|
PureBytes Links
Trading Reference Links
|
> Here are some that have affected me;
>
> 1) slippage is greater than anticipated,
You must include realistic slippage in your tests or your tests will
give you bogus results. If the slippage is greater than you
expected, then figure out what slippage you are ACTUALLY SEEING and
use that figure in your system backtests. THEN see what results your
system would have given you.
> 2) the tradable security moves quicker in anticipation of
> the item being tracked, in my case, the money index,
So you're running the system on the cash index (e.g. SPX or NDX) and
trying to trade the futures? This is certainly possible -- I've done
it for years -- BUT you have to understand that the results you see
on the TS report for the cash are NOT going to match the actual fills
you get in the futures.
Compare a 1min chart of the cash and futures side-by-side. Notice
how the cash index moves smoothly, with very little noise and jitter.
The 1min bars are very short (H-L is a small number) because the
cash doesn't vibrate around a lot. Now look at the equivalent
futures 1min bars. Notice how much taller the futures bars are.
That's because the futures are much "noisier," with lots more
movement on a short-term basis.
In particular, the futures tend to move farther, and sooner, than the
cash does. This is especially problematic if you use stops (on the
cash) to trigger your entries. The cash moves slowly and cleanly to
your stop, but the futures (which is what you have to actually trade)
race past the futures. By the time the cash moves up to your long
entry stop, the futures have moved much farther. This variation in
the premium (futures minus cash) results in bad "slippage" between
your theoretical cash fills and your actual futures fills.
If you want to run your systems on the cash to take advantage of the
cleaner signal, you have to observe your actual slippage and factor
that into your system tests.
> 3) No trading discipline,
> 4) chasing market,
> 5) failing to take all trades,
> 6) second guessing the system
Those are all different aspects of the same thing: discipline. If
you have a system that works (assuming you use realistic slippage
&etc) then it's up to you to learn the discipline to trade it.
Without that, the greatest system in the world is useless to you.
Gary
|