Candle aggregation inaccuracies....

Suppose you open an MT4 account with a broker, and then click the H1 button. A few seconds later, your broker sends you H1 data that populates your History Center database.

Fine and well.

But then, you download M1 data, open an offline chart, and run the "Period Converter" set to 60. Va La, you have just overwritten your H1 data that you imported from your broker with aggregated M1 data.

I have noticed a lot of variation between the H1 candles created by an aggregation vs. a broker's H1 import.

My question "Is the M1 data bad, or the H1 data? They can't both be right."

Case and point.... Alpari. Just aggregated M1 back to 2004, and my results went up 100%s when the MQ went to 90%.

Now I am using the Alpari MT4, and the H1 they provided did far worse without the aggregation.

The thing is, my EA trades only H1 bars, so this is bizarre, indeed.

Well I am clearly distraught now. My EA is opening trades against my rules, even with 90% modeling quality. It does so even if I remove all M1 data.

I have a rule that clearly states to enter a trade only if the last 10 H1 candles O-C is less than 10 pips. Sure enough, there is a trade opened where one of the 10 candles is 10 pips from open to close.

90% my butt. MT4, get your act together!

EDIT: After further review,I found that the History Center's data is correct.... that bogus candle's O-C was 9 pips. But on the chart, the popup displays O-C as 10. This is frustrating.