Way of the turtle

203

---

Monte Carlo Reports
What can we do with the simulated alternative equity curves that
are generated using Monte Carlo? We can use them to build dis-
tributions of the results for a specific measure to determine the
range of possible values one would see if the future was like any of
the alternative universes we created through the simulation. Fig-
ure 12-3 shows a distribution that was created by generating 2,000
simulated alternative universe equity curves, computing the RAR%
for each of those curves, and then plotting the distribution of those
curves on a graph.
The vertical line intersecting the curved line at the top of the
graph shows the RAR% value that 90 percent of the 2,000 simu-
lated equity curves exceeded. In this case, 90 percent of the 2,000
alternative universe curves showed more than 42 percent RAR%.
Graphs like this one are good because they help you realize
that the future is unknown and will come from a set of possibili-
ties. However, one should be very careful not to read too much
into the specifics of reports like these. Remember that these num-
bers are taken from an equity curve that is dependent on past data
and therefore suffers from all the potential pitfalls outlined in
Chapter 11. A Monte Carlo simulation does not make a poor test
better since the simulated alternative universe equity curves are
only as good as the historical simulation from which they are
derived. If your RAR% is overstated by 20 percent because of the
effect of the optimization paradox, a Monte Carlo simulation
using the same optimized parameter values still will overstate the
RAR% for the alternative universe equity curves generated in the
simulation.
204
•
Way of the Turtle

---