To end up with a positive TWR, we must make sure APM is larger than SD.
If we can achieve this, the end result for TWR will only be a function of f(the frac-
tion of available equity to risk) and N (the total number of trades).
However, if we ever were to encounter a loss of the same size as the worst
historical loser, the loss for that trade also will equal f. Thus, the higher the sys-
tems profit-generating potential, and the higher the f, the more you stand to lose
in one single trade. Therefore, it’s better to replace the worst historical loser with
the stop-loss level for each individual trade for all trades. By doing so, the worst-
case scenario will be based on a reasonable amount that you’re willing to risk for
each trade, instead of a terrifying amount that you were willing to tolerate in the
past, but would rather not encounter again.
However, the dilemma for the short-term trader is that short-term trading
usually means a low average profit per trade and the use of tight stop losses that
makes sense in relation to the estimated and desired average trade length. But the
tighter the stops on a one-share basis, the fewer trades he can be in simultaneous-
ly for any given f.
To work around this, separate the stop loss (SL) from the money manage-
ment point (MMP). Originally, we assumed SL to be equal to MMP, but by sepa-
rating the two and placing MMP further away from the entry price, we can lower
the number of shares to buy and the amount of our capital tied up in each trade.
As a consequence of this, the fused to calculate the number of shares to buy won’t
represent the true fraction risked of the available equity, but rather just a fictive
fraction to feed the formula:
ST (AC * ffict) / MMP
Where:
ST Shares to trade
AC Available capital
MMP Money management point in distance from the entry price, to cal-
culate the number of shares to trade
ffictFictive fraction of capital to risk to calculate the number of shares to
trade
To calculate the actual amount risked in each trade, we can use the following
formula:
f= ffict* (SL / MMP)
Where:
SL Stop loss in distance from entry price
To figure out how many positions we can be in simultaneously, we only need
to divide MMP by ffict, or, if we already know how many positions we would likeCHAPTER 29 Consistent Strategies 381