equity curve generated by most system testing environments (or your own real life trading) is “one-dimensional”
in that it does not say much about the expected variability of the results. It
is therefore difficult to effectively position-size a trading system to meet your objectives simply from a single (hypothetical
or real) equity curve.
technique for assessing what “good and bad” could look like for your trading system or method is to simulate the
variability of your results to generate a series of “what if” situations (i.e. a set of equally likely equity
curves). This gives you a much better indication of the possible range of results
of your trading simply by showing what an equity curve would look like by randomly sampling the trades (or sequences of trades)
in different orders.
In this mini-eBook
we look at simple but effective techniques for simulating the variability of your trading results in order to more effectively
position-size a trading system to stay within your targets for reward and risk.
The main caveats
of simulation are also discussed and these include:
- Sequential versus concurrent trades as part of the simulation
- Intra-trade simulation versus “atomic” trade simulation
- Serial dependency between trades being “lost” in simulation
If done properly,
simulation of the variability of results can be a great benefit in setting sensible position-sizing rules and also setting
your expectation for “what good looks like” and knowing when a trading system or method is operating within or
outside normal parameters. If you know what the likely range of results is for
your system you can have rules in place that monitor, detect, and actually do something about situations where your trading
is not acting “normally”.