Qualifying for a 5%ers Funded Account

Trading Discussion Trade Journals
341

5 trades using the new system today,5 losses, so much for an 80% win rate!

The problem is I need to go back to the backtesting to look at SL in more detail and how big the moves are as there were some potential wins in there, everything did show profit to begin with - see below.

The moves from the 7am breakout were much bigger than for the 8am one which isn’t normal.

I’ve had another look at my backtesting and sure enough the move at 8am is usually (~90% of the time) much bigger than the one at 7am so today is just odd I suppose.

342

Late entries had 1 win from 2 at 1/1.5 so at least that has a profitable day.

343

Hi, Do you use some kind of backtesting software and run a decade’s data through it, for all these breakout systems, so you actually know there’s a real edge there, over a statistically significant sample-size?

Good luck with this one. It sounds good - but many things do, often without a verifiable edge? :neutral_face:

344

No I only tested about 100 trades, so not enough to be certain. I’ll have a look at some back testing software.

1 Like
345

It’s about the single most helpful and useful thing you could do, at this stage, IMO. :slight_smile:

1 Like
346

Any recommendations?

Something I can automate. I’ve used MT5 for backtesting bots over several years.

347

Sorry, not my subject. Second-hand info but I’ve heard a lot of good things and never anything bad about “Forex Tester” and I think they offer a free trial, too (they definitely used to).

1 Like
348

OK I will try to code this as a bot and use MT5

349

OK so I have a working bot, it is profitable but only just over 40% win rate (1/1.5RRR),I need to run it in real time (in demo) as I don’t think it is doing exactly what I want.

Makes a profit every day apart from Tuesday :smiley:

Once I know it is doing what it is supposed to I will try different SL/TP etc.

1 Like
350

Yeah, AI is dumb, it is basically doing WTF it likes :smiley:

I asked AI to tell me what the script did and it isn’t opening a trade only when both breakouts are in the same direction, I told it that several times, so I have now got modified code to test again.

351

FGS :smiley:

ds
ds914×226 10.9 KB

352

Notes for me, test the following ideas;

  1. Compare 7am and 8am breakouts for win rate versus a combination of both
  2. Repeat above but only when going with the trend
  3. Only open trades based on size of range ie omit large/small ranges
  4. Try SL equal to the range or half the range
  5. Test different TPs/RRRs
  6. Try later breakouts (after 8:30 etc)
  7. Trailing SL
353

So, it is profitable for MOST currency pairs over a decade but the returns are so modest not sure if it’s worth progressing, around 5% over 10 years.

It does seem to work best on Mondays, I may see how it goes ONLY trading Monday, I suppose there is some logic to that as pent up demand after no trading days? Funnily enough Fridays also look better than Tues-Thur?

This isn’t working 100% yet so needs refining and SL/TP RRR might make a bit of difference.

354

Main account (5%ers) had 2 winners from 3 today. :stuck_out_tongue:

1 Like
355

The bot is a bit too keen to take trades, so I will tweak the code a bit.

EDIT - done that and wow, much better, 50% profit in a year, I will do more testing!

5%ers - 2 wins and a loss.

Noob - 1 win and a loss

Late Entries - 1 win and a loss

1 Like
356

I probably need it more than you do. I’m not taking nearly enough, at the moment. :sweat_smile:

1 Like
357

It is August, you know?

You don’t have to take many, in August?

Even Quark does less business in August.

2 Likes