The reasoning used in the stop-loss version of the system also holds true for
the trailing-stop version, with the addition that the trailing-stop version even is a
tad better still.
All in all, we can be very sure that we will be able to repeat these historical
and hypothetical results in the future when we trade meander on previously unseen
data. The only two negatives are the amount of time spent in the market and the
percentage drawdowns, which still are relatively high compared to the version
depicted by Table 10.5. Nonetheless, we’re looking forward to continuing our work
with this system and believe we will be able to address both the time spent in the
market and the drawdown shortly, with the help of a few (trend) filters.Volume-weighted Average
This system presented us with a couple of interesting choices. When tested using
the stop-loss method, the percentage-stop concept indicated the system would
work best with a stop loss of 0.4 percent, a profit target of 4.5 percent, and a seven-
day maximum trade length. The estimated average profit came out to 0.3 percent,
with about 35 percent profitable trades. The ATR concept, on the other hand, indi-
cated the system would work the best with a trailing stop of 1.4 ATRs, a target of
2 ATRs, and a six-day maximum trade length. The estimated average profit per
trade came out to 0.4 percent, with about 45 percent profitable trades.Stop-loss Version
This is a good illustration of how the very same set of entry signals can produce two
distinctly different types of systems, even though both versions are fairly short-term,
with maximum trade lengths of six and seven days, respectively. Note that for the per-
centage method, the initial risk–reward relationship going into the trade is more than
12:1 (4.5:0.4). This is not the true and final risk–reward relationship: Given that the
stop is fairly tight, while the target is rather generous, it is reasonable to assume that
most losing trades will be stopped out at the maximum allowed loss, which explains
the relatively high amount of losing trades. Most winners, however, will be stopped out
on the seventh day in the trade, before they reach the target. Assuming the average loser
also equals 0.4 percent, we can estimate the average winner to be 1.6 percent [(0.3 * 1
0.4 * 0.65) / 0.35] for a more accurate risk–reward relationship of 4:1.
We really can’t do the same type of calculation for the ATR concept. The
most obvious reason for this is that we use different measuring sticks for the input
and output variables. Because the inputs are measured in ATRs and the output in
percent, we can’t combine them into one formula, as we just did for the percent-
age concept. But even if the inputs and output where the same, the numbers for
this system still tell us we cannot use the formula to come up with a more accu-
rate risk–reward relationship.CHAPTER 20 Adding Exits 241