Metatrader 5 importing Tickdata

About

In July 2017 Metaquotes introduced the capability to create custom symbols including the possibility to import your own price data.
The lacking possibility to import custom price data has always been one of the major reasons for me to pass on Metatrader 5.
Now, more than 8 (!) years after its initial launch Metaquotes finally implemented this much asked for feature. Time to take a look how it's actually working.

Creating M1 Bars

Expected Format

The first thing to note is that as of now you need both Tick and Bar data to perform a backtest on custom price data as MT5 seemingly can’t downsample the provided Tickdata by itself. Luckily, creating these M1 Bars just takes a couple of lines in python code.

We have to take a look at the format of our saved Tickdata first and adjust it to the expected MT5 format. My sample data looks like this.

tickdata format

Metatrader expects the imported Bar data to have the following structure:

Whereas Tickdata needs to be formed like this:

Preparing the Data

Creating OHLC candle data from Bid, Ask prices is easy enough with only a few lines of Python code. The first step is to read our file containing the Tickdata. Keep in mind that if you want to convert a huge amount of price data (multiple GB) you might run out of memory. In this case, you will need to split this data into chunks first before going forward.

import pandas as pd
FILE = 'EURUSD.txt' #our tickdata
usecols = [0, 1, 3] #we only use time, bid and ask
df = pd.read_csv(FILE, 
                usecols = usecols, 
                names=['DateTime', 'Bid', 'Ask'], 
                parse_dates=['DateTime'], index_col=0, 
                float_precision = 'round_trip')

After we have loaded our Tickdata we are going to use Pandas resampling feature and downsample our Tickdata to M1 candles with a single line of code. To avoid having invalid candles once there is no data during a 1-minute timespan (e.g Rollover) empty rows are dropped.

bid = df['Bid'].resample('1Min').ohlc().dropna()

As Metatrader 5 expects the spread in points for its M1 Bar data, we will do another resample of the ask prices and calculate the corresponding opening spread of the M1 bars. Additionally, we add “fake” columns for Volume and Tick-Volume as MT5 also expects those values. In order to draw OHLC candles, Tick-Volume must be greater or equal to 4.

ask = df['Ask'].resample('1Min').first().dropna() #needed for spread calculation
downsampled = pd.concat([bid, ask], axis = 1) #merge bid and ask
downsampled['Tick_Volume'] = 4 #need some tickvolume to draw candles
downsampled['Volume'] = 0 #fake volume
downsampled['Spread'] = (downsampled['Ask'] - downsampled['open']) * 100000 #spread for m1 data in points
downsampled['Spread'] = downsampled['Spread'].map('{:,.0f}'.format) #format spread
downsampled.drop(['Ask'], axis = 1, inplace = True) #not needed anymore

Finally, we just write our downsampled M1 candle data as well as the Bid, Ask Tickdata to *.txt files in order to import them into Metatrader.

downsampled.to_csv("M1_"+FILE, mode='a') #write M1 data
df.to_csv("Tick_"+FILE, mode='a') #write Tickdata

If your TimeDate field has a different format than expected by Metatrader, you need to adjust the date format first before writing to disk. Otherwise, your data won’t correctly import later.

Importing the Data into Metatrader

Now comes the final part. Open your MT5 Terminal and the Symbols Window (CTRL + U) to create our own custom symbol. I’d recommend just copying the needed values from an existing Symbol (in my example EURUSD) and just changing the symbol’s name.

Our freshly created symbol is already preselected in the below picture. Click on “Bars” to import M1 OHLC data or press “Tick” to import Ask/Bid data.
After having clicked on “Import Bars” and selecting the previously created M1 *.txt file you should see an overview of the imported M1 OHLC candle data. Now repeat this same process for your Tickdata.
In the MT5 Backtester we just need to select our newly created symbol and can start the backtest on any desired timeframe as MT5 downsamples our imported M1 candles to the correct timeframe.
You can also select your preferred backtesting method. E.g making use of our freshly imported Tickdata using “Every tick based on real ticks”.

For your convenience, you can download the complete python script from our website. If you have any remarks, questions or suggestions please leave a comment!

python_mt5_prepare.py (972.0 B)

Disclaimer: Metatrader is the registered trademark of MetaQuotes Software Corp. KlondikeFX is not associated with or sponsored by MetaQuotes Software Corp.