Trading Systems - What's the deal with hypothetical results?
Jan 21st, 2006 09:51
Markus Heitkoetter, http://www.futures-trading-systems.net
If you have been looking into trading systems, then you realized that
most published results are marked as hypothetical.
This fact might have made you a bit skeptical. Does it mean that the
system won't work in real time?
The answer is no.
Published results should always be marked as hypothetical, even if they
have been achieved in "real" trading. Here's why:
- It is impossible to predict the slippage when using stop or market
orders. Two traders, let's say YOU and John, place an order at the same
time, and John might experience 1 tick slippage, while you are filled
right on your specified entry price.
- You don't know whether a limit order will be filled or not. If your
trading system requires the use of limit orders, you might experience
the following situation: Your order is filled, John's isn't, and the
market retraces. While you took some profits using a limit order, John
is still in the trade and sees his profits shrinking, and in the worst
case turning into a loss.
- Another factor is the account size: If John is trading a rather small
account, then his broker might liquidate his position because he
experiences an intra-day drawdown that issues a margin call. Many
electronic platforms are set up in such a way that they immediately
liquidate a position, even if the market turns around and he would end
the trade with a profit. John then experiences a loss, while you might
realize profits on the same trade.
- What about the ability to withstand losses and your discipline to
follow the trading strategy no matter what? Let's say that after a
couple of losses John decides not to follow the system any longer, and
that's exactly when the system produces some winners. You strictly
followed the system and realized these profits, while John is missing
- Published results are always PAST results. If more traders would have
been trading the system, the prices might have behaved differently.
There could be a difference in price movement when 100 traders try to
enter the market at a certain price point instead of only 1 trader. And
what if 1000 traders placed a 2-lot order at a certain entry signal?
All these factors cannot be fully accounted for when publishing the
results. That's why every serious vendor should mark his results as
Back to your question: "Does this mean that the system won't work in
No. It just means that you should be aware of the limitations of past
No serious vendor can guarantee that a trading system will make profits
in the future, but professional development and thorough testing of a
system increase your chances of making money dramatically.