Page 1 of 1
Placing Orders in the Market
Posted: Tue Jun 17, 2008 8:37 am
I am also trying to ascertain what to do when one generates a order at a certain price but the market opens above that price.
Do you wait or do you just take the best price you can at the open?
I suppose the most important question is what does the system do in testing?
Posted: Tue Jun 17, 2008 9:49 am
It sounds like a system based on EOD (end of day) data and transactions at the open. The difference between the signaling close and subsequent open is "slippage" and quite possibly should have been a parameter in your testing .... Trend-following and breakout systems have reason to believe that the price move will be disadvantageous between signal and transaction.
Did you test for differences between signaling close and transactionable open prices?
Did that test find any patterns in the slippage? Was it random with mean approximately zero, or did it tend to work to your system's benefit or detriment?
If the answer to the above two questions is appropriate (i.e. "yes" and "to my system's detriment"), then you should include those parameters in your evaluation of system results and decide if you still like the system.
If you didn't test for differences, you are in essence assuming that the difference has a mean of approximately zero and should be willing to live with that assumption.
If you did test, and the mean was zero or the slippage tended to work in your favor, then what's the problem?
FYI my four systems break down as follows:
In two I did not test this, assuming it was approximately zero over the long term. Since these are not predominantly "breakout" style trend-following systems, I have no reason to assume a disadvantageous move and I'm willing to live with that assumption.
In one (partially trend-following) system, I tested results with closing signals and opening transactions, including the slippage by default.
In one (trend-following) system, I tested the slippage and found it to be non-zero, to my system's detriment, but small enough to be non-critical considering the low number of annual transactions, that is, I still like the system.