A journal of using a Forex back testing application

  • Definition of Terms
  • Setting Boundary Conditions of the FT5 tests
  • Creating test strategy statements and test plans that confirm viability of strategies

The last test was our first exposure to the outputs of a backtest.

We will now follow the same procedure with the next series of tests, but we will apply some real world constraints to simplify the overall scope of the tests and to avoid getting lost in detail.

In terms of Strategy (the what and why) for each “What” there will be an associated “Why” using the same test number sequence 1, 2, etc. This will allow us to identify which test is which, and will be important when we move from unit testing to integration testing.

Once we have written down the strategy that describes the multiple tests, we will move on to writing the Plan for each of these strategies. This is the beginning of our Test Strategy and Test Plan.

Definition of Terms:

Boundary Conditions – defines what is in scope of this series of tests

Integration test – a test whose scope is a combination of unit test indicators

NNFX method . No Nonsense Forex Method that can be searched in a browser. Author acknowledged as “VP” (Patrick Victor).

Output – A tangible and measurable result at the end of a unit or integration test.

Test Plan – Who, when, where, how

Test Strategy – What is going to be tested and why

Trading Bank – the total amount of funds committed to all Trading activities.

Unit test – a test whose scope is for one indicator at a time, whose settings may be changed during sub-tests to determine any optimum settings

Setting Boundary Conditions of the FT5 tests

There follows a list of statements, all of which are included in the overall Forex Trading Plan, to which the Test Strategy and Test Plan relate. These prerequisites will be referred to as Boundary Conditions of the FT5 tests.

These are extracted from the Requirements of the author’s 2021 Forex Trading Plan.

  • The Trading Plan shall follow the NNFX method
  • The NNFX method is a trend following strategy only
  • The NNFX method uses the daily timeframe (D1) only
  • The NNFX method uses a maximum risk per trade as 2% of Trading Bank
  • The initial duration of back tests shall be six months
  • The scope of instruments shall be Forex currency pairs – majors and crosses
  • Initial back tests of different indicators shall be limited to five currency pairs
  • The back tests shall include realistic live account estimates of broker commission, spread and rollover where possible
  • One test success criterion will be that wins > 60% and losses are < 40%
  • One test success criterion will be that the bank drawdown over the period shall not exceed 10% of bank.

Getting used to creating test strategy statements and test plans that confirm viability of strategies

Test # What Why
1 Rerun a test in FT5 of the built-in “Strategy” referred to as “Awesome EA”. Use defaults for the test period, timeframe, currency. Do not adjust settings of the Awesome EA To confirm that FT5 returns the same test results for the same parameter settings on two consecutive tests.
2 Apply all boundary conditions to Awesome EA strategy. Define all settings for each currency and reference source of settings data. There is an unlimited number of tests we could do to demonstrate the usefulness of FT5. We are going to put some practical limits on these tests. To limit the number of tests in scope of FT5 testing.

To inform the user of the change of output resulting from a change of instrument settings that replicate real live trading accounts.

To determine if test plans deliver minimum criteria to warrant further testing.|

TEST PLAN 1

Who? FT5 tester Mondeoman

When? 5Nov21 – estimated duration of the test including write up of results on Babypips forum = 2 hours

Where? At home/office location.

Define Prerequisites to perform the test 1:

Access to a test end user computer and registered version of FT5, Access to EURUSD data, Default test period, Default currency pair, Default settings, Default timeframe (all as previous post).

Define How?

As per previous post.

TEST PLAN 2

Who, when, where? Same as Test Plan 1. Added input from @Alg626 on the subject of real spreads and rollover costs

Define Prerequisites to perform the test 2.

Access to a test end user computer and registered version of FT5, Access to EURUSD data, (all as previous post).

Default test period – six months 01May 2021 to 31Oct21

Default currency pair EURUSD, Default timeframe H1

Default settings – list each setting applicable

Real spreads and rollover costs.

The associated thread in which I requested input from other members on the subject of spreads and rollover fees has been updated. Thanks to @alg626 for the input.

Spread 2 pips (two way turnaround)

Commission $0

Rollover cost $0.00

To be continued……………

4 Likes

Is this platform an accurate backtetser ?

Hi Jordan,
Sorry I cannot answer your question.

I have managed many infrastructure and application software rollouts in my time as an infrastructure project manager, so thought this would be a good idea for my training in Forex

I had to suspend this work back in Nov21 due to ill health. It was a commitment I made to the software proprietor that I was unable to finish. My health has now recovered but the downside is that I have had to curtail my activities a lot, to get my time back to be able to look after myself and my family.