Backtesting and forward testing may be different because of the following:
- EA coded on open bar;
- EA was coded on high/low of the close bar, or using some indicator with high/low of some mode for some indicator;
- we are backtesting using Alpari rus data but many indicators are showing very different results for different brokers;
- different broker's datafeed;
- some unknown coding cases.

Besides there are some scalping EAs which are working fine for 1 or 2 months. There are some EAs which is working fine for 6 months. But after those 1-2 months or 6 months the account will be down up to 0 (day X): so, the day X may be tomorrow or may be next year ...And sometimes we can not see it using backtesting. There are a lot of EAs like that in this public forum and some of them are in elite section as well.

It may be good if we see the code and can look inside the code so we can estimate those cases. But if we can not look inside the ode so we must believe in seller but sometimes the sellers are not telling the true.
And because iof that the general conclusion is that backtesting is not reliable.