Community Spirit: Profit Shares for Quants and QC, New Income Streams

Back

I am seeking feedback from admins and the community.

It would seem that prop firms, investment clubs, hedge funds, and non-quant traders are huge untapped resources for the QC community, as a whole, and that quants' work is way undervalued.

It also appears that the SEC has discriminated against poor class retail traders who have less than $25,000, by enforcing an unconstitutional "Pattern Day Trade Rule".

As a quant, I feel that if my algo cannot generate profit for a client, I deserve no payment, regardless of any agreement, and, that a client should not have to risk any fixed fee amount, gambling that my algo will perform for them, 

I also feel that if  a quant, with the help of QC, makes a significant amount of profit for a client, that the quant and QC deserve a fair share of those profits. 

I also feel that, where a quant is willing to risk their own money on a venture using their algos, the quant should have the opportunity to attract new funding sources to QC, and command a profit share.

That is all fair business practice, and congruent with the community spirit of QC, and would attract new profits for both the quants and QC.

In typical proprietary trading firms, new traders pay a security fee and the firm matches that money, 2 times, or more, and the trader trades that money for a share of the profits generated, usually, 60% to 85% of the profits go to the trader, and the firm keeps the rest.

The trader also pays a monthly desk fee of about $200, like the QC monthly premium account fee of $250, or $3000 yearly.

In Alpha Streams, traders , who are connected with institutional funds, are, apparently, paid an arbitrary licence fee for the use of thier algorithms.

It is reasonable to see that a quant, who offers a monthly license of $250 or $1000 for an algo that generates $10,000 or more, profit in a month, has significantly undervalued thier product, and has been underpaid for their merit.

In the QC user agreement, QC asks 30% of all profits generated from the use of QC services, in extracurricular business agreements.

It is unclear if QC derives a percentage of profits from the funding party's profits generated, using the traders' algos.

Alpha Streams' highest prize represents $8000 to the 1st place winner, but there is no mention of a share % in profits generated from the use of the winning algos, for the winners.

$8000 equates to 2.7 years' free premium QC account status, at $250 monthly.

For example, if a trader's algo is used to manage $1,000,000, and the trader's algo generates 25% profit, in a year, that equates to $250,000 profit.

If a trader's share is the propietary trading industry standard minimum of 60%, that equates to $150,000, to the trader, for one year's use of the trader's algo.

At the industry standard high end, 80% of profits, that equates to $200,000, for the trader.

If QC makes 30% of profits generated from all extracurricular business agreements , using QC services, that equates to $75,000, for QC, from the use of just that one algo.

If an arbitrary licesning fee of $1500 per month is paid to the trader, that equates to $18,000 per year, which is $‭3,720‬ less than the poverty level for a family of 3 , in the USA, in 2020..... minus $3000 for the QC yearly desk fee, that is $6,720 less than the poverty line.

In Alpha Streams, there is no apparent means for quants to be identified and contacted, for other opportunities, within, or without, QC, beyond licensing other algos from the same anonymous quant.

In Alpha Streams , it appears that only a very small number of algos are readily identifiable for licensing. 

There appears to be no means for quants to bring new firms and funding parties into the QC community and negotiate business with them, that includes profit shares.

Currently, QC must trust that quants are being forthcoming with commercial accounts, not readily identifiable to QC.

Currently, quants must trust that their algos are not being used to generate profits for others, not readily identifiable to the quants.

It is well proven that trustless, transparent, and mutually beneficial business models are the most sustainable and attractive.

Conficius said, "a merchant will grow grow more wealthy, more quickly, selling 100 people something for one coin, rather than selling one person something for 100 coins."

Poor class citizens , and society as a whole, would benefit greatly by being able to day trade and use algos, amnd they are the largest, most excluded class of citizens.

QC and quants could benefit greatly by giving quants true ,ultimate control over the quants' algos and business arrangements for the use of their algos, and expanding quants' ability to bring in new sources of fnding, that, otherwise, would not be participating with QC, or trading at all.

I designed a money management module that is specifically designed to manage money investment for any budget, and I apply it to high probability rules.
A poor class person puts in all their monthly expenses and incomes, and defines their risk tolerance and goals, and the algo allots accordingly.
Simple algo leasing cannot offer this level of tailoring, currently, as the quant would have to communicate with the client and enter the subjective variable data, for each client, prior to deployment.

For small accounts, a fixed licensing fee presents a mathmatical wall , higher risk, and disincentive, where a percent share of profits generated does not.

Because QC and the quants do not offer any investment clubs, or other means for non-quants and poor class citizens to participate, an entire class of traders is excluded from pattern day trading, and a significant recurring income stream class for QC ,the quants, and the commnity as a whole, is neglected.

Because QC does not allow quants to share in profits generated, quants remain underpaid for their merits, and there is very little incentive for quants to devote themselves to QC for the long run.

Just like treasure hunters enjoy thier work, and are looking to strike it rich, so are QC, the quants, and licensees.
Everyone wants to be paid very well for their work and contributions.
A "community spirit" wants others, especialy those who facilitate their success, to be paid very well for their work and contributions.
Anyone who says any different is not being truthful.

All this being said, I would like feedback from the community and QC admins ,regarding the following proposals:

1. I propose that QC create a protocol and system that allows quants to bring in new funding sources, like proprietary trading firms, hedge funds, individuals, and investment clubs, and negotiate , or present, pre-approved contracts, where profit shares for quants and QC are requisite, fair, and competitive in the market.

2. I propose that QC allow quants to command profit shares, of total profits generated, from licnsing the quants' algos, in which the quant and QC would share, and not limit quants and licensees only to licensing fees, for the use of algos. It is reasonable that QC and the quants share an equal percentage of the profit-share generated. Neither QC, nor the quants, can operate without eachother.

3. I propose that QC create a system that showcases each, and all, quants' algorithms ,that each quant wants to present for negotiation, where algos may be searched by quant's name, intended account size, backtest report metrics, investment horizon, risk tolerance, vehicle/market type, industry, sector, etc etc, and a means to communicate with the quants.

4. I propose that QC enforce a clause in all quants' fixed fee licensing agreements, where , once a defined percent profit of starting capital is exceeded, the current recurring fee is waived and the quant and QC are entitled to a royalty percent of profit gnerated there-after, which is competitive in the market, that serves both QC's and the quants' long term interests and well-being.

5. I propose that QC develop and offer an investment club system, to allow poor class citizens exposure to day and algo trading , and allow quants to form their own investment clubs, to meet the $25,000 minimum requirement, where QC and the quant share equally in the profit-share generated.

6. I propose that QC develop a "Tip Jar" routine for QC, the Quants, and the licensees, where any party may "tip" another party in appreciation, for their contributions. This could help QC pay for new features requested by quants, and foster good, lasting relations between all parties. Quants can also tip other quants and admins who help them with their algos.

Update Backtest







 
0

The material on this website is provided for informational purposes only and does not constitute an offer to sell, a solicitation to buy, or a recommendation or endorsement for any security or strategy, nor does it constitute an offer to provide investment advisory services by QuantConnect. In addition, the material offers no opinion with respect to the suitability of any security or specific investment. QuantConnect makes no guarantees as to the accuracy or completeness of the views expressed in the website. The views are subject to change, and may have become unreliable for various reasons, including changes in market conditions or economic circumstances. All investments involve risk, including loss of principal. You should consult with an investment professional before making any investment decisions.


Hey Cary!

Firstly thank you for taking the time to post such a thoughtful post. I appreciate its spirit and goals. We started QuantConnect in 2012 with a plan very similar to your post above and ran headlong into reality as all start-ups do. Before I get to the issues, let's start with the things we can do from your post:

1. Referral System - Awesome yes we should have it and will build it in the next couple of months.
6. Tip Jar - Love it. Yes we'll build it in the next month or two so people can earn tips/contributions to their QC-Wallets. 

Now for 2,4,5: The USA does not permit any form of revenue share unless you are an institutional investor, or are registered as a financial advisor. Aiming to make revenue shares available (anyone <-> anyone) is unfortunately not legal at this time. There are some combinations that are allowed: (institutional investor -> anyone), (registered financial advisor -> anyone). We are a Delaware C Corporation and covered under US Laws. 

The next restriction is a practical one; revenue shares with institutional investors require the quant trust the institutional investor is only investing what they claim and limits the use of the algorithm to simple allocations. If you're using an algorithm as a factor, making a tiny contribution to a 1-million factor model your revenue share will be extremely small. QuantConnect elected to use a simpler subscription model which provides the same revenue potential with cleaner licensing. 

For 3: This is Alpha Streams. Due to US laws it's restricted to Institutional investors at this time. We'll try and lower the barrier here as far as the law will allow but likely that means an accredited investor ($1M net worth, $250k income). Practically we can't showcase all algorithms as there are fixed hosting costs that have to be paid.

We'll keep bucking and fighting to give the individuals in our community access to global funding, and world-class infrastructure. Thank you for supporting the community and being a great part of it!  

3

The material on this website is provided for informational purposes only and does not constitute an offer to sell, a solicitation to buy, or a recommendation or endorsement for any security or strategy, nor does it constitute an offer to provide investment advisory services by QuantConnect. In addition, the material offers no opinion with respect to the suitability of any security or specific investment. QuantConnect makes no guarantees as to the accuracy or completeness of the views expressed in the website. The views are subject to change, and may have become unreliable for various reasons, including changes in market conditions or economic circumstances. All investments involve risk, including loss of principal. You should consult with an investment professional before making any investment decisions.


Jared,

Awesome!

Thank you.

I would like to respectfully bring to your atention some legal issues you may not be aware of.

Investment ckubs are not regulated by the SEC noe FINRA.

Please see: https://www.sec.gov/reportspubs/investor-publications/investorpubsinvclubhtm.html

1. Software programmers, system technicians, and network administrators do not make financial decisions for anyone. With regard to investment clubs, the members vote on execution triggers and allocation. Under  FINRA 1230.(a) Associated Persons Exempt from Registration, programmers, technicians, and network administrators are exempt from FINRA registration and regulation, especially if they are vested  members of an investment club.

2. Investment clubs vote on compensation for it's officers and employees, as well the finincial decisions of the club, and this activity is unregulated, as it is protected private activity, under the 1st and 4th Amendments.

3. All tax liabilities pass through to each member's individual taxes. 

4. At worst, some states require that investment clubs register as partnerships, which retain the same protections from unlawful regulation and taxation, and the club may register in any state ,any of it's officers are located in.5. Where registration is required, and yearly reports are requisite, those routines can also be automated.6. Where membership is free ,and the club does not offer it's own shares of stock for sale to the public,  the SEC has no jurisdiction, as free membership does not represent any security. See  Investment Company Act of 1940.

6. Following the Ethereum smartcontract model, all proceedings and binding obligatons can be attended in the code itself, and all meetings and communications done virtually, commencing under 15 U.S. Code § 7001. General rule of validity: Electronic Documents and Signatures. I can send you an example.

7. Where any startup seeks crowdfunding from the public, or it's own members, and does not issue shares of it's own, they are exempt under 15 U.S. Code § 77d.Exempted transactions. This was part of the crowdfunding acts.

8.  An investment club is a "person" in the legal sense, and may, itself, join other investment clubs. Thank George Bush senior for that one, not that I am a fan. See 29 US Code § 53 - “Person” or “persons” defined.

I will stop here, and simply say, that, investment clubs are a feasible alternative to financial advisory.

Itis only a matter of how one structures and executes transactions  and agreements and how decisions are made.

For example.....

A quant and QC want to offer a club.. They are the first 2 members of the club and are the first officers. They invite parties into the club and offer a beginning algo as the club's first execution system. The trade plan is submitted to all members ,who then agree to the signals and execution and allocation variables, and contribute whatever amount to the club's pool they wish, and the program manages the money and executes trades under rules approved by the members.  Members may submit issues for votes at any time. The members approve the officers' seats.  No party is obligated to remain a member if they do not agree with the decisions of the club members, and may be canvassed for membership in better suited clubs, relative to their goals and risk tolerance.

Writing an investment club algo entails the algo,  the smart contract and money management routines, as well the signup, interview, elections, meeting minutes, and yearly reporting roitines, as well seting up a forum for virtual meetings., using google groups . 

I will let you chew on this and message if you'd like me to send an example.

0

It is definitely interesting and I'll ask our lawyer next time we meet. From your link: "To qualify, an investment club: must not make, nor propose to make, a public offering of its securities; and must not have more than 100 members.". I'm pretty sure there's no hack around it and QuantConnect would need to register as a fund and meet all the compliance needs associated with that. We chose to not do it as its a burden on our execution, and instead, we focused on providing the best service to the community -- and second, it opens us to conflict of interest acquisitions from those worried we will appropriate their IP. That said there's nothing stopping you from making an investment club =).

2

The material on this website is provided for informational purposes only and does not constitute an offer to sell, a solicitation to buy, or a recommendation or endorsement for any security or strategy, nor does it constitute an offer to provide investment advisory services by QuantConnect. In addition, the material offers no opinion with respect to the suitability of any security or specific investment. QuantConnect makes no guarantees as to the accuracy or completeness of the views expressed in the website. The views are subject to change, and may have become unreliable for various reasons, including changes in market conditions or economic circumstances. All investments involve risk, including loss of principal. You should consult with an investment professional before making any investment decisions.


Hey Cary,


thanks for your contribution! Although I don't agree with everything you said, I think you made some important points that I would like to take up in the hope that Jared will respond to them. 

Your suggestions concern the business model of QuantConnect (profit sharing instead of monthly licensing income in the sense of a subscription model).

Personally, I think the QuantConnect business model is great and I see many advantages in the approach used. 

For example, a Quant can generate a regular and steady stream of income in this way.  At the same time, institutional asset managers have a clear basis for calculating their costs. Moreover, they do not need to disclose their finances completely, as would be the case if remuneration were based on profit sharing or a performance-related fee. 

The strength and advantages for institutional asset managers in using QC Alpha Stream market, I believe, is not only that they can find an Alpha Stream that offers huge returns, but also that they can reduce costs enormously. A fund can use QC to manage its AUM at a fraction of the usual cost by subscribing to one or more corresponding Alpha Streams that meet its requirements and investment philosophy.

The 30% required by QC, to my knowledge, does not relate to the profits the funds make with their AUM, but to the subscription income. So if a fund subscribes to an Alpha Stream for $1000 per month, QC earns 30% of that, which in this case is $300 in that one month on that particular Alpha Stream. In particular, it actually costs the fund $1,300 per month. Can you confirm my statement, Jared?

You pointed out that the payment as a quant is worse with the existing model, but that was just an example. It should be noted that an offered Alpha Stream can be subscribed to by several funds at the same time. But apart from that, I agree with you that $250 and even $1000 per month is very little and no one can make a living this way.

However, I do not see the solution in a change of the business model, but in the following two measures: 

    1.) Better and fairer pricing of Alpha Streams

    2.) More funds (larger customer base) for the Alpha Streams Market

Allow me to elaborate on the two points below. 

To item 1.) : A market is only reasonably efficient and capable of fair pricing if there is competition on both sides (supply and demand, more on this in point 2.) and the market itself is transparent for all market participants. 

I think that there is room for improvement in terms of market transparency. For example, the quant community is currently not able to see all offered Alpha Streams in the market and therefore you cannot see the prices of competitors for comparable products /Alpha Streams. At this point I also ask myself why only a small selection of Alpha Streams are visible to us in the market? @Jared: Can you comment that, please?

In addition, I think that it can be helpful for everyone if QuantConnect develops a kind of guideline for the pricing of Alpha Streams. The documentation mentions some helpful aspects, but in the end, the developer should be able to judge if he is charging a fair price. 

QuantConnect could suggest additionally a price when submitting an Alpha Stream based on backtesting metrics such as PSR, scalability, correlation to other Alpha Streams, accuracy, number of insights generated, average value of an insight, etc. 

Currently, a default price of $100 (respectively $250 for exclusive) is initially suggested for the license. However, the performance metrics of the Algos are not considered here. 

A good solution would also be, for example, if QuantConnect suggests a price or a price interval after the Alpha Stream has been Live on the market for at least 3 months. Then the proposal would not only be based on backtesting results, but on performance in the live environment as well. In this case, the Live Sharpe Ratio could also be included. 

What do you think about this?

For me, transparency also means that the community can see how many Alpha Streams are currently subscribed to and what the metrics and prices of the offered Alpha Streams look like on average. 

Last but not least, transparency also means that you know the needs of your customers well and can put yourself in their shoes. What type of Alpha Stream are the funds looking for that may not yet be available on the market? How far should an Alpha Stream be scalable and what percentage of their AUM do they intend to allocate to a specific Alpha Stream? Maybe QC could conduct interviews with the clients and publish them here on the blog? The interviewee can remain anonymous if he/she wishes to do so. 

Now I finally come to the second point. 

To item 2.) :  When I registered on QuantConnect about a year ago, there were 5 trading firms with an aggregated AUM of 10 billion. Now there are 11 trading firms with an aggregated AUM of +$ 65 billion. 

First of all, congratulations to the QuantConnect team. This is a strong performance!  :-)

In other words, if QuantConnect would be a pubicly listed company, I'd definitely buy some shares. ;-)

It is in the interest of all parties involved if QuantConnect can attract more funds and expand its customer base. This not only increases the chances of getting subscribers for an Alpha Stream, but also generates additional income for QC. 

The way I see it, QC is currently focusing on quantitative funds. Why not include the whole range of asset managers as an addressable market? All that would be required is to add a user-friendly GUI to the existing API, so that non-quantitative funds would also be able to benefit from the Alpha Market. 

Basically, I am convinced that the market is much larger - even if you only consider the quant funds in the USA. That means there is still a lot of potential ahead of us. 

What does the process of client acquisition actually look like for QC? I'm just curious. :-) Let us know if we as a community can somehow help you to attract more institutional investors to QC. 

As you can see, 1) and 2) are connected and have a positive interaction with each other. 

 

So, this has become a long text, but it was worth it. 

I am happy for every feedback. :-)

 

Side note: As English is not my first language, there is a possibility that some of my passages are linguistically not correct. If this is the case, just ask me and I will gladly explain it in other words. 

2

Thanks for the detailed post Arthur! Its been a crazy week please forgive the brief reply:

> In particular, it actually costs the fund $1,300 per month. Can you confirm my statement, Jared?
Correct, the Alpha Stream model passes costs to the funds licensing the model. Alphas were not intended to be a single mission you work on for months, but more of a library of factors you could make in less than 10 hours - so $1000/mo for 12 hours work is a solid living =). An Alpha doesn't need to perform well all the time, just not incur large drawdowns. 

> Alphas Visible
The market place page was never designed to be a fair pricing mechanism but more of a marketing tool to easily share individual alphas. I understand your point about finding a fair price factoring in the participants in the market but generally, alphas are worth nothing in backtesting and slightly > $0 with a 4-6mo track record. It's only with a fund who is interested in that asset class, trusts the historical track record, and is willing to bet its performance will continue that a match is found and capital deployed. This has little to do with the price of other alphas in the market. However, we will make it public as part of the changes we're making soon -  

> Funds Participating
Agree and we're changing things here in the next couple of weeks. We are changing how QuantConnect can be used to make it more friendly for institutional clients to directly use QC as their infrastructure. This will impact all the services we offer (fixed plans broken into "AWS for Quant"). With this push, we're also streamlining Alpha Streams to be easier consumed by funds operating on the QC Stack. This should also increase the funds licensing as the barriers to using a community-strategy in production will be much lower.

> Alpha Pricing on Performance
This is really hard. Even some of the best alphas struggle to find buyers if no one in the market is trading that asset class. This should be addressed by increasing the participating funds which we're working on with the organization refocus above. ETA 10 days. 

>Alphas Clients are Seeking
Unfortunately very few are open to even disclose publicly what they're seeking so we've had to do generic competitions. This is mainly a problem of the mega-funds (>$1B) and should be easier with sub $100M organizations. 

> Client Acquisition Process
I'd love the community to help with this! We'll set up the referral system+organization focus above and then there will be a solid mechanism for bringing new clients to the platform. 

> In other words, if QuantConnect would be a publicly listed company, I'd definitely buy some shares. ;-)
=D Thank you! 

3

The material on this website is provided for informational purposes only and does not constitute an offer to sell, a solicitation to buy, or a recommendation or endorsement for any security or strategy, nor does it constitute an offer to provide investment advisory services by QuantConnect. In addition, the material offers no opinion with respect to the suitability of any security or specific investment. QuantConnect makes no guarantees as to the accuracy or completeness of the views expressed in the website. The views are subject to change, and may have become unreliable for various reasons, including changes in market conditions or economic circumstances. All investments involve risk, including loss of principal. You should consult with an investment professional before making any investment decisions.


Based on what you are saying JAred, maybe as part of the alphastreams "checklist" submission process you can let the developer know that the securities used in the algo have a high, moderate, low, zero level of alignment with investor interest. 

ie - what good is a 99% PSR over a 20 years period if no-one is interested? This type of information can save us all time and help us developers change focus appropriately?

1

Possibly Grant! With the organization shuffle above we could even have funds publicly list what they're seeking to license so it is more direct. Give a week to roll out the new layer, catch up on sleep for a day, and then we'll start with the integrations above. 

2

The material on this website is provided for informational purposes only and does not constitute an offer to sell, a solicitation to buy, or a recommendation or endorsement for any security or strategy, nor does it constitute an offer to provide investment advisory services by QuantConnect. In addition, the material offers no opinion with respect to the suitability of any security or specific investment. QuantConnect makes no guarantees as to the accuracy or completeness of the views expressed in the website. The views are subject to change, and may have become unreliable for various reasons, including changes in market conditions or economic circumstances. All investments involve risk, including loss of principal. You should consult with an investment professional before making any investment decisions.


Thank you for responding to my input, Jared. I'd like to take this up. 

 

> Effort in developing Alpha Streams

I think that's true if you're already familiar with the QC/LEAN API and Docs. But if you're going to allocate your own money to such an Alpha Stream, it should have been thoroughly tested - preferebly by several independent people. And a fund who is managing others people money will expect this even more. 

Robustness and the avoidance of overfitting is a top prioritiy here. 

Let's take the requirements from the competitions. There we have PSR>80%, short recovery durations (< 6 months), DTW < 0.2 and so on. 

Now let's assume you want to develop something that qualifies for the competitions and in addidtion to that it should be really robust with very low risk to overfitting (not only measured by DTW and correlation of returns) and it should meet the following criteria: PSR > 0.8, CAGR > 20 %, Beta to SPY should be close to zero. 

I think then it starts to get difficult and become a non-trivial task that cannot be done in a few hours. 

It sounds to me a bit like it would be better to have many mediocre alpha streams that perform well only in specific market regimes/phases than to have one very strong alpha stream that is relatively constant profitable. Am I getting this right?

> Pricing of Alphas, licensing revenues and the investment reluctance of institutional clients

You mentioned that an Alpha Stream with a 6 month track record is worth almost nothing. That is good to know and helps me to estimate what I can expect here.

I think many quants, myself included, are creating excessive expectations when it comes to earning money by developing Alphas Streams. 

How long should the track record (of live trading) be at least to let the Alpha Stream be worth something? One year? Three years? The latter is usually the minimum in the industry if you want to get funded. 

And is a track record here on QuantConnect even sufficient for the clients or should it not be a track record which is audited by an independent third party as it's usual in this business? Wouldn't the verification of track records be a good element to build up customer confidence and lowering the hurdles? What do you think about this? Can we integrate such a machanism here at QC?

> Upcoming changes and other points

That sounds very good and is certainly another milestone that will move QC forward. I am very excited about the upcoming adjustments. QC is doing very well. :-)

 

0

One small addition: I forgot the very important factor of scalability in my first point above. Of course, an Alpha should be scalable and be able to meet the metrics mentioned at an AUM of +10 mio USD. 

So in short: an Alpha like RenTec's Medallion fund cannot be developed in a few hours. ;-)

1

Extremly interesting discussion here! Thanks to all of you for your inputs/proposals that could help build a better QC's market, which would be in the best interests of everyone (QC Quants, QC Clients, and the QC Team!)

1

A most interesting discussion and I agree with many of the points raised, in particular incentives.

BTW, there are other platforms that allow any user to subscribe to the signals of any other user's algo, such as Portfolio123, Collective2, and likely others. This involves paying a monthly subscription fee as set by the algo provider. There is no profit sharing. It's all seen as 'personal' or 'private' investment. Users are able to directly link an algo's signals to their brokerage account.
Collective2 boasts that users can switch to any algo within 20 seconds and have it directly linked to their brokerage account. The platform take is ~30% of the subscription fee. A single algo could potentially have 100-1000s of subscribers if it has good performance.
I hope I'm not causing a problem by mentioning these other platforms - just wanted to highlight the model that other algo subscription platforms are using.

I had assumed that QC would eventually also support a similar retail subscription model as well as institutional.

In my view, there are some excellent features that are unique to QC, such as the open source Lean engine, support for Options, and in general is more sophisticated technically compared to the above platforms.
QC has enormous potential given these unique features.

Best,
ES.

 

0

@Ernest: Regarding collective2 I'd like to quote the user R1234 from the elitetrader Thread:

>>"From what I've seen, the majority of subscribers at C2 have a gambler's mentality where they chase the hot systems that have made mega returns. Then the mega drawdown happens and they all exit at the bottom. Quite the comedy show.


There's not much of a subscriber base there for decent systems making consistent but modest returns over the long term.

And because C2 now takes 50% of the fees, there has been a tendency for managers to jack up their fees extra high, and then trade extra aggressively to overcome those high fees."<<

 

While I really like the idea to connect good traders and quants with their ideas and strategies on the one side with the money seeking for alpha on the other side, I must admit that in practice it's often not realized well. 

There are too many gamblers and no mechanism integrated to filter them out. And even if you have good trading systems on offer which are making consistent returns at decent drawdowns, unfortunately, most retail traders don't appreciate that (they prefer huge returns) and will not be ready to pay a monthly subscription fee. The latter is understandable as a retail trader usually have a small account and the cost for subscription (copy signals/trades) will probably eliminate the profits generated by the copied trading system.

1

Hi Arthur,
thanks for the info from the eliteTrader thread. I totally agree with those sentiments.

On C2 I've seen volatility selling algos being the "top of the pops" so to speak, as they look to provide a nice steady "income", but most of the subscribers wouldn't understand the huge risks of these types of strategies.
And then it happened, in January 2018, there was a massive market sell-off and volatility event, and many of these strategies blew up, and from estimates something like (30%) or $30M was lost from the C2 total AUM.
Interestingly, this sparked a new approach from C2 to offer a 'C2Star' certification for strategies that traded within specific risk parameters and requirements. The primary idea here I believe was to get subscribers to think about balancing risk vs reward. What's interesting is that C2 pays $1000pm to the developer for a certified strategy, so this provides a developer with a guaranteed base income.
But, the subscription level to these strategies remained relatively low, with most users still gunning for the high reward and high risk vol selling algos, as well as algos using highly leveraged ETFs.
And when the even bigger COVID-19 event occurred in Feb 2020, the total AUM got hit again, but this impact hasn't been revealed by C2. Their response is that the C2Star strategies reduced the potentially massive hit to the AUM.

So, in response to your concerns that there is no mechanism to filter out gamblers, this certification approach from C2 is an attempt to do this.
In theory, there's nothing preventing QC from using a similar model.


Best,
ES.

0

I'm sure Jared's open mind to hear new offers is the strong foundation for Quantconnect's tremendous growth,
I have a little idea and I would love to get feedback from people who are more expert and have more experience than I do,
I'm sure there are many people who have great ideas with good strategies but have no programming knowledge
The idea is to have a Quantcount website for freelancers to do the job for them,
This will help everyone:
1. It will increase the amount of visitors in Quantcontact and add a lot of good algorithms that will not go the other way
2. It will allow us another source of income that will allow us to work more here and so we can focus on our ideas more
3. Of course, Quantconnect will also earn commissions for any work done through the platform

There is one example of great success in this field:
https://www.mql5.com/en/job

A slight check shows that every hour (!) People upload a new job, there are about five hundred jobs from the last months waiting for freelancers

I would love to hear from the experts

1

We've quite intentionally avoided becoming "just another robot site" through years of conscious decisions. I don't see QC embracing anything like C2/MT5 in the near future for a few reasons:

  1. I think the SEC has a point that most retail investors don't understand how algorithms work or the risks involved. By Alpha Streams requiring sophisticated investors (>$50M AUM) or hopefully soon "accredited investors" we're ensuring vulnerable people aren't being exploited. There are about 10,000 funds that meet these criteria so there are plenty of participants for the Alpha market. 
  2. The goals of "outsized, steady returns" can be better obtained through a fund vehicle applying a diversified pool of algorithms managed by someone who knows how algorithms work. Possibly with the buy-in of the community, we could explore this idea but only if we don't lose trust as its a conflict of interest we've also carefully avoided for years. 
  3. We've carefully built a brand/reputation for enabling high-quality strategies, and high caliber engineers. Trust is critical in the quant space where there are so many things that can go wrong. I would not want to risk that reputation by lowering this too much. Relative to other platforms LEAN is a Ferrari, better than many large institutional systems.  

We've always been razor-focused on providing the best technology and opportunities to the community possible. Keeping this institutional-quality+ approach will be better in the long run for the quant community we serve.

@Miriam - we've considered a job board but continue to recommend people post on job sites like UpWork and Freelancer that specialize in the edge cases of freelancing field. Given the increased risk (capital loss and time losses) doing this right is particularly important which is why we've preferred to focus on doing what we do well. As an entrepreneur, the hardest job is picking what gets your primary focus. The tips/community incentives idea above compliments our primary goal which is the growth and success of the quant community. We want to find ways it can grow beyond the support capacity of the QC team. Scaling by enlisting the help of the community in a mutually beneficial way sounds great and doesn't carry the same risk as directly manipulating a client's codebase.  

3

The material on this website is provided for informational purposes only and does not constitute an offer to sell, a solicitation to buy, or a recommendation or endorsement for any security or strategy, nor does it constitute an offer to provide investment advisory services by QuantConnect. In addition, the material offers no opinion with respect to the suitability of any security or specific investment. QuantConnect makes no guarantees as to the accuracy or completeness of the views expressed in the website. The views are subject to change, and may have become unreliable for various reasons, including changes in market conditions or economic circumstances. All investments involve risk, including loss of principal. You should consult with an investment professional before making any investment decisions.


Update Backtest





0

The material on this website is provided for informational purposes only and does not constitute an offer to sell, a solicitation to buy, or a recommendation or endorsement for any security or strategy, nor does it constitute an offer to provide investment advisory services by QuantConnect. In addition, the material offers no opinion with respect to the suitability of any security or specific investment. QuantConnect makes no guarantees as to the accuracy or completeness of the views expressed in the website. The views are subject to change, and may have become unreliable for various reasons, including changes in market conditions or economic circumstances. All investments involve risk, including loss of principal. You should consult with an investment professional before making any investment decisions.


Loading...

This discussion is closed