Algo Performance Poll

Khaled · Sun, 11 Apr 2021 09:05:42 +0000

Hello Great Community! I wonder if we can organize a poll of actual results of Algos run by the members (Real or Demo accounts) who are happy to share. No need to discolse how much money people are making or investing, just % Win Rate, Risk/Reward, Drawdawn, etc. And for those who want, they can share one or two tips on indicators or methods they use in their best winners (Entry, Exit, Money Management). Backtest with WF/OOS is good but, in my case, actual results have always been surprising (Up or Down) although I’m not fan of curve fitting. The idea is not to compete over results, but just want to see based on actual performance if there’s room to outperform buy & hold of major Indices on the long run. I dive first. Enclosed the result of my live algos running on a Demo account at IG… I whish I had them running on real account… I’ve some losers that I don’t show because of the size of the table. Sharing is Caring ! Thank you

This topic has 34 replies, 12 voices, and was last updated 2 years ago by Eric.

Viewing 5 posts - 31 through 35 (of 35 total)

← 1 2 3

01/09/2022 at 9:08 PM #184980

GraHal

I never noticed … it read fine and the meaning is understood!

GraHal

Participant

Topics: 204

Replies: 8399

Been thanked: 2041 times

01/09/2022 at 9:20 PM #184981

nonetheless

Khaled wrote:

Any thoughts?

What I think is that this is a good example of what I meant by abandoning First Principles; you are trying to think your way into it. There’s nothing wrong with that, it might be a good idea, but the only way to know is run loads of tests on as much historical data as you’ve got and see if your idea holds water. In my experience, after countless hours of BT and demo forward testing, separate values for long and short yield the best result. That will remain my position on the subject until I’m shown otherwise, at which point I shall happily revise my opinion 😁.

Same regarding closing on Friday for fear of a disastrous gap opening on Monday. Lots of people harbour this prejudice (as did I), but it’s v easy to check, here’s the code:

//friday close
fridayclose = 1
if fridayclose then
if opendayofweek = 0 and time = 230000 then
tradesok = 1
endif
if opendayofweek = 5 and time = 210000 then
sell at market
exitshort at market
tradesok = 0
endif
endif

(add tradesok to entry conditions)

//friday close

fridayclose = 1

if fridayclose then

if opendayofweek = 0 and time = 230000 then

tradesok = 1

endif

if opendayofweek = 5 and time = 210000 then

sell at market

exitshort at market

tradesok = 0

endif

(add tradesok to entry conditions)

Run your BT with fridayclose switched on and off and tell me which works better. I’ve never seen it yield any advantage – for the simple reason that, as often as not, the re-open can also gap in your favour. But if you get a different result, I won’t argue with you and happy to give it another chance 😁.

1 user thanked author for this post.

Khaled

nonetheless

Participant

Topics: 83

Replies: 1237

Been thanked: 978 times

01/09/2022 at 9:48 PM #184982

Khaled

@nonetheless, I run a backtest on 100k on NQ and guess what? You’re absolutely right cutting Friday night would have made me lose 13% of my gain on 100k. The point I wanted to make is with Gaps your loss can exceed by far your SL, but I forgot to mention that your gains can also exceed your highest expectations 🙂

1 user thanked author for this post.

nonetheless

Khaled

Participant

Topics: 59

Replies: 314

Been thanked: 65 times

02/25/2022 at 3:09 PM #188832

deletedaccount051022

Great topic, thank you Khaled for starting. Some very helpful and insightful contributions, as is always the case in this forum.

Here is my contribution:

I started in May 2020, completely new to algo trading on a non-professional basis and learnt very quickly that it is a completely different ball game to what I have been used to before. Thanks to this great community I was able to learn very quickly. Everything you need can be found here on this site. Thank you to Nicolas, GraHal, robertogozzi, JC_Bywan and Vonasi for all your efforts in running this site.

I started a couple of strategies on both a 1min and 5min timeframe late May ’20 in incubation and scaled up through to July, adding a few more strategies as I went. After my first year I was up 484% but it was a choppy ride and not one I would wish to repeat. I also benefited from a year of strong momentum which really helped and I was able to leave the aglos alone whilst I invested a lot of time into learning more about coding and strategies in general.

What I have learnt that has really helped;

-The wholly grail is very much about finding strategies that fit your personality.

-Whilst everyone wants to make the big bucks, the path to achieving that goal is more important. If the path includes huge drawdowns, low win rates then it is unrealistic to keep running such a system.

-Not wishing to run strategies that are active in all market conditions (overtrading). Sitting on the sidelines is a legitimate trading decision that is overlooked sometimes.

-Higher time frames fulfill my preference in reduce number of orders but still allow the smoothing out of an equity curve. Also I no longer reoptimize once a strategy is launched like I did on lower timeframes.

-Price action is a leading indicator. Any indicator that derives its value from price over a defined period is automatically lagging. However, they form great tools to filter out and define different market conditions.

-Authors I would recommend: Van K Tharp, Kevin Davey, Andreas Clenow and Ernest Chan to start.

I also trade options manually and use scans which I have coded to create daily signals. The approach I use there has benefited my algos. Alignment on multiple timeframes. This will ensure that you only trade an A+ setup every time and help reduce the number of losing trades to a degree.

Hope this helps.