Congratulations, Jeff! A year is a real milestone. I've seen you do pretty well in the contest too - maybe you'll be managing some of our money soon, too.

We are full speed ahead on #1.

#2 is indeed a good idea. We've talked a lot about retail offerings. I expect that in the long run we'll make a more formalized offer, but for now, focus on #1 is most important. That said, there is no reason that you or anyone else can't launch a business like #1 on the Quantopian platform today.

Dan, thanks for your note. And, thanks for all of the Q team's incredible work...I think Q is the coolest, most innovative thing out there in wealth management. And, I think the Q hedge fund has truly massive potential for accredited investors, Q, and the quants who get in.....I can't wait to see the continued evolution leading up to a launch date.

Regarding #2, I've thought about starting a low cost online RIA using Q's platform because it's highly scalable and very needed for long term investors. Assuming I charge a 0.25% management fee and based on the internal expense ratios of the ETFs I'm using and based on IB's fees, here are the approximate total annual fees investors would be facing (based on account size......"or less" means I'm showing the highest possible total fee here. Total fees could be less based on how much cash is held at certain times during the year based on signals from the algo and based on IB waiving certain fees based on activity and size):

0.58% approximate total fees (or less) for an account with $50,000
0.50% approximate total fees (or less) for an account with $75,000
0.46% approximate total fees (or less) for an account with $100,000
0.39% approximate total fees (or less) for an account with $250,000
0.36% approximate total fees (or less) for an account with $500,000
0.35% approximate total fees (or less) for an account with $1,000,000

These fees would be slightly higher than Schwab, Wealthfront, Betterment, etc...but, I'd be offering something different, something with true risk management for investors worried about potential multi-month, protracted declines in the stock market.

Conversely, these fees would be significantly lower than big wealth management firms and other traditional RIAs.

I have a business proposition with some questions for Q:

• If I started an on-line RIA using Q's platform, would Q be interested in a 50/50 revenue split (assuming the following)?
  
• I get all the necessary licensing required to be an SEC registered RIA (my own cost)
  
• Q writes annual compliance filing for SEC (ADV and any on-going state notifications if necessary as the company grows...the thought here is to leverage the compliance work being done for the hedge fund)
  
• Q and I collaborate on writing the required compliance manual, privacy notice and advisory contract
  
• I cover my own E&O insurance
  
• Q provides a world class website including process for new client account sign-up with e-signature capability and integration with Interactive Brokers
  
• I agree to make accredited investor clients exclusively aware of Q's hedge fund (only) should they desire diversification with a more elaborate strategy. As a referral fee, I would be paid the 50/50 split normally paid as if my algo was managing the money and Q would keep the rest. The client's investment in Q's hedge fund would remain visible to me as the RIA and listed on client's quarterly statements.
  
• Q would provide free trading service to each client's individual account, in terms of linking the algo to IB
  
• Q and I would collaborate / brainstorm on marketing ideas to get new clients
  
• Q and I would collaborate / brainstorm on innovative future ideas to offer more to clients

The reason I would like to have this conversation in public is because there are probably others out there possibly considering starting an RIA using Quantopian's platform. If Quantopian can offer some help / incentives, then maybe many more quants would be able to start their own RIA wealth management firms (based on their own unique algos).

Hi Jeff,

It sounds like the whole idea is predicated on your ability to write algos that would provide "true risk management for investors worried about potential multi-month, protracted declines in the stock market." Would clients have access to the algos, so that they could understand exactly what they'd be getting for their money? And if so, how would you protect your intellectual property? Or from a client's perspective, would you just provide the strategy details required by regulations, without revealing the algo code? And even if you don't reveal code, I'm wondering if regulations would require you to reveal so much that in the end, those "skilled in the art" could simply replicate your strategy on Quantopian and avoid paying you fees.

Grant

Hi Grant,

Yes correct, that's the idea.
No, I would not show the code, but as an RIA, the philosophy and strategy of the algo would be described in the form ADV. Plus, I'd like to show on-going performance (which you have to be careful about as an RIA, but it is possible).
Maybe they could replicate...if someone is that skilled then they probably invest on their own.

You're a great algo writer...would you ever be interested in starting your own RIA, or partnering with someone on an on-line RIA? Or, are you solely focused on Q's Managers Program?

Hello Jeff,

For me, my involvement in Quantopian is as a hobbyist/DIYer, since I can only commit time intermittently. Maybe the Q fund will be an opportunity for me down the road, but I have a lot of work to do to develop an appropriate algorithm. Regarding partnering, I'm always interested in helping and tinkering using the Quantopian tools. My current focus is to figure out how to use the research platform to it's fullest capability.

Grant

Hi Grant,

Thanks for sharing...I'm a serious hobbyist/DIYer, too...I want it to be more, but I'm trying to find the best/right way to do it. I'll be curious to see more of your posts on the research platform. Thanks in advance for all your contributions.

Jeff

Hi Jeff,

My sense is that if you are looking to set up shop as a formal adviser, it's gonna be tough going. It's too regulated and there's too much competition. And at an 0.125% annual fee, you'll have to have lots of capital for it to make sense as a personal business (e.g. $100M*0.125/100 = $125K). And targeting low-end retail is gonna have lots of overhead that will suck up meager profits (e.g. billing, advertising, customer support, etc.).

Another approach would be to consult as a programmer/implementer, but without any fiduciary responsibility. Say someone comes along with $5M and has an idea/hunch that could be implemented via a Quantopian algorithm. It could be speculative/conservative/crazy/whatever. You could charge $150 / hour for research & code development. So, for 40 hrs. of work, the bill would be $6000. You could collect $2000 up front, $2000 upon delivery of the code (e.g. after validation via Quantopian & IB paper trading), and $2000 after 1 month of successful real-money trading (meaning that the code runs properly, not that the algo makes money). You could offer a kind of warranty, freely fixing any technical bugs that crop up within 90 days. Any substantive changes would cost $150 per hour up front, and a major overhaul would be treated the same as writing an algo from scratch.

From your client's perspective, he's done alright from a cost standpoint, 100%x$6K/$5M = 0.12%, and you'll likely have a product support revenue stream in that eventually, he'll come back to you and want you to tweak his algo.

I don't have the legal and business background, but I'm guessing you'd need to set up an LLC, and make sure that you have the right insurance so that when your client loses his $5M, you are covered. Also, I'm guessing that you have to make sure that you only provide technical assistance in implementing your clients' ideas, since if you get into offering investment advice, you'll be setting yourself up to be sued or shut down by regulators.

Grant

Hi Grant,

You are correct, the business case for starting a really low cost on-line RIA is difficult. The hardest part of it to figure out is the customer service. You could setup an on-line RIA and auotmate everything from account setup to billing to trading, but how would you do customer service? You'd need a really accurate virtual assistant, but that would be a big investment I think.

Jeff

Jeff, it's a very interesting idea. We're just not ready for it yet. Our plate is full with the fund - it has a boatload of work attached to it, plus the continued improvement of the platform. It's worth revisiting in the future, but for the next 6+ months, we're heads-down on the fund model.