TechTalk

TechTalkblog

InvestmentNews takes advisers through the developments and innovations in technology that’ll change the way you do business today—and tomorrow.

A chat with the founders of Quantopian

The co-founders like, respect and think quants are underserved — and have too little access to tools of their own

Feb 6, 2013 @ 12:31 pm

By Davis Janowski

+ Zoom
(Quantopian)

I enjoyed my interview with the founders of Quantopian so much that I thought I would just share it here on my blog. Here it is in its entirety or rather as close to verbatim as my short hand will allow me to flesh out. Enjoy.

Meet John 'Fawce' Fawcett, co-founder and chief executive of Quantopian and co-founder Jean Bredeche.

Davis: So from where did the idea for Quantopian spring up?

Fawce: STEM, science technology engineering and math. They make up the base of our user community, people in those fields. Whether undergrads, grads or professionals. The vast majority are not on Wall Street.

Some have interest in signals analysis (and several other areas I was not fast enough to get down). This is all applicable to the financial markets. There are all these people out there with the skills and talents but the pathway to turn them into quants has not been available.

When we talk to students that are graduating and going to top firms like DE Shaw, Bridgewater [Associates], etc. what appeals to them is the mentorship. People entering those fields really need that networking opportunity and mentorship.

STEM professionals and students mixing. Some of them will explore [Quantopian] and backtest and never trade while some of those will end up opening accounts [to trade and test what they have built].

One thing the capital markets are not really good at is discovering talent. So people will develop this experience and develop a track record [at Quantopian] and then be able to be discovered.

Being a quant is a talent game and our intention is to open things up and cast as wide a net as possible versus what is a very closed world today.

They look at a very narrow set of schools these days, maybe 10 top schools. That is a very small number given the millions of bright people out there.

It is very easy to clone a sample algorithm and then run a backtest on it.
+ Zoom
It is very easy to clone a sample algorithm and then run a backtest on it. (Davis Janowski, created using sample code and tools on Quantopian)

Davis: So how old are you guys and how long ago did you get the idea for this?

(Both are in their 30s)

Fawce: I began thinking about this while I was back in school.

At that time, the majority of websites still had 15-min.-away pricing. At that time I couldn't figure out where to get the data or even where to start…My career took me in a different direction.

(Both men said they worked together at Tamale Software LLC but in the financial arena, though not quantitative analysis).

Fawce: My take on it was that there were so many smart interesting people on the quant and analytical side. But there is a crazy imbalance between those raising the money in the fund and those managing the quants in terms of compensation and recognition. I liked the quants so much I really wanted to make a product for them.

Then as I learned more about it, [I came to realize] that quantitative is really product. Lots of researchers, a treadmill almost that they remain constantly on, an assembly line where you have to bring lots of them through and that is where our whole drive to open this up to a much wider audience and try to work with them came from.

Jean: The way I see it, I worked with FAwc for a long time…built tools that were around qualitative research not quantitative.

Actually both of us have engineering backgrounds and probably could have done it [become quants or work in that area] but not really the inclination. On top of being a very interesting engineering problem [becoming a quant], which everyone has to do [pursue] themselves because getting the practical experience is so difficult no one will share it (because most folks learn the practical aspects working for a firm, the work product of which stays proprietary). Without having a standardized backtester that is shared you cannot ever truly backtest or compare apples to apples.

With the big firms on Wall Street, when I meet quants they immediately have a vested interest [in what Quantopian is doing]. A lot of them get restless after three years [working at a firm]. They dream of trading on their own. They spend six, 12 or 18 months building the systems they would like to use [on the outside] inside a firm. The quants are very encouraging toward us, whether in contributing to Zipline or advice on how to [this or that]…

They see this as a path to a certain kind of freedom so they help us a lot.

Fawce: People on the management side are dismissive. 'We spend a lot more money on this than you ever will.' This definitely fires me up. I think it is total nonsense and hubris.

There are a lot of people out there that are talented and do not want to work for this market but love the idea of competing in the market…80% of any of the trades are done by algorithms. Whether a force for good or not, it is still a force. I feel that the communities we are building and team we are building here could help to balance that out.

I think it [the current situation] is a tremendous danger, for such a small number of people having such control over the markets.

If you read anything from some of the famous people outside the market world they will say they are lost, they don't understand how the markets even work any longer.

The idea that it should be kept secret is very dangerous. Things like the flash crash, hacking, and even regulators playing to catch up with what is going on should prove that.

If you have faith in humanity like we do, this is a reason to be hopeful.

Jean: I think it was Marc Andreessen who noted that three-quarters or 80% of the market is controlled by computers.

Our dream is that now that power will be possible for everyone.

We are taking advantage of all the computing trends and trying to share it with those rather than it residing behind a walled garden of a couple thousand quants on Wall Street.

Davis: So who is supplying you guys with your data, I'm assuming you still have to pay for it?

Both men noted that the data Quantopian runs on — minute by minute US equity history for the last 10 years — is currently purchased from/supplied by Tick Data Inc. (which should be lauded for publishing its price list, a rarity among any firm on or tangential to the Street).

Fawc: We are in the process of getting new datasets and people are asking us for all the things you'd expect like options, forex and futures and commodities.

Davis: So what do you guys see as trends that will affect consumers and/or advisers that might intersect or perhaps prove tangential to what you are developing?

Fawc: One is the kind of business service vendors you are talking about that are cropping up (I threw out HiddenLevers.com and MacroRisk Analytics LLC as examples). There is also a trend of services and tools like Wealthfront, Personal Capital, etc that I find a lot more interesting.

The RIA and wealth management industry is a very sales driven business. You build a book and it requires a lot of time to manage and does not leave a lot of time for asset management.

The end consumer in coming years is going to have tools where they can have a high quality yet automated service.

Davis: How do you envision people or perhaps other developers using Quantpian?

We see people building all sorts of things on top of Quantopian, like tax optimization for example. You can imagine in five or 10 years as our platform matures and people become more mature in terms of how to use it, for example trying to predict things that could come up and affect your 401K.

Davis: What do you think of the technology being built by custodians and online brokerages?

Fawc: They are much cheaper than current [Wall Street/wirehouse] offerings. There are pieces of them that we admire. Etrade, Fidelity and Schwab and TD Ameritrade, they have great customer service and are amazingly efficient in terms of their onboarding process and infrastructure and in creating accounts. Interactive Brokers is remarkable in how they lower costs and have an execution API and can keep their costs down but remain streamlined.

Davis: And what about fellow startups like Motif Investing?

Jean: Currently, the staff of Motif actually constructs the motifs. It would be more amazing if members of their community could do that. (Which is a feature they are working on if they have not already rolled it out).

And we really like macro investing and it is an area where only the big boys play but quants love and so it is one of the things we gravitate toward.

For more information and even a short video on how it all works visit Quantopian online.

Davis For those that want to see some highly active threads on Quantopan, the guys encouraged me to search there using the term “ thisisamazing

Get Daily News & Intel

Breaking news and in-depth coverage of essential topics delivered straight to your inbox.

X

Subscribe and SAVE over 72%

View our best offer
Subscribe to Print