2 algos that I wrote... | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Contacting owner who Wrote MT4 to NST Script 7 replies

Machine Learning Trading Algos 10 replies

How to Make Money on Tested Autotrader Algos 0 replies

Trading news events with banks and algos 15 replies

Anyone know who wrote the lsma3? 1 reply

Attachments: 2 algos that I wrote...
Exit Attachments

2 algos that I wrote...

  • #1
  • First Post: Edited 9:35am May 20, 2018 8:48am | Edited 9:35am
Hello guys,
I am looking for advice.
I am a full time trader working on the forex, and the future and options french markets. I live...in France
My trading giving me a lot of free time (I am not a heavy trader) and having computer science background, I spent the last 2 years modelling and coding 2 of the strategies I use with the EUR/USD pair.
Those 2 "EAs" are now completed and they have been running 5x7 in one of my accounts for the last couple of months. The yearly return is in the 15% range and I built them with the core idea to protect my capital, so very tight protection systems are included in the code. Basically my philosophy has always been "before winning try to not lose". I also built a small website where I introduce them and publish their live performance.
Today I am thinking of making them more..."commercial" and this is where the advice part at the begining of his thread comes in.
Indeed, I have no clue if such software or signals derived from my algos, or managed third party accounts with them, could be of interest to people and where I should start, who I should talk to, etc....
So if you have serious recommendations to make, ideas to throw, even if you would like to try a pilot to assess my work, well anything relevant, please let me know.
In the meantime, have a nice day.
Ben.
  • #2
  • May 20, 2018 10:49am May 20, 2018 10:49am
If you have figured out an edge that can make you 15% annual returns safely I can think of the following ways to convert it into real $$$

1) You sell your EAs. You will be required to prove the performance of the EA based on back test and front test and some real live performance data.
2) You rent your EAs. Same conditions.
3) You invest other people's money. Depending where you live and operate from, there will be regulations to follow and customer base to build. This takes time and lot of hard work. But if you are thinking long term. This is the best way to build a lot of wealth.
4) You start a own signal service or use existing one, where your clients can duplicate your trades paying you some fees.


The above said, 15% ROI annually would not be interesting to most people here. But I am sure there would be few still be interested (myself included) but it depends on how did you conclude the "15%" What kind of testing have you done? Have you traded this live? What % of draw downs did you see?
  • #3
  • May 20, 2018 11:26am May 20, 2018 11:26am
Quoting pipsify
Basically my philosophy has always been "before winning try to not lose"..
//------

and there is the key to what is said can't be done............

if(! ) {then} .......h
to trade and code, keep both simple... no call to impress....h
  • #4
  • May 20, 2018 11:32am May 20, 2018 11:32am
I think if someone truly feels like they've figured out a consistent edge, selling that edge in any format would be the last thing on their mind.

Before presenting any product for sale you'll need to gather live forward testing performance data. The reason for this is that most strategies fail after a short period. Counter trend strategies are likely to fail sooner than trend strategy.
  • #5
  • May 20, 2018 1:10pm May 20, 2018 1:10pm
Hi Yalgaar,
thanks for sharing your views.
Actually, having thought about it myself my focus would be on 2) (rent my algos) and 4) (provide signals).
Regarding 3) I would be thinking of working with established fund managers that would like to add my algos to their existing arsenal, but I'm not sure 15%/year would catch their attention, and furthermore I do not know people who could help me with meeting such guys.
This beeing said :
- For 4) it looks like the french regulation is so tough that I would need to pass many certifications to be allowed to provide my own service, and I'm not eager to do that. That's why I'd like to be able to find partner(s) with existing signal services that would be interested in piloting my stuff. And I'm not talking about guys like zulutrade but about real pros selling signals.
- For 2), my algos are currently written in LUA (technological choice), the language used in FXCM Trade Station II platform, and I have not ported them to MQL4 yet, plus I am not sure I'd like to do it because MQL4 is way lesss secure than LUA and it looks like MQL4 EAs can be cracked pretty easily. However I can (and do) easily push signals from Trade Station II to any MT4 account at any broker, so my EAs could be rented this way, pushing signals into MT4 accounts.

Basically, whatever model we talk about, my main issue is my lack of connections to companies, individuals,...that I could talk to, people who are already in that business. So this is probably where I would need most help.

As per your questions, I'm no a big fan of backtesting. I did some and perf are good but I'm more on forward testing, so :

1) I let the algos run for 1 year in a demo account to be able to work on possible bugs and validate them.
2) Since end of march 2018, they are running in one on my live accounts 5x7.
3) DD is very low, and the algos are designed to keep it this way (2.7% max as we speak)

And finally I realize that I talked about my website in my first post but did not provide the URL. Here it is : www.malori.org
Here you will find a description of the algos, the live perf and more.

Eager to find a good way forward, I hope this thread and your comments will help out.
  • #6
  • May 20, 2018 2:28pm May 20, 2018 2:28pm
Quoting BLSH
I think if someone truly feels like they've figured out a consistent edge, selling that edge in any format would be the last thing on their mind.
Well yes and no...if you do not have enough capital to make a living out of 15%/year...then selling your edge makes sense...at least the signals coming out of it.
  • #7
  • May 20, 2018 2:52pm May 20, 2018 2:52pm
can you post the link to your website ?
  • #8
  • May 20, 2018 3:02pm May 20, 2018 3:02pm
Quoting MXT
can you post the link to your website ?
www.malori.org
  • #9
  • May 20, 2018 3:13pm May 20, 2018 3:13pm
Quoting pipsify
{quote} www.malori.org
thanks Pipsify , appreciated
  • #10
  • May 20, 2018 3:21pm May 20, 2018 3:21pm
Quoting pipsify
Hi Yalgaar, thanks for sharing your views. Actually, having thought about it myself my focus would be on 2) (rent my algos) and 4) (provide signals). Regarding 3) I would be thinking of working with established fund managers that would like to add my algos to their existing arsenal, but I'm not sure 15%/year would catch their attention, and furthermore I do not know people who could help me with meeting such guys. This beeing said : - For 4) it looks like the french regulation is so tough that I would need to pass many certifications to be allowed...
simple Q
how do u know that it will work in d future w/o backtesting
I dont mean that backtesting guarantees future results but gives a hint [w/o optimization]
WaBaS Where are Buyers and Sellers
  • #11
  • May 20, 2018 3:33pm May 20, 2018 3:33pm
Quoting mr.brown
{quote} simple Q how do u know that it will work in d future w/o backtesting I dont mean that backtesting guarantees future results but gives a hint [w/o optimization]
I did some backtesting (last 10 years) w/o optimization. When I'm saying I am not a big fan of backtesting that's because you can always make the backtest look the way you want with proper optimization. So for me the real proof of work is forward testing but backtesting is of course part of the whole story.
  • #12
  • May 20, 2018 4:06pm May 20, 2018 4:06pm
Quoting pipsify
{quote} I did some backtesting (last 10 years) w/o optimization. When I'm saying I am not a big fan of backtesting that's because you can always make the backtest look the way you want with proper optimization. So for me the real proof of work is forward testing but backtesting is of course part of the whole story.
do u have BT report(s)?
WaBaS Where are Buyers and Sellers
  • #13
  • May 20, 2018 4:10pm May 20, 2018 4:10pm
i've noticed your system is very conservative

you call them scalping systems , but they actually trade quite conservatively from what i've seen , trading 1 - 2 times per day - trading 1-2 times a day is not scalping but swing trading rather

they scalp 2-3 pips but they also take longer trades . like 155 pips from what i've seen

so your systems are both scalping and swing trading ?
  • #14
  • May 21, 2018 2:40am May 21, 2018 2:40am
Quoting mr.brown
{quote} do u have BT report(s)?
Not at hand cause I stopped backtesting more than 1 year ago and did not keep my reports, but I can generate a couple if you want (in TS II format).
  • #15
  • Edited 4:08am May 21, 2018 2:47am | Edited 4:08am
Quoting MXT
i've noticed your system is very conservative you call them scalping systems , but they actually trade quite conservatively from what i've seen , trading 1 - 2 times per day - trading 1-2 times a day is not scalping but swing trading rather they scalp 2-3 pips but they also take longer trades . like 155 pips from what i've seen so your systems are both scalping and swing trading ?
So...
ApaChE is a pure scalper, but not a mad one, it does enter the market only when the right conditions are there...so 1-2 trades per day on average. Scalping (for me) is when you do not stay for a long time in the market (not more than a few minutes) and try to get just a few pips.
KproteKT is more a day trading algo. In some very rare cases it will hold a position for several days trying to take a big chunk of the move . As a matter of fact for the last 2 months that "rare" situation occured several times and that is why the performance of my live account is already at 15% after 2 months whereas those 15% should be the annual performance.

Hope this helps.
  • #16
  • May 21, 2018 4:29pm May 21, 2018 4:29pm
Quoting pipsify
{quote} So... ApaChE is a pure scalper, but not a mad one, it does enter the market only when the right conditions are there...so 1-2 trades per day on average. Scalping (for me) is when you do not stay for a long time in the market (not more than a few minutes) and try to get just a few pips. KproteKT is more a day trading algo. In some very rare cases it will hold a position for several days trying to take a big chunk of the move . As a matter of fact for the last 2 months that "rare" situation occured several times and that is why the performance...
gotcha
  • #17
  • May 22, 2018 3:50am May 22, 2018 3:50am
Guys,

still in listening mode for proposals to allow a larger community to benefit from my algos performance.
Please keep this thread alive ;-)

In the meantime, check the live perf at www.malori.org
  • #18
  • Edited 2:43pm May 23, 2018 10:33am | Edited 2:43pm
FYI, we just passed the 17% mark after 2 months of trading in real account.
Attached Image
  • #19
  • Edited 8:33am May 25, 2018 8:19am | Edited 8:33am
Quoting pipsify
Hello guys, I am looking for advice. I am a full time trader working on the forex, and the future and options french markets. I live...in France My trading giving me a lot of free time (I am not a heavy trader) and having computer science background, I spent the last 2 years modelling and coding 2 of the strategies I use with the EUR/USD pair. Those 2 "EAs" are now completed and they have been running 5x7 in one of my accounts for the last couple of months. The yearly return is in the 15% range and I built them with the core...
Hi Ben

If it was me....the major hurdle to overcome is the reasonable but inevitable sceptical outcry from traders regarding the performance metrics.

To make your EAs more saleable and to overcome this stigma faced by any commercial EA provider..... I would be putting them on Darwinex or say FundSeeder where you could be getting benefits from investor allocations as a signal trader and at the same time building awareness from your verified performance to the investment community through that channel. There is no substitute to verifiable performance and Darwinex/Fund Seeder are pretty good in defining worthwhile options.

Once you have the demonstrated track record....it would all follow from there regarding commercial application with a lot less grief.
  • #20
  • May 25, 2018 9:17am May 25, 2018 9:17am
Quoting Copernicus
{quote} Hi Ben If it was me....the major hurdle to overcome is the reasonable but inevitable sceptical outcry from traders regarding the performance metrics. To make your EAs more saleable and to overcome this stigma faced by any commercial EA provider..... I would be putting them on Darwinex or say FundSeeder where you could be getting benefits from investor allocations as a signal trader and at the same time building awareness from your verified performance to the investment community through that channel. There is no substitute to verifiable...
Thanks for the advice, Darwinex is in progress actually, I'll check Fund Seeder.
With the same idea, what do you think of psyquation.com ? Do you know them ?
0 traders viewing now