Hi Seong,

Cool, but how does this fit with the idea that Quantopian's business model is a crowd-sourced hedge fund, and not a trading platform (which I gather was abandoned as a way to make money)? And I see elsewhere (https://www.quantopian.com/posts/how-to-implement-algo-to-other-trading-website) that you are considering other venues (https://www.robinhood.com/). When I heard your announcement at QuantCon, I had this ringing in my head, "Why are they doing this? Isn't it going to distract from their main mission?"

I don't mean to be a sourpuss...I'm just not following the strategy here. Why put resources into this kind of thing? Don't get me wrong--I'd like you to have wild success. It is just not obvious how this fits in the overall business plan, as I understand it as an outsider.

Grant

Grant,

I think this is a very valid question and I'll try my best to answer it.

We're here to provide quants, whether you're a software engineer at Google or a equity trader at JP Morgan, the tools to test, implement, and profit from your ideas. That could mean that you simply use Quantopian to backtest your strategies, or maybe you're here to take a stab at managing $100,000, or maybe you just want to use our curated, interactive coding environment, or maybe you want to profit and trade your algorithms live in the markets. Whatever your goal is, we're here to make sure that you can reach it.

Speaking personally, I believe our mission is to break down the barriers that Wall Street has held high for so long. The crowd-sourced hedge fund is one way of doing that because it gives everyone an equal chance of doing something only an elite few could do before. Quantopian is here to democratize finance. Adding a new brokerage is one way of doing that.

Just came as a surprise...thanks for the feedback. --Grant

Grant,
Over the last few month I also have been wondering about the divergent goals the Q has stated. Seong seems to thing Quantopian's original goals of democratizing Wall Street are still in place. I very much doubt that. There's no money in it as has been noted on various threads. In fact, as you point out, the expansion of broker offerings is actually counter productive. Maybe you recall the movie Mercury Rising as well as the recent The Imitation Game where cryptographers created intensely difficult puzzles looking to filter in only those people smart enough to prove their intellectual mettle. The mathematically challenged need-not-apply. That philosophy would seem to tie in well with Quantopian's more recent goal settings; cultivate the best algorithms from the smartest yet undiscovered quants and become wealthy leveraging said algos as a hedge fund. If it works maybe somebody will even make a movie about them...

No doubt the research platform will assist those targeted quants. That's a worthy and necessary addition in support of the fund. And of course the Open contributes to the algorithm-trust. But adding cheap brokers? A better forum? Additional help docs or how-tos or simplified tool expansion to assist the uninitiated? You're right. Why? Better to make it HARDER to get started. Raise the entry bar up where the noobs give up sooner and don't waste the fund's time or resources. Truly. Nix all this OWS claptrap. The VC and people with bank don't want to hear about Quantopian supporting the little guys, the simple traders? They want Wilmott, BattleFin and Kaggle caliber quants crunching their stats on the Q's servers.

Seong, are you sure you're working for the company you think you are? This is Wall Street after all.

Market Tech,

Well, opening up to alternate brokers is a means of advertising, thereby pulling in other prospective Q fund managers. Also, in the context of the fund, having only one trading venue (IB) might not be the best practice. If the Q fund gets to $10B, then maybe they'll have enough weight to play one against the other, to drive down costs and get better overall service. Or maybe under the hood (the back-end API), there are advantages to E-Trade and other brokers over IB?

It doesn't seem like this is an equal partnership, yet. I don't see Quantopian mentioned on https://us.etrade.com/active-trading/api. And I don't see any press release or even a peep out of E-Trade on the web. I did see a couple E-Trade guys at QuantCon (by their name badges), but they kept a low profile (e.g. no sales pitch).

Regarding "trading" with $500, I'm skeptical that it'll happen, unless Q subsidizes commissions. I'm also wondering how fractional shares are treated. Say I have a portfolio of 20 stocks. For equal-weighting, I'd put $25 toward each one. So, unless fractional shares are supported, the share prices need to be pretty low. No time now to sort it out, but I do see that E-Trade has (https://us.etrade.com/active-trading/pricing?ploc=at-nav):

Short-Term Trading Fee for Commission-Free ETFs
$15.99 for customers with 150+ stock and options trades during quarter
$19.99 for customers with 0-149 stock and options trades during quarter
You can buy and sell the exchange-traded funds (ETFs) available through the E*TRADE Securities commission-free ETF program

I can't figure out if that's $15.99/$19.99 per trade (surely not), or a monthly/quarterly/annual service fee. Also, which ETFs are available? Anybody know?

Grant

The other thing that is not clear is if I port my (hypothetical) precious strategy over to Quantopian, from E-Trade (or elsewhere), what will Q be able to access? There is a (legal?) commitment not to look at code directly, but what about returns and other performance metrics versus time? Securities being traded? Cloning of algos for backtesting & simulated trading (again, without looking at the code directly)? Or is there truly a separate business from the Q Fund, through which users can just trade without participating in the fund in any way (i.e. fully opt-out)?

As Market Tech says, "This is Wall Street" so the only way to "Hack Wall Street" is to try to think like Wall Street. A bit of a mind-bender...

Seong or other Q person, what is your interpretation of the Q terms of use, in this context?

Of course, one of our goals is to make money for ourselves, our investors, and the people who will eventually invest money in our fund. But our goal is also to democratize Wall Street. These two goals are symbiotic, not conflicting.

The easier we make it for people to engage in quantitative investing, the more people we will enable to do so. The more people we enable to do algo trading, the more effective our strategy will be of using "crowdsourcing" to find good algos for the fund. The size of our pool of potential algo authors is a key component of our pitch to potential investors in our fund. It is entirely to our advantage to continue to do whatever it takes to grow that pool, and that means continuing to improve the site, not just for people whose algos are managing money in our fund, but for anyone who wants to try their hand at algo trading.

There are already more than enough barriers to entry in this field. We do not see "Better to make it HARDER to get started" as an effective strategy. That's the strategy that the existing funds and banks are using, the strategy that makes it hard for people to become algo traders. Our approach is entirely different and will remain that way.

Regarding e*trade, it's my understanding that yes, it's significantly expensive to trade on e*trade than on IB. It's therefore highly unlikely that the algos trading in our fund will use e*trade as their brokerage. The higher cost of trading through e*trade means that while it may be suited for algos that trade only occasionally, e.g., rebalancing algos, it's will be difficult to implement a profitable strategy that trades frequently with e*trade as its back end. We've chosen e*trade as our next integration because many of our users have asked for it, presumably because they would like to run algos that trade infrequently to make sense with e*trade, and because e*trade has a solid integration API.

Regarding what Quantopian looks at, as outlined in our terms of use, we look at algorithms' data exhaust for the legitimate purposes enumerated there, and not for anything else. Certainly, we would never look at an algorithm's exhaust with the intent of reverse engineering, front-running, or in any other way taking advantage of the algorithm's IP without the knowledge and consent of its author.

I think it is important for people to have as easy a path to trading live as possible. For many people, myself included, seeing their algo actually place trades to their account is a huge rush. Automating anything is a bit of a thrill for me, but investing is uniquely exciting. I just imagine all the people in the world trading in the same exchanges, and my bot is out there competing.

IB is great on costs because they have a higher account minimum. E*Trade is the opposite - lower minimum but higher commissions. I think more people will be able to try trading real money with more brokers.

Tell your favorite brokers that they should integrate with us.

A link to more info. on commission-free ETFs through E-Trade:

https://us.etrade.com/e/t/estation/help?id=1204000003

To discourage short-term trading, E*TRADE Securities will charge a short-term trading fee on sales of participating ETFs held less than 30 days.

Seems enticing. No commissions. No Quantopian fees. I wonder what the expense ratios, etc. are? In other words, are there internal expenses of the ETFs that make them less attractive?

Hi Fawce,

Would E-Trade provide you with data on how profitable their users are, particularly at the low end of the capital spectrum? I'm just skeptical. Maybe the histogram of returns, with $500 is so skewed to the left that Quantopian is setting folks up for a high probability of failure. Of course, the cool thing about Quantopian is that anyone can do their own due diligence, before they jump in...nice job there! Part of the traditional "Wall Street" business is to get folks to pay excessive fees with the idea that they'll be able to beat the market with "secret sauce"active management (and of course, there is the gamblers rush and addiction problem, as well). It just doesn't leap out that you are doing the world a favor here, but rather might be trodding a well-worn path.

Full disclosure: Just for yucks, I may try to figure out how to "trade" with $500 at E-Trade via Quantopian's API, with a goal of not paying a dime in commissions and fees. We'll see if it makes sense.

Grant

Hi Grant,

I can ask them if they have profitability data on accounts. Since they are a public company, if that data isn't in their quarterly updates, I very much doubt they will disclose it at all.

One of the reasons we have not pursued forex is because forex brokers only have ~25% customer profitability. The regulatory body for forex is the CFTC, and they started requiring forex brokers to report "profitability statistic". Forex also has extremely low balance requirements and very high leverage.

The simple fact is that you can't trade with any frequency at $10/trade. With Quantopian, you can set the commission model and check what you'll be paying, and decide if the strategy can overcome the transaction costs. The load of the commissions is worse with a smaller balance, so you are right that $500 starting balance will be hard to make profitable. However, I also think it is good to start with a tiny balance just to be sure everything is working as you expect in a live environment.

No commission is the best commission :). Be sure to look at the underlying ETF fees, there is usually an annual percentage fee that you'll be paying inside the ETF. Do you have a link to the no commission ETF list?

thanks,
fawce

Thanks Fawce,

It takes a bit of digging, but here are some links:

https://us.etrade.com/e/t/estation/help?id=1204000003
https://us.etrade.com/e/t/invest/prospectus

As an example, the WisdomTree Earnings 500 ETF has only an annual management fee of 0.28%, when I finally get into the prospectus.

So, I'm thinking that if someone does their homework, and wants to give Quantopian a try, a simple use case would be to do a monthly or quarterly re-balancing of a basket of the commission-free ETFs offered by E-Trade (rather than the ones Seong posted above, which would get slapped with ridiculous commissions). It would fit under your product category of "algorithmic investing" if, in fact, you still have it in your business plan.

Grant

Grant,

I think that's a great idea. Will be updating this thread shortly with a sample basket of commission-free etfs when from that list (https://us.etrade.com/e/t/estation/help?id=1204000003)

Hi Seong,

Sounds good.

Best wishes,

Grant

One consideration mentioned in http://www.forbes.com/sites/investor/2012/02/13/are-commission-free-etfs-really-cheaper/ :

In the case of DEW, it had gaps in its trading history and its bid/ask spread–the difference in the lowest price a seller is willing to accept and the highest price a buyer is willing to pay as of the last trade–was wide. When the spread is wide, you can end up paying more for an ETF than it is actually worth–and this can easily add up to more than a broker’s $10 trading fee.

Is this still in-process? Or has it stalled?

Grant,

Thanks for checking up. We're working on getting ourselves ready for beta full-steam :). Once we have a date in mind, I'll be sending an email out to our waitlist attendees. Thanks for being patient all!

Seong

Hello Seong,

How's this coming along?

Grant

While it is true that IB has a 10K USD account minimum, they do allow 3K minimums for 25 years old and under.

The real kicker is that you need a $25K USD account minimum if you are a pattern day trader, meaning more than 4 trades a week. If you don't have this, your stopped from trading. Ouch.

This $25K minimum isn't an IB rule, it is an SEC rule. E Trade has the same $25K minimum rule for pattern day traders.

Regardless, incorporating E Trade is awesome. Right now, the only real platform to trade with Python with is IB. To bring in E Trade really puts Quantopian in a nice niche. Very nice move.

Hi Seong/Q support,

Just checking back to see how this is coming along. Does it have legs, or is it dead in the water?

Grant

Hi Grant,

We've found this work to be taking longer than expected as we tend to be careful with these sorts of things. We're going through a new planning cycle for expanding our brokerage integrations. It's made for much slower progress than we expected earlier this year.

Thanks
Josh

Josh,

What happened with E-Trade? There was a big announcement at QuantCon 2015 and then apparently it fell apart? Why?

Grant

Grant, I have a few ideas I would like to bounce off your experience. Can you send me your email address or simply call me.
Leo Jones [email protected] / 336-202-0022

This thread has been idle for a while...

Is eTrade still supported?

Also, is there a timeline for supporting options?

@Liz ForPrez

On the options: Maybe there's a more recently update, but I've heard them unwilling to nail down any official date, but they did say "within 2017" in one of the chatwithtraders podcasts.

Etrade brokerage support is not currently a part of our roadmap. As an alternative, I'd encourage you to look into our integrations with Interactive Brokers or Robinhood.

Our current focus in terms of markets is futures. Futures are currently in a closed alpha. We're hoping to open up a beta the first half of this year.

Thanks Harrison & Josh,

IB is a bit inconvenient but not an insurmountable problem. eTrade/OptionsHouse would have been really sweet. I'm not ready to try RobinHood yet (their customer reviews are terrible) although the concept & their VC backers are very intriguing.

Unfortunately, futures don't help me because the bulk of my investing/hedging is in retirement accounts & it seems nobody allows futures trading in IRA accounts :( Plus, I don't really have experience with futures so there would be a significant learning curve for me personally.

Please, please, please consider adding options soon, maybe right after futures!!!