Way of the turtle

The Unknowable Future
In many cases, the reality is that we simply do not have enough infor-
mation to make these sorts of decisions with any certainty. The rea-
208
•
Way of the Turtle

---

son for this is that we do not have enough data. Consider the
sequence QQQVVQ. If this sequence represents the periods of quiet
versus volatile markets, can one determine with any reliability the
relative probability of future markets being volatile or quiet? If you
were paying attention in earlier chapters, you’d realize that a sample
size of six is not enough to draw any certain conclusions. Even if we
have a larger sequence such as VQQVQVVQQQQVVQ, it may
appear that there is a cycle, but there is not enough data for us to
make that assessment with any reliability.
In these cases, it is best to come to terms with the fact that we
do not have enough data and therefore do not know what the future
will bring; therefore, we cannot predict the relative performance of
systems in the future accurately except in very broad terms. A
mature understanding of this reality is critical to building a robust
trading program. As is the case with many aspects of trading, see-
ing the truth is a crucial first step. Once you see it, you can make
decisions that reflect that truth and adjust your actions accordingly.
Robust Trading
Robust trading is about building a trading program that will perform
well no matter what the future brings. It is founded on accepting
the reality that no one can know the future and that there is a very
large margin of error inherent in any testing based on historical data. 
Ironically, you will find that your trading performance becomes
more predictable, once you build a trading program that takes into
account the unknowable nature of the future, The reason for this
seeming paradox is simple: If your trading program is built on the
premise that the future is unknowable, you are assured that the
Bulletproof Systems
•
209

---

future will bring conditions that your trading program anticipated,
something you did not predict. In contrast, a trading program built
on the assumption of a particular set of market characteristics—
almost any assumption at all, in fact—will suffer if the conditions
on which the program is based are not present.
So how do you build a trading program that is not dependent on