Bastard Testing

I'd like to real-time test a strategy I'm working on. Since it is a short term strategy, slippage will play an important role in identifying profitability. I'd like to check what kind of fills I get, but want to keep my potential losses to a minimum. I'm wondering if it is unreasonable to trade 10 share lots across a wide range of stocks? Obviously it's an imposition on the specialist...

 

do it just to p$ss them off

but stay away from the naz with those odd lots...or we'll kill you :eek:

 

Yup! The slippage is killing me. I've only tested it for 3 days now and every day has been profitable ( I used 100 share lots), but the slippage is murder. I'm only realizing about 30% - 60% of the theoretical gains.
Any brilliant (or otherwise) ideas on how to cut down slippage costs?

 

My series 7 manual states that the specialist has to fill anything under 100 shares at the next traded price after he receives the order.

So this should be quite accurate to backtest as you know the next tick will be yours, no?

 

To cut down on the slippage, try using NQ in your thoretical trading.

 

Try to reduce the number of stops that will be elected.

 

Try using limit orders on entries, exits, or both.

 

You are going to have worse results as your size increases, but I would try NX limits to reduce slippage.

 

what the hell is bastard testing?! sounds interesting..

 

Testing a stock using bastard methods. For example, the 10 share limit gave him exact fills. Unfortunately once you go to the round lot the specialist has the edge.