It's possible, you'll have to edit the data to conform with FSB's requirements though. (in other terms, make it identical in format to Dukascopy data once you've downloaded it from Oanda)

It's something I'm currently considering myself given Oanda's pretty extensive offering of data, just a matter of a time investment.

There's a number of threads here in the forum where you can find good information on data formatting and so on, running a search should net some good resources.

Regards

My reasoning in looking into it was that because data feeds differ by brokers, the tick by tick data might also vary significantly between brokers for any given period.

So someone using Oanda as a broker would benefit from tick data coming directly from the data feed on which they're trading, and the same for someone using the Dukas. data etc.

Whether or not this difference between ticks by data feed actually exists, and if it does, whether or not it actually results in significantly different trading results when backtested upon I don't know (perhaps it would provide a more accurate representation of how that broker's feed behaves during news events, and so on)- which is why running the test seems beneficial.

If anyone else has already given this a try or has any further insight and wouldn't mind posting it, it'd be greatly appreciated.

Bottom line really: Oanda tick data could yield more accurate/relevent results for someone trading on an Oanda platform. Dukascopy tick data could yield more accurate/relevent results for someone trading on Dukascopy.

It's just a thesis, but if it does yield even a small advantage/difference as far as backtesting is concerned, then it would seem to be worth considering.

Using data specific to the broker which you're trading on would seemingly yield testing results which more accurately reflect the results you'll receive in live trading using that broker.

In other words, if there's commonly a small price difference (and this could become more pronounced during news trading etc.) between the data from one broker and the data from another, then the price movements reflected in the tick data (notwithstanding the spread) would be different. So using the data from your broker would mean that you're testing on a data set which is more in tune to what you'd be running trades on.

This would be especially true if the feed from one broker tended to behave a certain way during big news, or peak hours, or down time and so on-- and the feed from another broker behaved in a different way. One feed could perhaps react more sharply or quickly than another to price changes due to the amount or type of liquidity available to the broker for example.

As far as I'm concerned it's just a theory or thought at this point until I get around to testing it. Again if there are others who have more experience with this it'd be great to hear from you.

Here's their ticks format:

(notepad)

2/1/2004 9:22:00,106.88,106.91
2/1/2004 9:22:00,106.89,106.92
2/1/2004 9:23:00,106.88,106.91
2/1/2004 9:23:00,106.88,106.91

(excel)

2/1/2004 9:22:00    106.88    106.91
2/1/2004 9:22:00    106.89    106.92
2/1/2004 9:23:00    106.88    106.91
2/1/2004 9:23:00    106.88    106.91

I can offer the data through something other than excel or notepad if that's helpful just let me know.

An import feature for the data would be fantastic. To get the data unfortunately however you have to be an account holder at Oanda. So perhaps this would only be helpful then to a few users here. You put in a request from the account management screen for whatever pair and time period you want and they send you the prepared file within a day or so.

The offering is good. Their ticks for majors go back into the 1990's.

Again as far as the qualitative differences with Dukas tick data I can't say at the moment as I haven't yet tested on it, although I plan to soon.

The default file names are bizarre, not that it's truly relevant except for the sake of convenience. Could be random or a function of their delivery database. This is the unchanged name for a Jan/01/2012-Jan/31/2012 ticks file:

502C1FF14E8BAA47F08A2A17FDCBC241

Can you upload somewhere data files for the major symbols. lets say 1-2mil ticks per file.
Is there bar data or only tick data?
Do we need to filter out some weekend data?
If only tick data are provided, we have to compile bar data from them.

I can make a separate converter tool easily. If it works well we can include it in FSB.

Many thanks to Mr. Popov for following up with this.

A good deal of time that would have been spent formatting data may be preserved for other profitable endeavors!

Please report if it works properly since I do not have access to Oanda data.