I need a good history file for EURUSd - page 3
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
i think no kind of data that you can get for less than a few thousand $ is fine for m1 (or smaller!)...
probably metatrader isn't fine for that at all...
Na... See the list phillip gave. There are some excellent free sources... They are just hard to convert into a usable format.
Have you measured the quality?
I think there's a reason why certain data vendors sell their data for several hundred dollars (each market).
How do you know that they are excellent?
Have you measured the quality?
I think there's a reason why certain data vendors sell their data for several hundred dollars (each market).
Yes. It's more convenient. That's all.
My question wasn't ironic: have you measured the quality objectively?
Go to Dukascopy's site. They have it by the tick. With mili-second accuracy (yes - the timestamps have .xxx at the end). Their reputation and the fact that they are one of the largest ECN's in the world is enough. But when u try to download their data u will understand what I am talking about.... It's an unbelievable pain in the ass.
even tick data is not a guarantee that it does not contain gaps or spikes. even if you aggregate their tick data to m1 to make it compatible for mt it's no guarantee for high quality.
and btw: concerning dukascopy's reputation there are mixed opinions... but afaik discussions concerning brokers are banned here.
so I guess you have not yet done an analysis like that. anyway: thanks for the chat, gotta leave now... good trades.
(...) even tick data is not a guarantee that it does not contain gaps or spikes.
Gaps and spikes are normal. If u want data without them then by definition u want an averaged source - indicative data. You are forgetting that eventually u will trade with a real broker; real brokers have gaps and spikes. All of them.
As for the 'analysis' - don't really know what u mean by that. What kind of analysis do u do?
Phillip: please see my question from page one, concerning historical data from disktrading:
disktrading data is indicative data, meaning (just as Gordon explained) that the data (prices) represents the averaging value from several price feeds at the time. It is not the specific price feed of any given specific forex broker (i.e. market maker in this off-exchange forex business).
Gordon: I think schnappi and I might be causing confusion with the terms/phrase "gaps and spikes"...I believe you are using the terms as characterization of momentary spikes (large orders placed during less liquid market conditions) and the expected missing candles (no market activity within the candle time period) whereas schnappi and I are specifically referring to the more problematic issues plaguing some historical datasets in which spikes of 100's of pips within a given candle can appear (in ways that are truly not characteristic of any financial instrument, then or now) and data gaps spanning days, weeks, and in some cases even months occur (again not the kind of gaps that represent real market conditions).
Schnappi: Yes I have a series of scripts that iteratively crawl thru datasets characterizing and identifying gaps and spikes using the standard litany of statistical tests and so forth. I have no experience with pitrading data. For my style of market trading I actually prefer the "errors" laden in the disktrading data as a means of building in more robustness to the trade triggers/etc.
My philosophy (at the moment) is that if my strategy really relies on pristine and historically accurate price data to make or break the profit/loss ratios then my strategy is guaranteed to be a failure when tasked with dealing with the future market space. There are strong corollaries to be found with weather forecasting (specifically the methodologies) and forex quant trading in my opinion.
Gordon: I think schnappi and I might be causing confusion with the terms/phrase "gaps and spikes"...I believe you are using the terms as characterization of momentary spikes (large orders placed during less liquid market conditions) and the expected missing candles (no market activity within the candle time period) whereas schnappi and I are specifically referring to the more problematic issues plaguing some historical datasets in which spikes of 100's of pips within a given candle can appear (in ways that are truly not characteristic of any financial instrument, then or now) and data gaps spanning days, weeks, and in some cases even months occur (again not the kind of gaps that represent real market conditions).
I see. Yes, those kind of gaps/spikes should be avoided... Agreed.
... For my style of market trading I actually prefer the "errors" laden in the disktrading data as a means of building in more robustness to the trade triggers/etc.
...
As long as you're aware of these limitations, then this is a legitimate way to handle this issue.
Could you maybe tell a few more details how you're defining - let's call it - bad ticks and data holes (instead of spikes/gaps)?
The way I'd do it: I think I'd write an indicator or script scanning for them. Bad tick=maybe a M1-bar > n*ATR or so and a data hole ... don't know ... actually you'd have to handle weekends and holidays (could occur on a week-day) - so how did you do this?