Strategy Tester

Is anyone else skeptical as to the reliability of the Strategy Tester. I've only been working with it for about 6 months now, so maybe I'm missing something.

Here is what I do:

1. I download data from Alpari
2. I convert to appropriate time periods
3. I back test until I find successful strategies
4. I move successful strategies to live demo accounts
5. I compare results of live testing to back testing for that same period of time

Step 5 almost always fails. There is almost no similarity between the back testing and live results. I am running at 90% modeling quality and I even shift back 1 period so that I am always using the last completed interval. This should give the most reliable results in my opinion.

I understand that the different brokers (data) would cause some discrepancy, but the difference between live results and back testing is just too big.

I am testing 7 different strategies on live demo accounts and so far the trades are quite a bit different when compared to back testing.

Anyone have any insight? I have all the data from both live testing and back testing if anyone would like to see it.

Quite a few people are skeptical of the backtesting functionality in MetaTrader.

The problem, at least in part, seems to stem from the way MetaTrader simulates tick data. MetaTrader basically 'makes up' ticks based on the price data it has available in history.

There are multiple posts discussing backtesting, the best advice I can give is to only use open prices on the one minute chart to insure you're working with real prices.

I'm not sure if this is this the only issue MetaTrader has with backtesting, but this should get you much closer to reliable results.

You can trust only demo/real trade results as Tester from build
to build may behavior differently not check this thread:
http://www.forexfactory.com/showthread.php?p=1507134
It generate ticks and open trades outside of candles !!!!

Hi there
Here go my experience:
Back testing is fantastic, but first you have to download data from a real account, and there tick are exactly same as real, for my broker I can ensure I have test and compare live results with back test for more then 75 days in 1M time frame, and I only get difference in one single trade because of re-quote.
For data on demo account is completely different.