Way of the turtle


Download 0.94 Mb.
Pdf ko'rish
bet94/164
Sana29.04.2023
Hajmi0.94 Mb.
#1401749
1   ...   90   91   92   93   94   95   96   97   ...   164
Bog'liq
Way of the turtle the secret methods of legendary traders PDFDrive

Random Effects
Most traders do not understand the degree to which completely
random chance can affect their trading results. The typical investor
understands this even less than the typical trader does. Even very
experienced investors such as those who operate and make deci-
sions for pension funds and hedge funds generally do not under-
stand the extent of this effect. Results can vary to an amazing
degree solely on the basis of random events. The amount of varia-
tion displayed in a series of historical simulations that include ran-
dom events is surprisingly high. This section examines the range
of possibilities that could be due entirely to random effects in the
domain of long-term trend following. 
In the discussion of the edge ratio I ran a simulation for an entry
that randomly enters long or short at the open, depending on the
computer equivalent of a coin flip. I created a complete system that
combines random entries based on a coin flip with a time-based
exit some number of days later within the range of 20 days to 120
days. I then ran 100 tests with the same data that was used in Chap-
ter 10 to compare trend-following strategies. The best test in the
simulation returned 16.9 percent and turned $1 million into about
$5.5 million in the 10.5 years of the test. The worst test in the sim-
ulation lost 20 percent per year. This shows that there is a good deal
of variance that is due entirely to random events.
What happens if we add a little edge? What if we make our sys-
tem similar to a trend-following system by including the trend fil-
158

Way of the Turtle


ter we use with the Donchian Trend system so that trades are
entered randomly, but only if they are in the same direction as the
major trend? The answers to these questions are interesting because
if you examine the track record of any group of trend-following
funds, there is a lot of variance in performance. If the performance
of a particular fund is good, the manager of that fund will say, of
course, that this success is due to superior strategies and execution.
Superior performance actually may be due to random effects rather
than superior strategies. You can understand this better if you con-
sider the extent to which those random effects can influence out-
comes when there is some edge.
If you add a trend filter with a positive edge to the completely
random system, the average performance for 100 tests moves up
considerably. In my simulation, the average return rose to 32.46 per-
cent and the average drawdown dropped to 43.74 percent. Even
with the addition of the trend filter, there remains a large variance
among the individual tests. Out of 100 random tests in the simula-
tion, the single best test showed returns of 53.3 percent, a MAR ratio
of 1.58, and a maximum drawdown of 33.6 percent. The worst test
showed returns of 17.5 percent with a drawdown of 62.7 percent.
Luck or random effects play a large role in the performance of
actual traders and actual funds even though the best traders do not
like to admit that to their investors. Investors believe that a track
record is more definitive than it actually is. For example, anyone
investing in a particular fund generally expects to achieve per-
formance after investment similar to the fund’s historical perform-
ance. The problem is that it is impossible to know the difference
between a truly great trading operation that is having average luck
and an average trading operation that is having excellent luck if one
Lies, Damn Lies, and Backtests

159


looks only at the track record. The random effects are too large and
common for that to be known with certainty.
Consider the best track record from the 100-test simulation that
was cited above. If one traded less aggressively, for example, at 25
percent of the level we did as Turtles, one of the tests would have
achieved returns of 25.7 percent with a drawdown of 17.7 percent
with a 10-year track record. We all know that a trader who entered
randomly would not be more likely in the future to perform at the
same level since there is no edge in trading randomly. Unfortu-
nately, for anyone looking only at a track record, out of a large
group of traders some will have been lucky enough to seem to know
what they are doing when they actually do not.

Download 0.94 Mb.

Do'stlaringiz bilan baham:
1   ...   90   91   92   93   94   95   96   97   ...   164




Ma'lumotlar bazasi mualliflik huquqi bilan himoyalangan ©fayllar.org 2024
ma'muriyatiga murojaat qiling