minimum trade samples to test a trading strategy

Backtesting Strategies

This clause explains how to backtest strategies in JForex.

Backtesting of strategies in JForex is done exploitation the Historical Tester , which aims to replicate marketplace conditions with historical data and runs the strategy American Samoa a pretending within the sampling history. The data sample is built from marketplace information (tick aft tick), therefore allowing the effectuation of powerful and efficient back examination methods.

                Past performance is no more reliable indicator of future performance.

Historical Quizzer

To open the Historical Tester and start the feigning:

Go to View in the menu bar at the top of the political program and select Historical Tester.
The Humanistic discipline Tester is now displayed as a new tab (next to other tabs so much as Orders, Positions, Messages).
Select your strategy from the drop down menu (if it has already been added to the workspace, otherwise add it).
Adjust the tester settings as required (more on the different options below).
Click connected the start push ( ).
The Define Parameters window appears. This is where you set the variable values (as defined in the strategy code, e.g. official document, slippage, amount etc.) that are used for the simulation.
Click on Run to set off the pretending.
After the simulation has over (this may take a piece), the testing report is opened in your browser.

Historical Tester Overview

Open (1) - opens the file dialog window which allows you to add new strategies.
Edit (2) - opens the strategy in the scheme editor (available only if source file .java is gettable).
Modality Mode (3) - if enabled, the simulation is drawn on the graph while it is being tested.
Optimisation (3) - allows to run multiple tests if the scheme uses variable parameters.
Show messages (3) - shows messages if enabled.
Account (4) - allows to define tester calculate settings (fairness, calculate currentness, etc).
Instruments (5) - historical data is loaded for selected instruments only (the testing instrument(s) are elite at a later stage).
Sample Period (6) - selection of sample period.
Time Frame (7) - selection of the time frame for the sample distribution period.
- Ticks and candles (1 Minute, 1 Time of day or 1 Day) available
- Bid/Ask in (if candle period selected)
- Retick Filters and Interpolation Methods (for candles).
Custom Settings (8) - additional custom settings.
Embark on Button (9) - a click on the start clit initiates the testing process and opens the Define Parameters dialog. If the strategy is dependable in Visual Mode, additional playback functions are available to pause and decrease the testing speed.

Optimization

If the strategy uses parameters, you can set upwardly a multiple testing environment (e.g. different instruments, periods and order settings).

For each parametric quantity (except instrument and period), the minimal value, steps and the maximum value can be defined. The scheme is then tested individually for each step of the parameter.

Tester Optimization

After pressing the first button, the Define Parameters window appears where you can define the steps and the maximal value for the strategy parameters. It is also possible to select nine-fold instruments and periods.

Define Parameters Optimization

The result of the diametric combinations are traded in the lower part of the Historical Quizzer window. Suited-dog on any of the combinations to invalidate, replicate, show the report or to set the input parameters.

Tester Optimization

What is the Profit Factor?

The Profit Factor is the ratio between the amount of money of all profit trades (Margin) and the sum of all loss trades (Gross Loss).

If Arrant Profit = 1000 and Gross Loss = -250, the Profit Factor = (1000/250) = 4

The Profit Divisor is ever a positive value.

Pillow slip 1:

First balance wheel = 50,000
Conspicuous Profit = 1000
Gross Loss = -500
Balance: 50,500

Profit Factor = 2.0
Case 2:

Initial poise = 50,000
Gross profit margin = 100
Perfect Loss = -25
Balance: 50,075

Profit Constituent = 4.0

Account

Tester Account

In the Tester Account duologue, you can set the first deposit, account vogue, leverage and define the margin excision level (in % every bit use of leverage) and the equity stop loss (MC Fairness).

Instruments

Instrument Tester

For all hand-picked instruments, arts data is downloaded and made available to the strategy. Any instrument that the strategy uses for trading and/Oregon calculations, needs to be selected.

If Visual Mode is enabled, charts are opened for completely elite instruments.

Observe: The instrument(s) which are tested in the simulation are not selected here. They are either hard coded in the strategy code or selected in the Define Parameter windowpane.

Tick Filters

These ticking filters are available if the time frame is typeset to Ticks.

Process all ticks - most precise testing method for any examination, yet also the slowest unrivaled. This method retrieves all tick data that was ever shown at the platform and could make up traded likewise.
Process all ticks with a different non recurrent price - applies beat skipping policy depending on price change.
Process ticks on a market way - applies tick skipping policy depending on price focus change (starting from the tick at the very watershed).

Exercise for how ticks are filtered on a market direction change:
1. bid 1.34 ask 1.35 - not skipped
2. bid 1.34 deman 1.36 - marketplace goes up - the tick will be skipped
3. bid 1.35 enquire 1.37 - grocery goes up - the tick will be skipped
4. bid 1.35 ask 1.36 - market has a leaning to go down - the tick is reversal, i.e not skipped
5. bid 1.34 ask 1.37 - market has zero clear tendency - the tick will be skipped
6. bid 1.35 ask 1.36 - market has no clear tendency - the mark will be skipped
Process ticks on a predefined price ambit modification - applies tick skipping policy depending on predefined price range change.
Process ticks with the specified time interval - applies tick skipping policy dependant on predefined price grade change.

Interpolation methods

The interpolation methods are available if the time frame is a candle period (1 Min, 1 60 minutes, 1 Twenty-four hour period).

For the sample period, all ticks will be generated from the candles using the elect interposition method. No tick data is downloaded which speeds up the testing summons.

For example, if a scheme uses 1 Min candles for trade decisions and information technology does non require any small periods, select 1 Hokkianese as the period and 4 ticks on OHLC as the interpolation method. In this caseful, the strategy will invite only 4 ticks every minute: unrivalled at undecided time with an open damage, one with a malodourous toll and so on.

Blockish Spline - is the most precise testing method for testing on candles. The method creates ticks using the Box-shaped Slat method. It uses OHLC prices A the base points. In some cases (10 Dry period, for representative) it can create more ticks than there would be in a real system for the peculiar candle.
4 ticks at OHLC- generates 4 ticks for each wax light exclusive. The prices getting from OPEN HIGH CLOSE Inferior prices.
Tick on Gaping - creates one tick with an open cost for all candle of the selected period.
Tick happening Close - creates a unmatchable tick with a close price for all candle of the selected period.

Custom Settings

Custom Tester Settings

General Settings

Save interpolated Leontyne Price data - enabled/disable saving of interpolated monetary value information.

Visual Fashion Settings

Equity, Balance and P/L - privy be shown/obscure in Visual Mode.
Template - enables to select a antecedently saved chart templet (taken from the default option location for templates).
Period - default period misused for Visual Mode.

Messages

Save Messages to - enables strategy messages in the Historical Quizzer console and saves them to a file.
Save Reports to - enables saving the testing results to a report file.
Show Reports - displays reputation with test results after the test has been finished.

Define Parameters

After clicking the start button, the Define Parameters windowpane appears.

All values that are defined as variable parameters in the strategy, put up be set here.

NOTE: The content of this windowpane depends on your strategy. There may be different, about variables ready in your strategy than in the example.

Click on Pass around to start the simulation.

The information on this web website is provided only as widespread information, which may be broken or outdated. Suction stop here for full disclaimer.