What is the optimum depth of history for identifying a useful signal? - page 8

 
tester and trade trades are the same
Mathematical expectation per trade reaches 50 pips
Payoff expectation
in January 33.4

when tested since 1999, the full system net profit over 18000 micro lot drawdown is not more than 100 quid a micro lot

take profit at least 1.5 times stop

max stop 100 pips

 
azfaraon:
Tester and trade trades are the same
Mathematical expectation per trade reaches 50 pips
Payoff expectation
in January 33.4

when tested since 1999, the full system net profit over 18000 micro lot drawdown is not more than 100 quid a micro lot

take profit at least 1.5 times stop

max stop 100 pips

Do you use indicators?
 

I'm not interested in the chart readings when entering the market... I only use the price chart when setting a stop and take out.

As for the off-topic, I got here the right way ... Because from my point of view I found the maximal depth ... For example I will make an opt optician built-in ... The obligatory condition is that the advisor should have a stop and takeaway ...

 
Why beat around the bush ... the topic is yours and you want to be in charge, so tell me not to write here ...
 

Strategy Tester Report

Moving Average

(Build 765)

Symbol

EURUSD (Euro vs US Dollar)

Period

4 Hours (H4)

Model

Every tick (the most accurate method based on all available minimum timeframes)

Parameters

Lots=0.1;

Bars in test

1489

Ticks modelled

1210061

Modelling quality

n/a

Mismatched charts errors

277

Initial deposit

1000.00

Spread

Current (1)

Total net profit

658.89

Gross profit

909.16

Gross loss

-250.27

Profit factor

3.63

Expected payoff

50.68

Absolute drawdown

102.50

Maximal drawdown

275.92 (20.84%)

Relative drawdown

20.84% (275.92)

Total trades

13

Short positions (won %)

9 (55.56%)

Long positions (won %)

4 (50.00%)

Profit trades (% of total)

7 (53.85%)

Loss trades (% of total)

6 (46.15%)

Largest

profit trade

342.78

loss trade

-65.46

Average

profit trade

129.88

loss trade

-41.71

Maximum

consecutive wins (profit in money)

4 (286.99)

consecutive losses (loss in money)

3 (-125.31)

Maximal

consecutive profit (count of wins)

431.08 (2)

consecutive loss (count of losses)

-125.31 (3)

Average

consecutive wins

2

consecutive losses

2

 
it's optimisation
 

StrategyTester Report

Moving Average

(Build 765)

Symbol

EURUSD (Euro vs US Dollar)

Period

4 Hours (H4) 2015.01.06 00:00 - 2015.01.30 08:00 (2015.01.06 - 2016.01.01)

Model

Every tick (the most accurate method based on all available timeframes)

Parameters

Lots=0.1

Bars in test

1111

Ticks modelled

283171

Modelling quality

90.00%

Mismatched charts errors

0

Initial deposit

1000.00

Spread

Current (1)

Total net profit

233.63

Gross profit

281.63

Gross loss

-48.00

Profit factor

5.87

Expected payoff

58.41

Absolute drawdown

77.23

Maximal drawdown

325.23 (22.11%)

Relative drawdown

22.11% (325.23)

Total trades

4

Short positions (won %)

4 (75.00%)

Long positions (won %)

0 (0.00%)

Profit trades (% of total)

3 (75.00%)

Loss trades (% of total)

1 (25.00%)

Largest

profit trade

145.47

loss trade

-48.00

Average

profit trade

93.88

loss trade

-48.00

Maximum

consecutive wins (profit in money)

2 (153.47)

consecutive losses (loss in money)

1 (-48.00)

Maximal

consecutive profit (count of wins)

153.47 (2)

consecutive loss (count of losses)

-48.00 (1)

Average

consecutive wins

2

consecutive losses

1

and these are the same parameters from the beginning of the year. i.e. the optimal depth that is used to optimize the Expert Advisor ... We can do all this with any Expert Advisor.
 
I took a piece of history with optimal depth from my point of view and optimised it, and then tested it with the same parameters for the current month
 
optimization scheduletest Here are 2 graphs the top one is optimisation, the second one I made a test taking into account January of the year
 
ZaPutina:

Well, that is, you have chosen the depth of history at which the optimization parameters change more slowly in the sliding window than in the current month. All of this can be put into a normal predictive function.

Well, if you know, why did you open the topic?)))