Search
× Search
Friday, December 20, 2024

Archived Discussions

Recent member discussions

The Algorithmic Traders' Association prides itself on providing a forum for the publication and dissemination of its members' white papers, research, reflections, works in progress, and other contributions. Please Note that archive searches and some of our members' publications are reserved for members only, so please log in or sign up to gain the most from our members' contributions.

How to backtest trades in Tradestation

photo

 Tyler Kling, Trader at Privateer Capital

 Tuesday, August 27, 2013

Hey guys, I am trying to backtest a rules based trading system where the underlying is a spread. Bloomberg only gives me 30 days on intraday data so I am trying to do it on tradestation. Can you guys give me some quick tips on how to set up the backtest? Is it possible to simply set up a backtest on a spread or would I need to hardcode and make custom tickers/parameters? The barebones of the strategy involves a spread and RSI on that spread. Thanks!


Print

4 comments on article "How to backtest trades in Tradestation"

photo

 Tyler Kling, Trader at Privateer Capital

 Wednesday, August 28, 2013



thanks!


photo

 Larry Kase, Financial Analyst and Hedge Fund Principal

 Thursday, August 29, 2013



I seem to recall a prior back test quest involve options not long ago. There is a serious problem that virtually precludes any test producing reliable results. The show stopper is the plethora of trades that cannot be executed in real life. The barrier is the inability to precisely align the underlying with the options or futures. Also there is the not so small matter of the bid/ offer spread. I developed a program using options paired with the underlying. Transactions required selling sufficient premium to cover the tail risk associated with the underlying. Essentially, it was a modified risk arbitrage process. In any event it was immediately clear that back testing was impossible and unreliable. Real time with real capital was the only way to test the program and achieve proof of concept. Back testing anything is extremely challenging due to the fact that too many assumed executions will not be possible in real time when applying most back testing methods. Options and futures are impossibly uncooperative and produce misleading results under the best conditions possible. Incidentally, the test period was 15 months and met the stated objective. It was grueling and particularly stressful since real personal capital was at stake.


photo

 Adam Cox, FFIN, MFTA, PMP, Proprietary FX Trader

 Thursday, August 29, 2013



Back testing requires both art and science and reasonable assumptions. Weaknesses such as execution price and volume will catch the unweary particularly in thin markets, and of course in time zones the trader would not otherwise trade in. However, if back testing is completed correctly a lot can be learned from testing, system development, rule redundancy etc., but it does take a lot of experience to set-up 'experiments' and test 'hypothesis' in a productive and valid manner.


photo

 Aubrey H., Business, Economics & Markets Analyst & FX Trader

 Friday, September 6, 2013



I too download chart data into a spreadsheet and manually populate the time fields (I mass populate assuming a consistent start and end time for a period, then code cells to id discrepancies and fix from there).

As long as you aren't using too small a t/f (say m15 or greater?), surely one could expect reasonably probable outcomes from your back-testing?

Also, are you referring to FXCM's TS II?

Please login or register to post comments.

TRADING FUTURES AND OPTIONS INVOLVES SUBSTANTIAL RISK OF LOSS AND IS NOT SUITABLE FOR ALL INVESTORS
Terms Of UsePrivacy StatementCopyright 2018 Algorithmic Traders Association