Way of the turtle

56
•
Way of the Turtle

---

30
40
20
10
0
Monthly Return Distribution
30%+
30
40
20
10
0
28% - 30%
26% - 28%
24% - 26%
22
% - 24%
20% - 22%
18% - 20%
16% - 18%
14% - 16%
12% - 14%
10% - 12%
8% - 10%
6% - 8%
4% - 6%
2% - 4%
0% -
2%
Losing Months
Number of Months
Number of Months
2%+
2% - 4%
4%
- 6%
6% - 8%
8% - 10%
10% - 12%
12% 
- 14%
14% - 16%
16% - 18%
18% - 20%
20% - 22%
22% - 
24%
24% - 26%
26% - 2
8%
28% - 30%
30%+
Winning Months
Figure 4-3
Distribution of Monthly Returns 
Copyright 2006 Trading Blox, LLC. All rights reserved worldwide.
•
57
•

---

Donchian - R-Multiple
™
 Distribution
2.5R+ 2R
1.5R
1R
0.5R <0.5R
2997
3746
2248
1497
749
739
1659
32
10
6
0
1483
1854
1112
742
371
0
80%
100%
60%
40%
20%
0%
Losing Trades
Number of T
rades
Number of T
rades
Winning Trades
1301
<1R
1R
2R
3R
4R
5R
6R
7R
8R
9R
10R 11R
12R 13R
14R 15R+
406
221
126
78
42
44
22
15
6
9
5
2
3
18
29
828
Figure 4-4
Distribution of Trade Outcomes 
Copyright 2006 Trading Blox, LLC. All rights reserved worldwide.
•
58
•

---

example, the 100 percent level for losing trades is 3,746, which
means that there were 3,746 losing trades during the 22 years of
the test. The corresponding 100 percent level for winning trades is
1,854 trades.
The trades are divided into bars on the basis of the amount of
profit they earned divided by the amount risked on a trade. This
concept is known as an R-multiple and was invented by the trader
Chuck Branscomb as a convenient way to compare trades between
systems and between markets. (R-multiples were popularized by
Van Tharp in his book Trade Your Way to Financial Freedom.)
An example will help illustrate this system. If you buy a contract
of August gold at $450 with a stop at $440 in case the trade goes
against you, you have $1,000 at risk since the $10 difference
between $450 and $440 multiplied by the 100 ounces in a contract
totals $1,000. If that trade earns $5,000 in profit, it is called a 5R
trade because the $5,000 profit is five times the amount of the
money that was risked ($1,000). In Figure 4-4, the winning trades
are divided into buckets at 1R intervals and losing trades are divided
into buckets at 
1
⁄
2
R intervals.
It may seem odd that in this histogram the losing trades out-
number the winning trades by so much. This is actually a very com-
mon occurrence in trend-following systems. However, although the
number of losing trades is very high, the system keeps most losses
close to the desired entry risk of 1R. Winning trades, in contrast,
are many times the entry risk, with 43 trades returning 10 or more
times the entry risk.
How does this help one think like a Turtle?
As Turtles, we never knew which trade would end up being a
winner and which a loser. We just knew the general shape of the
Think Like a Turtle 
•
59

---

distribution of possible outcomes we might encounter: A distribu-
tion much like the ones shown in the graphs above. We thought
that each trade possibly could be a winner but that most probably
would be losers. And we knew that some would be medium-sized
winners of 4 to 5R and some would be large winners of 12R or even
20R or 30R. But ultimately, the Turtles knew that the winners
would be large enough to cover the losses from the losing trades
and that there would be profit left over.
Thus, when we made a trade, we did not measure our personal
worth by the outcome of the trade because we knew it most likely
would be a losing trade. We thought in terms of probabilities, and
that gave us the confidence to make decisions in the face of large
degrees of risk and uncertainty.