99% Model Quality

[difenp]

Hello everybody. I have been using Tickstory with MT4 to help me to backtest strategies to achieve a better degree of accuracy.

Unfortunately I still have problems with the size of the FXT files in MT4 that has a limitation. I wanted to ask if there is any alternative to this product sold from an independent developer which seems to solve two of the major issues releated to backtesting in MetaTrader

<link removed by moderator>

I would prefer to receive support from the community or to post a job request on the job board to develop a similar set of scripts to have the same result.

I look forward to hearing from you, Francesco.

[Carl Schreiber]

zen4x:

Hello everybody. I have been using Tickstory with MT4 to help me to backtest strategies to achieve a better degree of accuracy.

Unfortunately I still have problems with the size of the FXT files in MT4 that has a limitation. I wanted to ask if there is any alternative to this product sold from an independent developer which seems to solve two of the major issues releated to backtesting in MetaTrader

...

I would prefer to receive support from the community or to post a job request on the job board to develop a similar set of scripts to have the same result.

I look forward to hearing from you, Francesco.

My 2-years-fxt-files are ~2GB. I guess 3 years ~3GB and below the 4GB limit which is I think a limit of 32bit.

But

1) do you think the behaviour of the markets of 2010 is similar to 2014 and their results are valid today?

2) if so what about 2 fxt files one 2011-2014 and one 2008 - 2010?

[[Deleted]]

I seem to have the same sort of questions:

Maybe someone can advise. Running back-test on my code for these years:

2013-2014 , €1,000 initial deposit turn into $2.1 million.
2012 - 2013 ditto $126,658
2011 - 2012 ditto $1,784

Then all the preceding years the account is blown.

This is euro/dollar on an M5 time frame. $1,000 dollars original deposit with 6% of the account re-invested. I have tried optimising a 'bad' year but still the same results. I am using an Alpari platform and data. I have read that back-testing should not be relied on but how else to test?

These figures, of course, are ridiculas but why the utterly different results from one year to the next. (Each virtual year has around 3,500 trades so I thought this would/should average out the results)

Chris

[Ian Venner]

You should go back through the trades and find where the big profits occur and where the big losses occur and look at the chart to see what the price moves were like and how your EA reacted to them.