Tickstory

Hi,

I have Ticstory

Im backtesting on EUR/USD with yearly timeframe.

I get 99,9% modelling quality

However, Bars in test and, Ticks modelled is always the same for each year ?

I have store location on an external drive

MT4 settings on internal drive

Please advice,

Hello,

The number of bars/ticks modelled is just a report of how many bars and ticks that were processed during the back-test. Not sure what you are expecting?

Thanks.

Many thanks for reply,

From my understanding, From tickstory, I am downloading FXT files (tick-by-tick data) and HST files (Historical bar data)

I am trying to understand how the bars and ticks are implemented into the backtest.

As stated in subject, I am observing the same bar and tick data appears regardless of the year I am backtesting (same FX pair), logic states it is definitely not the expected behavior. Each year's historical data should be unique. The price movements, volatility, and overall market conditions are different from year to year.

Then why/how come bars and tick data remain the same regardless of period of time in backtest ?

My assumption --> is that the bars represent the chosen timeframe on backtest and, the ticks represent the total number of tickdata downloaded from tickstory to MT4. As a corollary, the bars will change with timeframe but the ticks in the report will -> always remain the same regardless of tested period and/or timeframe. However, the bars will also remain the same if testing the same timeframe but different periods of time (year/month etc). Ie: Backtest daily will report same nr of bars for each and every year, aslo true if backtest stretches over several years.

Thus, Ticks represent downloaded ticks and will always stay the same regardless of backtest. But why does the number of bars remain the same when I backtest one month/year/or five years ? It seems to me that the bars represent chosen timeframe (ie M15 or Daily) during period of one year, and then always will be using that nr regardless of time period tested..?

Please advice,

It seems to work correct when I download and test data for just one year, but not when I download multiple years,

Thus, either the problem is how MT4 handles large FXT files or how Tickstory exports data over longer periods.

Is there some setting in Tickstory's export function that causes it to create an FXT file that isn't optimal for MT4 when I export data over multiple years ?

Assumption: MT4 always uses the entire FXT file for backtesting, ignoring the date range I set in the Strategy Tester.

When I export data from Tickstory for longer periods (e.g.five or ten years), Tickstory creates a single FXT file for that entire period.
MT4’s Strategy Tester will use the entire FXT file for every backtest, regardless of the date range I set in the Strategy Tester’s settings.
Meaning, even if I set the test to only run for one month, MT4 will still use the entire FXT file that covered ie five or ten years, and my report will show the same number of bars and ticks as if I tested the whole period ?

Please advice,

PS: based on your previous answer (The number of bars/ticks modelled is just a report of how many bars and ticks that were processed during the back-test), demonstrates that this support function (sitting in bahamas ??) have no clue. So it would be highly appreciated if you who could connect me to someone that actually knows how the product work so I can get a proper answer on how to solve this issue without having to use a fresh FXT file for each test, download and delete data for every year. DS

Hi Anders,

I'm not sure if this answers your question but the FXT file that Tickstory creates contains all the tick data for the entire exported period. In usual circumstances, MT4 would create just the FXT file for the date range you have specified and therefore the "Bars" and the "Ticks modelled" metrics would change accordingly.

If you need this information for a specific purpose, you could potentially calculate it within your EA and then print it to the Journal at the end.

Hope this helps.

Apologies, but I dont get it.

If everything was working accordingly and all things under usual circumstances, why would anyone ask for support.

As stated in the subject, I receive same bars and tick in backtest regardless of date range.

Please consider, with reference to my previous question:

"Thus, either the problem is how MT4 handles large FXT files or how Tickstory exports data over longer periods.
Is there some setting in Tickstory's export function that causes it to create an FXT file that isn't optimal for MT4 when I export data over multiple years ?"

Hi Anders,

Apologies but your question is still not clear. If you are still asking why the tick/bars reported in the "Report" tab is always the same, this is just reporting the number of ticks and bars in the FXT file. Tickstory has no ability to change this number as it is an internal function inside Metatrader 4 which reports this information.

Regarding your question "Is there some setting in Tickstory's export function that causes it to create an FXT file that isn't optimal for MT4 when I export data over multiple years?" What "isn't optimal"? What is happening vs what you would like to happen (bearing in mind what we have said above)? Perhaps a screenshot might help clarify your question?

Thanks.

Thank you for your response,

You wrote: "What is happening vs what you would like to happen"

-> What is happening <-
When I run backtests in MT4 using FXT files generated by Tickstory, the number of bars and ticks modelled is always the same, regardless of whether I select a short date range (e.g., 1 month) or a long date range (e.g., 5 years). The profit, drawdown, and other metrics do change, which indicates that MT4 is processing some data differently based on the date range. However, the constant number of bars and ticks modelled suggests that MT4 is not fully utilizing the data within the FXT file for the specified date range. It's as if MT4 is only using a fixed subset of the data, no matter what date range I select in the backtest settings.

-> What I would like to happen <-
As you previsously wrote: "In usual circumstances, MT4 would create just the FXT file for the date range you have specified and therefore the "Bars" and the "Ticks modelled" metrics would change accordingly."

I want: the "bars" and the "Ticks" metrics to change accordingly to specified date range. As I have previously stated, Im trying to establish if the problem is how MT4 handles large FXT files or how Tickstory exports data over longer periods.

When I say 'isn't optimal', I mean a setting or that the FXT file might be structured in a way that prevents MT4 from correctly identifying and using the appropriate number of bars and ticks for a given date range.

So question remains: Is there a known limitation or setting within Tickstory's export function that could cause MT4 to consistently model the same number of bars and ticks, regardless of the selected date range in the backtest ?

If not, then as I in my very first post described, I have my FXT files on an external drive, could that cause MT4 not to correctly read the FXT files ?

Rgds / A

Hi Anders,

Thanks for the additional information. Let's see if I can address your points:

> I want: the "bars" and the "Ticks" metrics to change accordingly to specified date range. As I have previously stated, Im trying to establish if the problem is how MT4 handles large FXT files or how Tickstory exports data over longer periods.

Unfortunately this is not possible. The report's information about ticks and bars modelled is merely a depiction of how many bars and ticks that are in the FXT file. I think the main issue here is how is understanding how this is impacting your back-testing and understanding if we can suggest a suitable workaround, if applicable.

> So question remains: Is there a known limitation or setting within Tickstory's export function that could cause MT4 to consistently model the same number of bars and ticks, regardless of the selected date range in the backtest ?

It is not modelling the same amount of bars - it is merely superficially reporting the specifications of what is in the data file. If you would like to see what actual data it is processing, I suggest that you use the "visual mode" of the MT4 Strategy Tester. This will show the actual data - tick by tick - that is being fed into your EA. If you absolutely need to have the actual bar and tick counts that were actually fed into your EA, the only way we can suggest to do this is programmatically in your EA.

> If not, then as I in my very first post described, I have my FXT files on an external drive, could that cause MT4 not to correctly read the FXT files ?

This has nothing to do with where the FXT files reside. As long as you are getting 99.9% modelling quality, it means you are successfully using the exported tick data and your back-tests are all correct and the system is behaving as expected. Again, the report of the amount of ticks and bars will be as described in the FXT file itself (for the entire exported period, not just the tested period). The other metrics of your back-test (profit calculation, etc) will be the result of the back-test in the period you selected (again, you can confirm this with "visual mode").

I hope this helps clarify.