Before founding Estimize, Leigh ran Surfview Capital, a New York-based quantitative investment
management firm. His hobbies include ice hockey and surfing.
1. How
did you conceive the idea for Estimize? Who were the first three people you told
and why? How did they react?
The background begins with a personal
story. I left school to work for a hedge fund here in New York, where we ran an
interesting model that looked at earnings acceleration and analyst estimate
revisions. We were basically searching for the confluence between those factors
related to the inefficiency of the sell-side data set and all of the skew and
bias wrapped up in it and momentum. We were looking for stocks that were
growing quickly, companies that were flourishing, exceeding their estimates
quarter after quarter. We did a good job of removing inefficiency from of that
data set.
That hedge fund is where I learned why the
data set was so inefficient and about all of the incentive-structure issues on
the sell side. The idea for Estimize was born there in 2007 basically a year
after I had begun my career. I am a member of the Millenial Generation. We
believe strongly in web philosophies including openness, sharing and pseudonymity;
in avoiding pay walls and crowdsourcing. These beliefs had not been brought to
finance at large, definitely not to the data set that we are now developing.
Fast forwarding a couple of years, I
left the hedge-fund world to work for a financial start up named Stock Twits.
It was there where we built this community of people who were willing to contribute
their ideas to an open community without
any direct incentive structure. This development set the stage for Estimize. Walking
into a meeting one day at a large, financial-data company with a colleague is
where idea struck me. Walking out of that meeting, I turned to my colleague and
said, “We have to build this! This is the perfect thing for us to build.”
Agreeing, he responded, “Yes, but our founder will not go for the idea.” I
spent the next couple months validating the idea by talking to people in the
industry. I then went to our founder at Stock Twits and told him I should build
this data set. He replied, “Yes, I think it’s a really great idea, but it’s not
for us.” I ended up leaving the company in 2011 to build Estimize. That’s the
genesis of how I started the company.
2. What
is Estimize's business model? What are your products? Who are your customers /
your revenue source?
We are economic terrorists. Thomson
Reuters, FactSet, Bloomberg, and Capital IQ all own a sell-side data set, which
they guard behind a pay wall. They charge at least $1500 a month just to let
you browse it. If you want to actually touch the data, you have to pay them
even more. We said, “Screw all that! We’re going to make the data completely
free at the front-end of the site. We are going to amass this orthogonal data
set that no one else has, and we are going to own it! We are going to own all
of the eyeballs.”
Then, we are going to sell the data
through the back end by launching an API (Application Programming Interface) in
a few weeks and using an Excel plug in. We will sell the API to large,
quantitative hedge funds that do statistical arbitrage trading and other types
of quantitative trading, and the Excel plug in is used by discretionary
traders, analysts, PMs (Project Managers), IR (Investor Relations)
professionals and anyone else who wants to incorporate our data in their Excel
models. Then, we intend to build some premium, front-end analytic features
available via subscription. These features will not be openly available on the
front end of the website. We are also building connectivity between the buy and
sell sides in regards to research. I believe that will develop into an interesting
business as well.
Our customers, the people who pay us,
are large, quantitative hedge funds, asset-management firms and hedge funds on
the discretionary side, investor relations representatives, and banks.
Follow-Up Question: What are you current revenues
and how do you intend to grow over the next couple years?
We do not release our revenues for a
good reason: we have a few serious
competitors in our space. Our current goal is focused on growing the data asset
as large as possible. We aim to become the quote consensus, the distribution of
estimates that everyone must use because we have a set of more accurate, more
representative estimates. Eventually, our focus will shift to generating as
much revenue as we possibly can, but currently the data set is the most
important focus. About a year from now, we plan to really focus on revenue.
3. What
are the top three countries contributing to your open-source platform? Do you
plan to add languages other than English to broaden the field of potential
contributors?
The top three countries are the U.S.,
obviously, followed by Canada, then Japan. Because the third largest group of
contributors is in Japan, and they are working in English, we actually plan to
begin building a Japanese equity-estimate data set before we even build one for
Europe or Canada. After Japan, we will do China, South Korea, and Taiwan. Our
next four equity data sets will be for those countries. However, before
building these data sets we are going to expand our macroeconomic estimate data
set from just the U.S. to international, which will include Japan.
Follow-Up Question: How do you attract analysts and
others who contribute to your open-data platform? How do you manage them?
Concept marketing is our primary method
of user acquisition. We have a content director who writes a lot of great
research with our data, which we publish to our blog. This blog is then republished
to Yahoo! Finance, Business Insider, Forbes and many other sites. These blogs quite effectively attract
people who want to use the data to our site. Many eventually end up registering
to contribute to the content.
We have a completely open philosophy
allowing anyone and everyone to contribute to the data set. We want more and
more people contributing, because we know that when we have more estimates, the
accuracy of the consensus increases. We have to be careful, of course, to
protect the data from those might want to damage it and safeguard it from less
sophisticated people.
We employ two protection algorithms. The
first is a reliability algorithm. It tests whether the information you have
entered is reliable or if you are trying to game the system or otherwise contribute
data that does not belong in the system. If the algorithm detects a problem,
the data are not uploaded to the system. Furthermore, by the end of the day, we
manually review all of these suspect estimates. We are launching our predictive
analytics platform, the second algorithm, next week. This algorithm predicts
which estimates are going to be more accurate for a particular earnings release,
then weights them in the consensus based on their confidence level. We will be
posting confidence scores ranging from 0 to 100 next to every estimate on the
site.
4. Did
you always know that you someday wanted to found your own company, or did you realize
that was your destiny after beginning work in a traditional company?
I’m an accidental entrepreneur. I did
not initially plan to start a company. I was involved in a variety of entrepreneurial
endeavors growing up like selling different products and giving tennis lessons.
I never really had a regular job working for someone else. But I assumed that I
was going to go work for a hedge fund, working my way up the ladder. I did that
at first. Then, I was given the opportunity to work for an amazing finance
technology start up. The idea to leave and start my own venture simply
presented itself there; it was not an end I was striving to achieve.
Follow-Up Question: Taking one additional step back, let me ask
how you initially decided to join a hedge fund with the intention of working
your way up?
I am an autodidactic, which means I do
not learn well in traditional school settings. Rather, I learn best basically
by reading books and following people around. I did not study finance in
school. I actually studied economics and war theory. I thought I was headed for
the CIA, the State Department or the Rand Corporation. I was planning on doing
a Ph.D., working myself through the ranks of one of these organizations. However,
I realized that I did not have the patience to endure another five years of
school by the time I was half way through my undergraduate program. I was
reading a lot of behavioral finance literature, especially books on different
types of investment strategies. Viewing markets as a puzzle, I was interested
in all of the different moving pieces, the behavior of the different groups of
people. I was especially interested in how that behavior was affected by
different market movements.
About this time, my father bumped into
David Geller walking down the street in New York City one day. I had grown up
playing tennis with him as a kid, but never knew his profession. It turned out
that David, who was about 42 at the time, had been running a successful hedge
fund. When they began talking about me, David remarked, “If he wants to become
a trader, have him come up an interview with me.” Shortly afterward, I returned
from San Diego to interview with him. One interesting aspect of the interview
was that he did not any finance or math questions. The interview was quite
strange, like talking to a psychologist. David asked me lots of psychological
questions. When I arrived, he asked me to put together a packing box in his
office, indicating that he had to take care of work in another corner of the
room after which they would begin the interview. I tried to put the box
together, but I could not. I could not seem to assemble it no matter what I had
tried. After about a minute and a half, I went over to David to tell him that I
was unable to assemble the box. Replying,”Don’t worry about it,” he began the
interview.
David hired me as a quantitative trader,
a job I loved. About a year later, when we were discussing some behavioral
finance issues at lunch, he asked, “Do you remember the box?””Yes. Was there a reason
you asked me to put it together?” I replied. David explained, “The box is
designed so that it cannot be assembled. The activity is intended to assess how
you react when you cannot solve a problem. I wanted to know if you become
frustrated and smash the box or neglect to tell me about it.” David then
explained that different types of psychological profiles are suited to
different types of trading. What makes a person good at the type of trading his
company did was not their intelligence or financial knowledge but their
profile. Elaborating, David said, “If you had not said anything but simply sat
their resolutely, you would probably be a good value trader. But you would not
be a good quant trader. “My behavior, namely notifying him of the problem, was
why he hired me.
5. Japan’s economy seems to have
become stuck since last April, when the consumption tax was increased. How would
you evaluate the Japanese stock market currently compared with the US market?
What Japanese companies interest you? Why?
I am not familiar with many individual Japanese companies other than Softbank and a few others. Regarding the market as a whole, I think what Abe has been able to do is interesting. He seems to have stimulated the economy and the stock market to begin growing again. However, because Japan has some serious demographic issues, I am not sure if I would be investing heavily there. I think that the economy is incredibly dynamic and the innovation amazing, but the demographics are going to hold the economy back. Consumer stocks do not do well in Japan as consumer businesses are not a growth sector.
6. What is the most challenging aspect of being an
entrepreneur?
When you run a hedge fund or manage a
money book at a fund, you follow a very specific strategy and process. Every
single day, you execute this same strategy. Over time, you develop a rhythm,
and that rhythm is what makes you effective. It’s what makes you successful. When
you found a company, when you wake up every day, you have to figure it out all
over again. Every day, you have to wake up and make it happen, push the work
forward. But you do not always know how to do so. You have to learn so many
different skills and ideas along the way, not only about how to build a
business, but also about product-market fit. You also need to learn what your
customers want and how to build a better platform. You have to attract people,
hire them into the team, and attract investors, too. There are so many different
moving parts! Figuring out how to configure them all is challenging, but also
fun.
7. What
advice would you give to young business professionals, college and even
high-school students interested in starting a business?
Learn how to code. My advice in a nutshell
is simply that. If you can code, you can build anything. And if you can build
it, you can test it. If you continue to test new things, you will eventually
hit on something that works. I would suggest starting with Python or Objective
C or the new, global Apple language Swift. Learn Python and Swift. If you can
code mobile, specifically, iOS applications, an amazingly powerful distribution
platform that allows you to gather people easily, you are set. Everything is
going mobile.
8. I
understand that you are a passionate hockey player? What are your other hobbies
and interests?
I surf a lot. Whenever I can take a
vacation, I go chase waves. Surfing is an amazing experience that calms me
down. My other passion is international relations. I still read a lot of foreign
policy and research out of several think tanks. I am fascinated by the large
gap between how research is done and how policy is actually created. I also
read a lot of behavioral finance for pleasure. The field is so interesting. We
have so much to learn about it. It’s a really fun hobby.
0 件のコメント:
コメントを投稿