H-system Squashed?
October 17, 2006 by Trader Rich
I decided this weekend that backtesting with 1-minute data just wasn't
accurate enough for backtesting daytrading strategies so I went out to
Gain Capital's site, http://ratedata.gaincapital.com
and grabbed 5 years worth of tic data for the GBP/USD. It took hours
upon hours of grabbing the CSV data, inserting into the database and
then modifying my backtesting scripts but tic data is going to give me
the most accurate results possible.
Unfortunately the results obtained were disappointing and the H-system
failed to show anything but a tiny profit over the last 5 years. This
pretty much throws a huge wrench into the only strategy I've been using
over the last month or so. I don't know how others feel about this but should I be making a decision to shut
this system down? It has been profitable since I started trading it but based on the backtesting results, this would be nothing but a waste of time to trade. I know that past results don't always translate into future results but the whole point of backtesting is to find a system that has past results that are favorable. This would mean going back to the drawing board. If anything, the results instill less confidence and trust in me and this alone can create problems when trading. The most disconcerting thing about all of this is the confusion that it causes and the feeling that I have to start over again. I guess this is all part of trading which isn't supposed to be easy. I don't always ask for advice explicitely, but I sure could use some on this one….
Popularity: 2%
H-system Squashed
October 15, 2006 by Trader Rich
I decided this weekend that backtesting with 1-minute data just wasn't accurate enough for backtesting daytrading strategies so I went out to Gain Capital's site, http://ratedata.gaincapital.com and grabbed 5 years worth of tic data for the GBP/USD. It took hours upon hours of grabbing the CSV data, inserting into the database and then modifying my backtesting scripts but tic data is going to give me the most accurate results possible.
Unfortunately the results obtained were disappointing and the H-system failed to show anything but a tiny profit over the last 5 years. This pretty much throws a huge wrench into the only strategy I've been using over the last month or so. I don't think I have any choice but to shut this system down. This means that I'm back to the drawing board. I will continue to see if there is anyway I can optimize this system to get better results but that will take some time.
Popularity: 1%
Loading ...
Hi Rich,
Needing tick data or not is really dependant on how the ‘h-system’ works, e.g. does it depend on ask/bid within the current candle, etc. Also, are you 110% sure there are no errors in your scripts? I have also gone down the roll your own path, but these programs are so hard to error check.
I developed a stratergy which performed amazingly in MT4 backtests and terribly in tick data backtest, but here is the kicker, it does good in forward testing. I think in the end forward testing is the best test, even better if you can automate.
You will know yourself Rich if market conditions in the last couple of weeks have been conducive to the h-system (or not).
I wouldn’t discount your system based on backtesing. Past performance is not an indication of future gains. The problem with backtesting is that there is no emotion placed into it. It’s just a mass of data. So that time that you rode the wave up a few pips more than your original limit order or got out before hitting stop becuase perhaps the indicators changed. Backtesting has no human element.
Also the FX market is still in it’s infancy and it will evolve over time. The same things that affected the market 5 years ago may be different now. If your making money at it I would continue.
Cheers,
Brian
http://www.canadafx.blogspot.com
I think that if you are making consistent profit with the system now you should continue using it. As brian said back testing is 100% automated hence you cant use desernment. And also the Market conditions are constantly changing, remember that your system is only 20% of your trading, So You as the individual have the ability to pick choose and refuse which trade you are willing to take.
The bottom line is if it works stick with it.