or one trading sequence, whereas cells F71 to F73 indicate how much the average
profit per trade is likely to vary over several markets or several trading sequences.
It only makes sense that the average profit per trade is likely to vary less than the
actual profit per trade, just as, for example, a moving average of prices fluctuates
less than the price itself. Normally, the values in cells F71 to F73 are of greater
interest than those in cells G71 to G73.
Cell H71 shows the average risk-adjusted return for one market or trading
sequence. The value 0.15 is calculated using the formula AVERAGE(H$2:H66),
where column H refers to the risk ratio as calculated with the TradeStation export
function. The value in column H also could have been calculated using the values
from columns F and G, so that the value in cell Hnwould equal Fn/ Gn.
Cell H73 shows the risk-adjusted return over several markets or trading
sequences. The value 0.96 is calculated using the formula F70/F71. Note that the
value in cell H73 is much higher than that in cell H71. This is because in cell H73,
we’re using the standard deviation of the average profit per trade from several mar-
kets, as opposed to the standard deviation of the profit per trade from individual
markets, as in cell H71. Because the standard deviation of the average profit per
trade is less than the standard deviation of the outcome of all individual trades, the
risk-adjusted return will be greater over several markets or trading sequences than
over one single market or sequence given that the returns are similar from all mar-
kets. The value in cell H73 is of greatest interest to us.
The calculations and values in cells C75 to H78 should be pretty much self-
explanatory. The most important cells or values to consider in the bottom half of the
table are the average profit and risk factors and their respective standard deviations.
An average profit factor above one means that the system tested profitably enough
on the profitable markets to make up for the losses made in the losing markets. The
same goes for a risk factor above zero. One way to get a dollars-and-cents feel for
the risk factor is to equal its value to dollars made per dollar risked. In the case of
Figure 7.2, a value of 0.28 indicates that the system is likely to make 28 cents per
dollar risked. Note, however, that if the standard deviation for the risk factor is high-
er than the risk factor itself, a number of trading sequences will lose money. In this
case, a value of 0.29 in cell D76 indicates that 68 percent of all trading sequences
will produce a profit per dollar risked ranging from 1 cent to 57 cents.
As already mentioned in the beginning of this section, the trade length and
time spent in the market variables are important in that we need to balance these
numbers against the number of markets we would like to trade and the nature of
the trading strategy. Aside from that, because this is a book about relatively short-
term trading strategies, we will try to keep the average trade length down to
approximately five to ten days. Finally, at this point of the analysis work, we won’t
care so much about the drawdown values. As mentioned in Chapter 5, the draw-
down is not a core system characteristic and is more efficiently dealt with via
mathematical expectancy and the probability for a winning trade.94 PART 2 Trading System Development