reliable intraday historical data

[Lok_CL]

If you are looking for a reliable intraday historical data for your trading system backtesting - visit This sourse
This product is designed for traders who wants to develop own trading system or/and back-test third party's one before use it on real market.
All data are cleaned from errors spikes/gaps by experts.
Data formats : XPO,ASCII and Metastock.
Intraday data for 11 currencies since 1997. EOD data since 1986.
Monthly updates.
Daily data are FREE of charge.

[rmanasa]

Greetings -

I've received different recommendations on data feeds and would like to get the input of the forum on this matter. I'm looking to create a database of historical data for all the pairs, so I can back test some trading strategies and systems.

Now, there are two different types of "data" available out here - actual market prices and something called indicative data. Actual market prices are what you'd expect - the actual O/H/L/C for each time increment of your chart. Indicative data seems to be a composite from a variety of sources, whether they be market makers, brokers, whatever. Wherever it comes from, it is not necessarily the actual tick-for-tick data that went through the FXCM Trading Station (or any other trading platform). While not exact, it's offered as "close enough for jazz" and seems to be a popular format.

I've gone back and forth on this with many porfessionals and beginners. Some feel strongly that you should only use real historical data, and I can understand that. On the other hand, others make a good point about granularity. For larger time frames, defined as the hourly charts and above, an indicative feed is going to be very, very close to the actual prices that had been traded at that time, and good enough for testing purposes. If you're day trading, and need tick level info, then maybe the actual trading prices will give you more accurate information, but for position or swing trading, the difference between a set of actually traded prices and a composite of prices from reliable sources may not be significant.

Hoping for someone with experience in this matter, but looking forward to all replies. Thanks!


Rick Manasa

[Truville]

Originally Posted by rmanasa

but for position or swing trading, the difference between a set of actually traded prices and a composite of prices from reliable sources may not be significant.

I believe it is significant for one big reason: candles.

If you use indicative data, the hourly candle may have a different shape to it than a candle using actual data. The difference between a doji and a regular bar. The difference between engulfing and not. The difference between buy and sell.

In a market as difficult as this, I don't think close enough is ever close enough.

[rmanasa]

Originally Posted by Truville

I believe it is significant for one big reason: candles.

If you use indicative data, the hourly candle may have a different shape to it than a candle using actual data. The difference between a doji and a regular bar. The difference between engulfing and not. The difference between buy and sell.

In a market as difficult as this, I don't think close enough is ever close enough.

Thanks for the reply, Truville. I appreciate the comments and insight.

To give some more background, I'm a relative newbie, looking to practice developing systems and strategies on free historical data before committing to a for fee data feed and the costs involved with that. I haven't found any freely available actual data, so it's indicative or nothing at this point.

Are there any freely (or cheaply) available actual feeds for testing purposes? If not, is indicative data so inaccurate that I'd be foolish to test using it?

Looking forward to your reply. Thanks!