If Normalize = True Then Begin
If FuturesMarket = True Then Begin
RecentVolatility = AvgTrueRange(ContractLookback);
NumCont = MaxList(IntPortion(NormEquity /
(RecentVolatility * BigPointValue)), 1);
End
Else
NumCont = MaxList(IntPortion(NormEquity /
Average(AvgPrice, ContractLookback)), 1);
End;ABOUT THE COSTS OF TRADING
Although there is no way around the costs of trading in real life, you should not
concern yourself with slippage and commission in the initial stages of building
and researching a trading system. To understand why this is, we once again have
to remind ourselves that there is a big difference between building and research-
ing a system and actually trading it.
When we’re back testing on historical data, we should not try to squeeze out
as many fantasy dollars or points as possible, but rather try to capture as many and
as large favorable moves as possible, no matter what the cost of trading happens
to be in each specific market. Even if a market turns out to be too costly to trade
in the end, the results derived from researching that market still help us form an
opinion on how robust the system is when traded on other markets.
For example, if a system that tries to pick small short-term profits seems to
be working equally as well in two markets, but one of the markets is very expen-
sive to trade when it comes to commission costs, the results from that market,
when the system is tested without slippage and commission, still help us get a feel
for how good the system is at finding the moves in any market. Had we performed
the testing with slippage and commissions, the poor results from the too-expen-
sive-to-trade market might have discouraged us from trading the profitable one.
Further, considering commissions at this point will also result in suboptimal
parameter settings for the variables in the system. Generally, considering these
costs will favor systems with fewer or longer trades, the results are therefore
skewed in those directions, and we might end up missing a bunch of good short-
er-term trading opportunities.
For example, if you’re building a short-term system for the stock market, it
can be tricky to even come up with something that beats a simple buy-and-hold
strategy. But with a buy-and-hold strategy, you are in the market 100 percent of the
time, with the same amount of shares or contracts for the entire period. What if12 PART 1 How to Evaluate a System