Dukascopy Price Feed

Attached is an AUDUSD chart plotted from Dukascopy's data (http://freeserv.dukascopy.com/exp/) for the last FOMC news.

Is it accurate?

Just look, each candle has a LOW, HIGH that is more than 15 pips?
If this is indeed accurate, heck, anyone can be a millionaire by putting Limit Buy/Sell orders 5 pips from Open prices!

Hi,

i wrote a function for a tradingsystem that aggregates time-bars out of streaming tick-data. From my experience i can tell you, that on small timeframes, like your 1-min chart, the resulting charts can look vastly different depending on what method you use for bar-aggregation. For example, you could choose to base your bars solely on the bid-price, or you could choose to use the highest ask and the lowest bid for the highs and lows and use median price for open and close, or you could strictly use the last-trade price, and so on and so on.

Like for any other high-frequency trading system, to test your idea you need much more data than just a 1-min chart (that uses an unknown aggregation method), and you need to know exactly what this data is and how it was sampled. You need to track the bid and the ask and their respective sizes, and also account for commissions, anything less will lead to skewed results.

Then, your million-dollar trading idea is incomplete, you need to apply some sort of stop. Only with bar-data, you cant tell if your stop or your 5-pip target was hit first intra-bar. Simulation and deployment of such trading-systems demand a very high level of data-integrity and sophisitcated, most of the time custom-developed, software.

In short, their price feed is junk useless.

Im not formulating any "million dollar" strategy but rather trying to use their data for historic charts nothing more. surely that comment was meant as sarcasm for their price feed.

Im not saying its junk, its just not entirly clear how they compose their bars and that its important to know that if you want to conduct tests in very small timeframes.

My feed for AUD/USD shows 8-pip ranges (on average) on the 1-minute chart in the minutes following the FOMC spike on Sept 7th.