Very interesting... I'm in a very similar situation as you just described (i.e. major pairs, stellar results, etc., minimal drawdown). I'm wondering how many other expert groups have developed their own algos like we have. Could we be a part of a new trend? ...and if so, how would the presence of many home-grown algos disrupt the forex trading market, if at all?

Your listed questions, as to what to do next, are the same ones that we are asking right now. It is an interesting proposition since our downside is limited since we can use our algo for our own trading, while the upside is that we could just as easily propagate our success for a fee.

If you'd like to share ideas further, I'd be happy to connect with you over Skype or similar. Send me a message via LinkedIn, if interested. Anyone else reading this message who has developed their own algo and would like to brainstorm as a group is also welcome.

Perhaps we can figure out a solution amongst ourselves.

2. Even institutional investors are often stupid, or better to say, financially innumerate. Once I applied to a big German Asset Manager and as I showed them performance of my "DUCKS" vs. DAX they said: "well, you might have 10 funds and show us only the profitable one".

After this feedback I wrote this review, explaining why my performance could not be a pure luck (I disclosed the complete trading history, don't know whether it is suitable for you)

https://letyourmoneygrow.com/2016/10/29/my-fund-somewhat-better-than-ducks-is-1-year-old-a-festive-but-fair-review/

3. Last but not least, if you can generate 40% p.a. and have $300.000 your own savings, why do you need third-party capital?!

Just let it run for 10 years and all of you are millionaires! :)

Hi David,

though I cannot tell you a "do like me success story" I encounter similar challenges and can share my experience.

1. First of all, nobody will believe you if you merely claim that you have such impressive results (sorry, there are so many cheaters, esp. in FX trading). I don't think you have to waste $10000 on audit but you have to provide your performance statistics on a third-party(!) server.

Like I do it for my portfolio "Somewhat better than DUCKS": https://goo.gl/GZJXW3

Hi Vasliy, first of all thank you for sharing your insights, it seems we are going in the same direction.

ad.1) We are currently working with fxBlue to publish our results, but are waiting for the SAXO Bank integration. Will post them asap.

ad 2) YES .. It is exactly this kind of feedback we are receiving, which led me to this post. We must be knocking on the wrong doors...

ad 3) Well i guess we eventually will be rich... but 10+ years is a long time and we feel a little all dressed up and nowhere to go :-)

Do you know any conferences - tradeshows that would be relevant?

//D

@David - you are facing two different challenges which interestingly enough are on opposite sides of the spectrum. First, it is difficult to even get the attention of asset managers unless you can get an introduction from someone they know and trust. Asset managers are approached all the time with the next big thing. Given that even for the few algos that actually do work as advertised, the returns are often a case of the algo just happening to fit the specific market particularly well. It is very difficult to break through that noise.

And then you have to opposite problem, your returns are in the range that virtually always (not always but a significant percentage of the time) signals high leverage. The sweet spot in returns is in the 10% to 20% range. You may want to consider trading a separate, more conservative version of the same algo, aiming for the lower returns.

I say first you would make a "due diligence" on the strategy, which would imply you need to have an experienced party to evaluate the overall metrics. Which would not involve disclosing your strategy. My commet would be, the metrics you post is not sufficient to valuate the strategy - it takes less than 5 minutes to create 100 strategies having same metrics across 8 months.

@Scott Yes i guess a lot of people try to make the perfect algo that fits all... our experience is that you got to model the market conditions before you configure your algo. (without overfitting) ...(btw not easy :-) Regarding leverage we use a dynamic model taking the individual trades profit projectory into account. (with a leverage between of 5-10), we can then lower the leverage off 2-5 with a ROI of 20% and a AOM target of 50 million USD

@Søren Lanng Sorry i am not sure i understand how this would help?

David some random thoughts from reading your post:

•8 months of live trading is too little, to improve your chances of raising external capital you want to have at least 2 years of track record.

• While I am sure you guys believe in the long term benefit of your method I can tell you practically speaking a 19% monthly drawdawn is massive. At 5% most institutional money will start doubting why they gave you money, at 10% you will have lot of redemptions, by the time you reach 19% whomever is left will want to kill you. Find a way to bring that max drawdown to a single digit number.

•I don't think at such early stage you need to have an audit, especially when both platforms you are using allow you to print out all stats you need to prove to a third party the veracity of the trades. I would invest the 10k in the system.

•If your algo is not scalable it will be hard for you to raise large sums of money from institutions. 10MM is a drop in the bucket, you algo should be able to take at least 100MM, work on making that a possibility.

•If u really have something special try to stay independent for as long as you can, growing and raising assets on your own might be a harder road but much more stable and fulfilling than hot money.

David, at the end of the day i think most so called institutional investors will end up asking you why you arent do this with your own money.. or if you do feel so confident of your algo why wouldnt u be borrowing and investing in your own system. The usual comment from the investing world has always been that the best systems and algos never get marketed ie they are used inhouse for personal profits. There is a logic to creating wealth by OTM (other people's money) but as some one pointed out earlier this is where it becomes running a business rather than making money from a smart trading system. Institutional class systems, back office, legal, compliance etc etc comes into play and it has made most small managers unviable unfortunately. Unless u target raising anything north of a 100mn, going down that institutional route is pointless. Not to pour cold water but if you really think u have found something that works consistently then patience and time will make sure you become a billionaire yourself.. the best of them have done this for themselves so why not you.. good luck

You say it is difficult to explain. Why? It should not be. If someone try to sell you something that pays well and can't explain clearly why and how, it should always be a massive red flag against investing in it.

All the best for your initiative. Btw, bullion ccy pair is XAU/USD and not AUX/USD, AFAIK

@David, as Kishore say, it is always suspicious when someone claim they have a killer strategy, but dont trade the strategy them selfs. 8 months of backtest and less than 200 trades is simply not sufficient to tell anything about the future performance of the strategy. Before you would waste any time or money, you need to perform a "due diligence" on the strategy, if you dont know how to do it, have an expert to do it for you. There is no book on due diligence of a strategy, this is very individual and based on experience. The best proof you have is years of live trading of the strategy. In terms of share of revenue, 20-25% is normal, I doubt someone will pay more, as all know a strategy works till it dosnt work any more.

David, I can not add much to what has already been said. However, live by the motto "the worst draw down is yet to come". So, I would agree to scale back the leverage. What is important to investors is the shape of the equity curve of your cumulated profits/losses. They will then decide what drawdown is acceptable (which will define/limit the deployed leverage).

godspeed !

My apologies David Junge. You did state the 19% drawdown was in your backtesting. Whether it's theoretical or actual, 19% is tough hurdle with the current state of institutional money

@David - those are solid numbers in terms of leverage; you seem to have a good handle on that part of the equation. Drawdowns are always an issue and I have never heard anyone hope for bigger ones but I wouldn't let that deter you. There are all manner of ways to ameliorate those types of issues.

those who have money they can't do profitable trades. but those who have knowledge will not have sufficient money to trades. it's like .. a coach can create one sachin, but he can't play like sachin,

the problem with algos is that first or later they may not work anymore or perform much less. Even if there is no guarantee that what worked in the past will continue working, 7 years backtest and 8 month live may still be a too little timeframe. Beside that, the real holi grail is diversification on a wide range on instruments and strategies and run everything as a portfolio with proper balancing. This is the most difficult goal to achieve.

Hi David,

yes, try to post your strategy online asap.

As to conferences/events, unfortunately I don't know any (and I doubt that there are any). Their enough fairs for retail investors - even in Germany, which has no mature investment culture - but it is not what you need.

I, myself, tried to get an advice from sales guys (all successful funds employ strong salesmen (and saleswomen :))). They were reluctant to disclose the info, which I could understand, but at least they told me that generally one needs at least 3 years of proven track record. In this sense I would say - patience and consistency are our only allies :)

Feel free to email me your trading stats , I will respond accordingly. Thank you .

Please feel free to email me your trading records , I will respond accordingly. Thank you.

Hi - send me your stats and I will provide some detailed feedback. We have come through the same journey as you. Best regards

Hmm... seems a lot of these comments are "post asap"... why would you do that?

how about you start with doing Due diligence on the brokers platform and read FCC complaint, spyware and reverse engineering of your strategy is #1 thing brokers do.. Fortunately your numbers are not so great so you might get lucky and protect yourself against spyware before making your ALGO really worth something int he eyes of investors... BTW IMHO on 80% p/a DD should not go lower then 5%. then investors will ask you to deleverage to deliver 20% with 2%dd. Good luck !

Hi.... you may get in touch with RCM Alternatives......they may be of help ....

RCM Alternatives Main Office

621 South Plymouth Ct. | , Chicago , IL 60605

855-726-0060 (local) | 312-275-8123(fax)

invest@rcmam.com

To be frank,I am surprised you are looking for funds. Most folks with advanced degrees in math, science, and business have jobs that provide significant disposable income. With an algo that effective, make your millions with it (money speaks louder then words) then write a book and/or hold seminars. You will make several more million.

Regarding Maximum Dra Down (MDD) The magnitude of our returns is proportional to risk expressed in terms of MDD. The returns reported here are based on a MDD of 20% of invested capital. , e.g. to obtain an expected return of 63% the investor has to accept a MDD of his capital of 20%. If the investor for instance is only willing to accept, say 10% MDD, then the expected return will be around 31%

The figure below shows the Expected Return on Investment (ROI) as a function of expected MDD in percent of investor’s capital. These returns are calculated using an assumption of no re-investment. The MDD is in terms of the initial investment only, i.e. we assume a constant “skimming” of the generated profit.

http://screencast.com/t/Uq0jgVBzmW

@Ken we are from Denmark... we pay 54% tax and 25% VAT :-)

What is the leverage employed?

David, please get in touch with my partners at United Signals here in Frankfurt, Germany

Not that I'm the first one to mention it, but would definitely be interesting to see some additional data and background info such as currency pairs traded, median duration of positions, p&l distribution,...

@natalia. we have used an average leverege of 10

David, "an average leverage of 10" is very high, especially if you want to trade institutional money. In view of this, the performance numbers are not bad, but not really to get excited about. FYI, our average leverage is just below 2 with a max allowed of 4.

What is more interesting is what is your maximum allowed leverage ? 15, 20 , 50 ???

Hi,

when in similar situation while developing a solution for a cryptocurrency based solution and later FX, Brian Peterson helped with advice a lot. You can contact him via his home page - http://www.braverock.com/ or, better, on #R-finance. Also I think people on Wilmott forums give a lot of actionable advice.

" we have a winning algo !" is such a wrong statement to make

Check out quantopian or numerai and get paid

Our online platform (Collective2) provides strategy managers and developers a walk-forward track record, hundred of statistics and data points, and exposure to investors via a subscription model. Disclosure: I work for Collective2. More info: https://trade.collective2.com/become-c2-trade-leader/

Quite the thread! Thanks for the many insights. I'll lend my two cents. We also have an algo looking for a home aside from our own. I'd be interested in anyone's reaction to our approach. It's 100% verifiable daily by its same users, hence no need to audit or suspect foul play. It is strategy agnostic, which means it currently runs 8 proven strategies simultaneously. We plan to add many more to smooth out the probabilistic 'sine curve' performance syndrome. You might be able to guess a strategy but it would be really hard to master all eight plus their interwoven permutations. It is asset agnostic. It machine-learns its way to optimize any assets performance in much the same way a voice recognition program works. I've just mentioned a few items that we have addressed. Feel free to poke holes at our approach. Thx to everyone for a great discussion!

Thank you everyone for your contribution... we already got great help and are looking into new opportunities. Feel free to pm me anytime. We will publish our results on FXblue (when saxo updates their integration)... thank you and happy trading

I think that the algo must dynamic for each pair. The hedge fund have a huge of scenario for many expert and automatic trade.

Startup the business venture by your own power. Most obvious path is to acquire licences( FCA, CYSEC, OFSHORE - my advice) for investment fond, or investment company, or even a brokerage , which is not to complicated to acquire, and create straightforward web presence where you will explain who you are and what you offer for investors. Most of investors dont care about tech behind the scenes, what they care about is profit. Bigger investors dont ask for 10-100% a month, they want a stabile 2-5%, without huge volatility on account.

Develop smart marketing and flexible licence which will allow you to board as much as many clients you can handle. Your development path should be 6 month, after this time you will attract attention of bigger sharks.

Financial industry is hungry for a new automated strategies(banks included). Audit from Deloitte, EY, PWC, KPMG is welcome and it will increase the price of your product - but is not necessary on start.

If you're confident about 40% in 8 months why not just compound the results and keep it to yourselves?

Hi,

If you have something working, with huge leverage avalaible on fx, you don't need to share something with any seeders. Build your own boutique and spend your pnl to build your infrastructure. This is the path I choosed since 4 years and its more suitable to concentrate on building new strategies than maintaining relationship with investors and audit and rules....

SIBI Alexandre D., Thanks for your insights. I'm in a similar situation with a recent winner algo looking for a home. Originally I designed it for my own trading but know myself well enough that unless I am held accountable, I tend to cut corners. Hence, I am launching a subscription service to keep me honest. The outcome so far is working. I've created a far better product and the feedback from my beta participants, useful. I'm at a crossroad though. The service is growing by word of mouth. I'm considering going the EA route. Starting my own fund sounds counterproductive. What are your thoughts? Anyone else, please feel free to chime in.

Add usdjpy as most players are there only

David Junge and Tom Kadala thanks for generating content on this topic. I am in a similar situation in developing a trading strategy. This post has helped bring up some great points and new ideas. Will reach out directly.

Chris - Thanks! I'm amazed at the number and depth of responses on this thread. Kudos to David Junge for blazing out this trail for us. Due to this thread, I've met some like-minded types and can now look forward to exchanging ideas. That's a potential game-changer.

Some may disagree with me, but the forex trading space harbors too many big egos for what is essentially a one-sided, binary play in a 5 trillion dollar market. At some point, we should all realize our collective asset value, if we worked together more often. This thread is an excellent start!

David, I would suggest on learning it from the experience. Either go work for a hedge fund that does similar things and find out how the big guys do it or get involved somehow with the people who dealt with it. What you described is a typical hedge fund with a certain niche strategy, therefore you need investors that are interested in trading such asset class with such risk. Having a real track record wouldn't hurt as well, but it must be registered on one of the platforms not your personal account because it's not professional and hard to verify. Basically you got a product and now need marketing people to sell it, you need connections or some popularity (which grants you connections) good luck

I would recommend getting in front of real allocators. I'm surprised your own brokerage has not contacted you already, but you can actually start with your acct manager and see what they say. That much money should be deposited with a 'true' Prime of Prime (AdvancedMarkets, Global Prime Au, Lmax, etc) and then you can contact your account manager and they will introduce you to investors who are ALWAYS looking for traders. I also have a few contacts as well with people who have allocated over the past.

The hardest part is attempting to build a track record to pitch and then having to start ALL OVER again when you realize that the broker or asset manager you were working with wasn't worth a damn to begin with.

Hi all. Again thank you for all the interest and support. We are currently regrouping and all the interesting opportunities this great community has given us... I will get back in the new year with more information :-)

A SIGNIFICANT concern is "we’ve found that when approaching traditional investors and HNWI’s it seems difficult to explain to them what we are doing and why it works." Resolve this hurdle first, then do the marketing. The issue is related to the risk adversion of MONEY.

numera.ai

Good point... John Burchfield We've addressed these concerns by offering 100% daily validation of signals by the 'crowd' of users followed by a 'real-time' machine learning component that combines industry accepted strategies with correlated parameters that accurately measure market sentiment. The key component is applying probabilities to determine, which signals to use. At the end of the day, what matters is sustainable profits/growing the account but a sound approach can go a long way to help investor buy-in.

Interested to try the algo and help it market in Australian market

How about maybe joining a auto trading broker? You've probably heard about this, but this will replicate your own trades on connected/following accounts. It's maybe a good way to take the next step without all the paperwork.

Hej David

I like your story, because is similar to mine, I would like to share it to help you

I am European, living in the US. I moved 2 years ago to build a Hedge Fund; after work more than 10 years in algo trading (Futures market - Scalping) with my robots with above 85% successful, Return on Initial capital above 200% (Yearly) and DD below 20% , my thoughts were "if I manage/trade around $10M, in five years I will become rich"

I was naive, retail investors prefer "a brand man from Wall Street", my robots and my "real backtracking" it was not enough. In the same way for all HF that I contacted, they are not interested in a small developer of code (this is the name for us. For these companies we are not "experts" managing money) Prop Trading and similar companies were the same outcome.

...........................................continue next post.....

So, I gave up to try convince to "big players" and I continue to manage my money, running the robot only for 1 hour, it is enough to get a good salary, that is my goal. Good salary, more time for doing another business and more leisure. (I rent my robots for retailers, but the effort and marketing investment, they is not worth)

I wish the best for your goal, but my experience says that 8 month is not enough for running a robot, backtesting is good for you to understand your algo, not for retailers or big companies.

If you believe in your algo, try in the same way like a new company, FFF (friends family and fools) to raise money, build "a European financial company that allow manage money". I don't know laws of your country, but when I was looking for a solution in Europe, in London or in Luxemburgo you could rent a legal structure to manage FX / Futures. This is your real "backtracking" to become a big player.

..........................................continue next post...

If you need more help, some contacts, etc don't hesitate to contact me

Good luck / Held og lykke

The future of small traders have to change, big players ignore us, and this situation it gives us wings to revolutionize the market and to be the next "UBER"

Hi Albert thank you for sharing.. i will pm you... :-)

1. There is no such thing as a fair split, use high water marks and Low management fees to show you will take only take money when you make money, a 50% performance fee above 20% annualised is more than acceptable.

2. Auditors audit accounts, not algos, if your algo were operating under individual account, not much to audit . Open a simple domestic company, have the algos work for 6 months and pay for auditors to audit the account.

3. Presentation of track reccord is most, if you are to use a short term track reccord, give investors week on week performance numbers. FX markets are quite nice recently, real performance is when everyone is losing and you are staying in the green.

4. As for the structure, any aum under USD 5 m is not worth going for tax efficient structures, running costs for an offshore start as from USD 50k minimum pa including everything basic. And the costs keep increasing YoY.

Albert makes valid points. We have had the same experience. Unless you have connections in the "real" banking world, setting up a decent fund is very hard. If you're not part of the "inside crowd", you won't get any serious interest. Sad but true.

We decided to abandon that route (failed several times) and choose to setup and build our own client base. Both through selling systems/signals and offering a Managed Account solution.

Message me please

You also mentioned that it is difficult to explain what you do. In terms of attracting investors, it is not necessary to try to explain your method, not everybody will understand why your strategy is “based on a stochastic process together with X Y and Z mathematical or statistical principles”. There are few “Standard” terminology that sophisticated investors will require to be informed, make a decision and diversify their portfolio, such as Excess Return, Volatility, Correlation to the Market, Variance and VaR just to name a few; which are basically telling them why your strategy is more attractive than their current alternatives and how investing with you will affect their overall portfolio risk/reward balance.

On the “fair split for the profit”, I believe there is an Industry Standard Rule used by Hedge Funds called the “20/2” which means that a 20% is charged upon performance profit after beating a High-Water Mark in a specific time interval, plus a 2% annual management fee regardless of the performance. This however can change drastically depending on individual cases. I met once an investor who would pay up to 60% of the generated profit.

In case you want to offer your services to Funds which already operate, it is important that you define your strategy and how it will impact their current performance. A Hedge Fund will not be interested in your strategy if it does not fall into their business model. Remember, a Hedge Fund plays different set of rules, where they have to maintain a hedged risk exposure to the market while creating an excess return regardless to market conditions.

If you offer a profitable trading strategy or algo which has a Beta value different than Zero, maybe its other kind of Funds you should be approaching. Maybe pension funds could be interested as part of their diversification strategy. I will attach an extract of my thesis on a research I did upon a Market Neutral Strategy in the FX Market.

On the other hand, something that may help you as a fellow CBS Alumni. Every Autumn/Winter semester, Copenhagen Business School offers a course in Hedge Fund Strategies, you don’t have to be a CBS student to attend, however, fees may incur if you want to take it. I could share the content if you are interested, however, the course closely follows Lasse H. Pedersen’s book called Efficient Inefficient, so I recommend you to get it. Let me know if you are interested and I could give you couple of contacts in CBS to take the course in 2017.

Feel free to contact me if you have any questions.

I think for now, the only thing that you should do is to get as many profits as you can, since it is possible that profits obtained in this year can not be replicated with equal success in the months and years that follow. If the algorithm has already been programmed in Metatrader, you should perform a backtest with a tickdata at 99.0% over years with available data to check performance, stability and quality of strategy, it would be interesting to be able to acquire a copy of this algorithm as an expert adviser or to follow and copy its operations through a managed account or investment fund . Please, I would like to keep informed and to know more detail of this development, my personal email is juanp.agro@gmail.com. Thank you in advance and the best of success!

David, also explore Fundseeder and Quantopian.

get the results audited as it would make investors feel more comfortable with the stated performance

The key is matching the investors with the correct strategy and structure. I know a fair few people looking to allocate funds in to FX from 100k to $2M but it isn't easy at that level either finding someone with a good structure.