TICK TO TICK SIMULATION WITH METATRADER

[newdreams]

Lately I've had some arguments with colleagues about if it's possible to make a tick to tick real backtest with metatrader (using Tickstory software). I think it's possible, but somebody say the minium temporary bit in metatrader is one minute and everything that happens below one minute is a simple mathematical simulation, it don´t have precise correlation with reality.
Please could someone provide solid evidences for or against?

Thank you

[tickstory]

Hi Newdreams,

MT4's historical data does have a minimum granularity of 1 minute which means that if you are using your broker's data, then yes, you are only going to get tests on interpolated points within each of these one-minute bars (for example, the Open, High, Low & Close is taken, resulting in 4 data points). To process these generated points (or "modelling types" as MT4 refers to them), MT4 creates an (FXT) data file which contains these individual "ticks". It then proceeds to back-test on this data. Modelling types can be 'Every tick' (every tick that was downloaded from the broker while your terminal was connected), 'Control Points' (OHLC of the bar - 4 data points) or 'Open prices' (The open price of the bar - 1 data point).

In the case of Tickstory, it bypasses MT4's generation of the data file and instead produces it's own 'Every tick' model data file with tick-by-tick data. That is, rather than having merely a few data points that occurred within that minute, it will have all the actual ticks that occurred. The MT4 back-tester will then process this data as usual.

Below are some additional threads which may be relevant in your discussion about the accuracy/value of back-testing.

http://tickstory.com/forum/viewtopic.ph ... 857&p=3954

The value in backtesting is the ability to evaluate a strategy to-a-point. You have probably heard the phrase "past results do not guarantee future results" and this is the case for back-testing. The best you can do is understand your strategy and the variables involved so you have some confidence it can handle various market conditions, brokers, etc. Once you have done this, you are still not guaranteed to get the results you see on your backtest report - you will then need to go through a period of forward-testing on your chosen broker where you will find various other variables that cannot easily be simulated (trading lag, price discrepancies, slippage, etc). Then, when you finally are confident enough to go live with it, you are still likely to see some discrepancies between your simulated results versus live. This doesn't mean that back-testing has no value. The process of testing your strategy is a matter of prudence to ensure you are not merely gambling that your strategy will work as you hope. When you work with tick data, this is just another tool to help you more accurately simulate market conditions (especially for scalping strategies).

http://tickstory.com/forum/viewtopic.php?f=2&t=520

Here are some factors to consider when trying to get as accurate back-tests as possible:

- Does your broker's data feed match with your test data? If they are markedly different, expect markedly different results for HFT/tick-based strategies.
- What is your ping to the broker's server? Remember that back-testing assumes a ping of 0ms where this is not reasonable in reality. If you have a HFT trading strategy, you should code assume some delay when placing your trades in the back-tester.
- Is your strategy reliant on spread? If so, test with varying spread values to make sure the back-test results aren't adversely impacted by small changes to the spread.

We'll leave the discussion to others to comment further

[newdreams]

thank you for your answer

[dmitrievsky]

Hello, another question about the simulation ticks. In MetaTrader tester minimum time resolution of 1 second (modeling time), but for 1 second can come a few ticks, a different number. Do the ticks in the tester timing inside 1 second or arranged randomly?

[tickstory]

Hi Dmitrievsky,

Ticks that occur in the same second are ordered (and executed) in the same chronological order as they originally came in.

Regards.