of trades or risk a smaller than optimal amount because of the lack of avail-
able capital. To come to grips with this, we probably need to place the MMP fur-
ther way from the entry price and revise the research, risking a little less per trade.
For example, Figure 26.1 tells us that by placing the MMP 20 percent away from
the entry price and risking 2.25 percent per trade, we should be in eight trades
simultaneously. The added trades should help us increase the return and the Sharpe
ratio even further. I leave this to you to research on your own.348 PART 4 Money Management