Getting realistic fills when backtesting in TradeStation.

Discussion in 'Trading Software' started by walterjennings, Jul 18, 2008.

  1. Hey guys. Im having trouble creating and accurately backtesting strategies using trade station. I believe the main problem is that when putting strategy limit orders through TS, it assumes there is a fill if the price touches the limit price, basically assuming my order was first in line at the exchange, which is unrealistic. This results in a far too optimistic equity curve when back testing. I was wondering if anyone knew any tricks or hacks to change the strategy fills to make them more realistic/pessimistic.

  2. You need to assume 1 tick slippage. In a reasonably liquid market, if your limit order is placed 1 tick shy from the strategy order, you will always be filled.

    You can also assume 1 Tick slippage in the strategy options. If this option obliterates your equity curve, it means your average trade profit is too small to overcome the frictional market costs. In that case, it would be prudent to expand your timeframe of trade to enlarge the average trade profit.

    Hope that helps,

  3. So strategies created with tradestation which average less than 0.01 per volume most likely wont translate into a profitable system when live? I wonder how much false value is added solely from the 'always front of the queue' bias TS gives strategy orders.

    I was playing around with the idea of placing the real orders 1 tick shy of the strategy orders, but then I was thinking Id run into the problems when the case where the prices would fill and reverse on my limit order, but not fill the TS Strategy order, which would cause a desync / divergence from the TS Performance and at worst would invalidate the backtested long term performance of the system.