Trade strategies assuming closing price after measuring it -- is there a trick?

R

Rocko Bonaparte

Something I have now seen more than once when looking up trading strategies from an indicator looks something like this:

Enter long at today's closing price if today's indicator (that relies on the closing price) crosses above zero.
More...

The indicator here symbolically represents all the indicators where I've seen this happen; it's just a demonstration.

How can one trade that? You'd need to know today's closing price to enter the position. But you can't then act on that price because the market has closed.

I know there are market orders that can execute at close, which are useful to daytraders who don't want open positions overnight. Still, it's my understanding those activate at some time before the market actually closes--they need time to complete. So the trader would have to make that order much earlier--before they had the closing price information they needed to act.

The only way I can imagine pulling it off is to use a limit on the closing price for the next day. Maybe I'd get it, maybe I won't.

Perhaps what is further frustrating to me is that these strategies then discuss their trading results, and certainly post unrealistic results. What kind of software allows these kind of orders? I began to doubt myself because software seem to tolerate doing that in backtesting. Really? Hell, I'm using my own little backtester in Java and it wouldn't fall for that.

I was torn on whether to consider this a problem of automation, strategy, or software; but in the end I figured somebody fond of this subforum would be best suited to knowing any tricks for it. Are such strategies realistic?
 
Quote from Rocko Bonaparte:



The only way I can imagine pulling it off is to use a limit on the closing price for the next day. Maybe I'd get it, maybe I won't.
More...

You'd get it on all the eventual losers, but not necessarily the winners.

I think the only reasonable thing to do is look at strategies that enter on the next OPEN after a closing price - based signal. But even then you'd have to factor in significant slippage, as fills at the opening price are not usually possible, obviously.
 
Some minutes before closing you will have to assume that the current price will be nearly the closing price. And than check if you would have a signal on that price
 
Quote from autocore:

Some minutes before closing you will have to assume that the current price will be nearly the closing price. And than check if you would have a signal on that price
More...
Then you'd have to act on it. Is there enough time for that too?

I don't see that getting addressed when people pulling that stunt. It sounds like it's non-trivial, and I'm assuming they're not considering it in simulation.

I may post a variant of this in the Software subforum to find out what package tolerate putting orders in like that. It would scare me to think a lot of tools are OK with it. My hope is that these folks in their papers and books are using old tools or Excel voodoo and hence don't have a software tool that would bop them over the head for trying to use that strategy.
 
I have not checked on it in years , but I once remembered something call "Stop Close Only" orders back in the days of the pit. Not sure if we still have them in the electronic age.

Linda Raschke used them in her famous Golf trade when she had epstein barr 25 years ago.

You get filled if the close is above a given price only.
 
I've been experimenting with Ninjatrader and noticed that for the swing trading backtesting that it does, it's using the opening price. So at least by default it isn't affected by this problem. So I wonder what kind of software these people use that lets them get away with tricks like that.

I'm paranoid that the software is "Excel." :D
 
You must log in or register to reply here.
Back
Top