from the chart where the worst of the trade was. However, for purposes
of analysis, it’s tedious to inspect a chart, look up the low
and the high, calculate the MAE, and record it. As long as you understand
the idea, this is a job better left to a computer.
Measuring Maximum Adverse Excursion
We’re interested in measuring the adverse price movement from the
point of entry In other words, how bad does the situation get when
we have a winning trade? Or a losing trade? This section lays out the
mechanics of actually capturing and displaying this information.
1. Define an entry and an exit. Recall from other article that our
sample trading system had the following rules: go long when
both averages turn up, short when both turn down, and stay
out when they aren’t in agreement. Any other rrading rules can
be used as long as they define the entry and the exit without
stops.
2. Tabulate how far things go bndfor each trade. For each day (in
this example) of each trade, calculate the worst intraday price
movement, if any. If there is none, enter a zero. For long
trades, use the low for the day. For shorts, use the high of the
day. Table 3.1 shows an example of this method of tabulation.
Although this method is straightforward, questions can
come up. For example, what if the price first moves against
the position and then moves favorably? The rule is that MAE
never declines. No matter what the sequence, we shall endure
the worst at some point, so look for the worst that happens
when we trade a specific rule set. Check each day to see if the
adverse price movement has grown.
3. Break the trading results into winners and losers. When you’ve
gone through all your data, you’ll have a list of trades and for
each an entry, an exit, and an MAE value.4 Next, separate the
trades into two categories: winners and losers. List all the
winning trades in one table and all the losing trades in another.
Sophisticates can use a third category,
draws (setting the range of profits and losses to define draws
is an education in price excursion all by itself), but a third category
is not necessary. Here we just need to see the difference
between winners and losers, as well as see if there’s a difference
we can use.
4. Tabulate the MAE for the winners and, separately, tabulate it
for the losers. To measure MAE, we’ll use trading points. Dollars
are just as feasible-and preferable for comparing different
tradables-but to emphasize the connection of adversity
to trading movement, Table 3.3 uses the actual quoted price
of Crude.
5. Sort the tabulations into categories of loss. Here, we must address
the size of the losses we can afford to take.
We want to aggregate the actual adverse movements into
chunks that will tell us something about the relation of adversity
to dollar stops. For example, every adverse movement in
If we broke up “any size” into two bins, one for MAEs less
than 30 points and one for MAEs greater than 30 points, we
could say, “There are three losses with MAEs under 30
points,” or “There is only one loss with an MAE over 30
points, trading this way.” That would give us more detailed
summary information. The trick is to establish whether to use
30 points or some other number.
I recommend you start from your trading capita1.j Of that
number, say $30,000, 2% should be the maximum loss on any
one trade. Before we even get started, we know we’d like the
stop to be no more than 2% of our capital.
Convert that 2% figure into trading points. That’s the
biggest bin size that would work operationally. Bigger than
that and we won’t be able to see if the MAEs fit our capital
stop. Smaller might be helpful, though, so I recommend we
take the 2% chunk size and divide it by 2. By the way, statisticians
call these chunks bins when they’re doing their work
sheets. This term refers to the idea of tossing each instance or
occurrence into a bin for counting.
6. Sort the trades’ MAEs into the bins. We have the trades separated
into winners and losers. Now we break them up by the
amount of adverse excursion they showed. Using the data
from Table 3.3, Table 3.4 sorts the winners and losers by the
sizes of their MAEs. Table 3.5 shows the frequency of winners
and losers in each bin.
7. Graph the summary table. Last, convert the summary table
into a chart. If everything so far has been done in a spreadsheet,
plotting won’t be difficult6 Figure 3.3 shows our summary
chart.
Read More : Measuring Maximum Adverse Excursion
