Reply

Analysis of data

5 replies

AndrEAs SQ

Customer, bbp_participant, community, 44 replies.

Visit profile

4 years ago #254786

Hi guys, when I analyze Dukascopy data (downloaded by tick) with the data manager analysis tool in M1 data, I see that the most common problem is gaps. Reviewing the different pairs of forex I find that most pairs have an average of 0.3% of gaps, which seems normal to me, because gaps are something that happens in trading, and more in M1 charts. However, some pairs have between 3.5% and 4.5% of gaps. Is that amount still normal, or should I download those pairs again?

Even when I get out of forex and check some indexes like the US30 or others, I find gaps of 20% of the data. I can trust in those data?

Thanks guys

 

The worst case

AndrEAs

0

tomas262

Administrator, sq-ultimate, 2 replies.

Visit profile

4 years ago #254795

There used to be a very low volume trade with Dukascopy for this symbol (few ticks). This caused many gaps in 1 min data. It is better to trade US30 (Dow Jones) using the E-micro Futures Contract being traded on CME https://www.cmegroup.com/trading/equity-index/us-index/micro-e-mini-futures.html

The spread is usually 1 tick (0.5 USD) and the point value also 0.5 USD. In SQX you can get data for this using our Futures pack https://strategyquant.com/data-subscription

0

AndrEAs SQ

Customer, bbp_participant, community, 44 replies.

Visit profile

4 years ago #254797

Yes, it is very logical that it is not at all advisable to use data with 15% or 20% of gaps as with the dukascopy indices. thanks Tomas for your advice; But what do you think about forex pairs (also downloaded from dukascopy on tick) that have some up to 4% of gaps, are still in a normal range of gaps, is it advisable to use them?

 

AUDUSD Gaps 4%

AndrEAs

0

hankeys

Customer, bbp_participant, community, sq-ultimate, 487 replies.

Visit profile

4 years ago #254798

analyzing 1M data doesnt make sense, we dont need to care about those small resolutions – if wee trade higher TFs as from M15 higher

about the indices you need to know, that gaps in here could be normal, because of trading hours – and because you are in CFD bussiness every broker could have different trading hours and they are changing through history

Trading hours for DAX UTC1 could be 08-22, 02-22, etc.

also the specifications of prices could be different – dukascopy could have 3 decimal prices, your broker could have 3,2,1,0 decimals

for trading indices with dukascopy data you need to know what are you doing…we learned months how to trade indices with those data

You want to be a profitable algotrader? We started using StrateQuant software in early 2014. For now we have a very big knowhow for building EAs for every possible types of markets. We share this knowhow, apps, tools and also all final strategies with real traders. If you want to join us, fill in the FORM.

0

AndrEAs SQ

Customer, bbp_participant, community, 44 replies.

Visit profile

4 years ago #254803

Hey Marek, nice to see you after so much time. Thanks for answering. I understand your point with regard to the fact that we trade in larger TFs, and consequently we should focus on them. However, for me it makes sense to check M1 in forex, because after reviewing many pairs and in many TFs, you realize that if an instrument has a large percentage of gaps in M1, it will also have a large percentage of gaps in Larger TFs, while the pairs with low percentages of gaps in M1, have lower percentages of gaps in larger TFs. The difference lies in the fact that when reviewing larger TFs, the percentage tends to decrease, but the proportion remains, and that is why I review M1 to more easily look at which data have higher gaps.

But I have no problem changing the question, because the only thing I want to check is that there is no error exclusively mine from data downloads. So, for H1 and H4, the forex pairs normally have gaps below 0.1% on average, but there are pairs that are far from the average such as the EURJPY or the AUDUSD that have gaps of almost 3%; So, based on the data you use, do you consider that almost 3% of gaps in H1 and H4 data are within normal range? Even, what you can see in the attached images for the end of 2017 and throughout 2018, is it normal for an H4 TF? Last year I was very involved in the world of cryptocurrencies and away from SQ and forex, but I can not imagine that it could have happened so big with the EURJPY and the AUDUSD so that 2018 was the year of the mega gaps, that’s why I find it weird.

 

EURJPY H4 2018  data Gaps:

EURJPY 2018 gaps

 

AUDUSD H4 2018 data Gaps:

AUDUSD 2018 Data Gaps

 

With regard to the indexes, thanks for the advice, that is a subject that I have a long way to learn.

 

Best regards.

AndrEAs

0

AndrEAs SQ

Customer, bbp_participant, community, 44 replies.

Visit profile

4 years ago #254804

Hi guys, I already fixed the error, I checked that if there was a problem in the data download, I overwritten the data from 2017 onwards and those errors were corrected. I knew that it existed, but I saw very far the possibility that the data would be downloaded badly, and if it were not for the new SQ data analysis tool, it would assume that everything was fine, just as I did with tick downloader; but I see that it is not so, and that the download failure is more possible than I thought, so, now I realize that it is worth analyzing the new data that we download.

Thanks to the SQ team for the new tools, constant evolution and good work.

Attach an image with the data already corrected for the AUDUSD H4. The gaps dropped from 2.1% to 0.2% in the H4 TF.

 

AUDUSD corrected

AndrEAs

0

Viewing 5 replies - 1 through 5 (of 5 total)