Aug 05, 2019 / 05:45PM GMT
Joseph Dean Foresi - Cantor Fitzgerald & Co., Research Division - Analyst
Okay. So here we are with Richard Newman, Head of CTS with FactSet.
I think probably the best way to start off is by giving a little bit of your background and maybe how long you've been at FactSet and some of the details around that.
Richard Newman - FactSet Research Systems Inc. - Head of Content & Technology Solutions
Great, Joe. Thanks. And I'm thrilled to be here. Thank you for inviting me to this conference. I run the Content and Technology Solutions business at FactSet, CTS. I actually arrived at FactSet about roughly 19 years ago now. In an unusual way, I was actually a founder of a software company called Insyte, and in 2000, we were FactSet's first acquisition.
We've done a number since then, but the world's come almost full circle from my roots in this industry. At Insyte, we built an object-oriented database technology that was designed to integrate large investment databases like Fundamentals and Estimates with the firm's internal information and it was a time series-oriented database, very analytical, very cross-sectional, and we built a nice business and FactSet acquired us for our client base and the technology. My focus was typically quantitative users. Back then, the Quants were almost in the corner. They weren't the cool people then. And basically, we built the solution. And at FactSet, I continued that by building up the quant business for about 5 to 10 years.
Many of you who know the FactSet business, our strength has always been integrating data, bringing together third-party content. But in about 2005, we began to collect our own content, or purchase content that we own, Fundamentals, Estimates, Ownership, and built a nice business around that as well in terms of our workstation, our core business. We realized there was a monetization opportunity, typically around the quant workflow to take that data, take that integration experience and build the business around data feeds and around APIs in a sense of business off platform outside of the workstation. And that is when CTS started, the business I run now, which is the business around our off-platform business.
As we move forward, it's grown fast. If you look at FactSet's earnings reports and if you listen to our earnings announcement or transcripts each quarter, CTS has now become a key driver of FactSet's growth. It's over 10% of FactSet's business and the fastest growing business in FactSet. And what the challenge for us a few years ago was, as we collected our own data, we have not only Fundamentals, Estimates, but also other types of data like GeoRevs and Supply Chain and Events and Transcripts was we saw the growth in new content, alternative data, a lot of things you've been hearing about. And the choice we had to make was do we go and create all of our own ESG sentiment information or do we start partnering? And that's when the Open:FactSet Marketplace initiative started, which is the latest initiative under CTS, and I'm sure we'll talk more about this.
And it's been really leveraging our strength in core integration of data, symbology of linking that data together and providing more and more data as the world is becoming much more data-centric.
Joseph Dean Foresi - Cantor Fitzgerald & Co., Research Division - Analyst
Got it. So one of the common themes in the conference, and we came across it not just last year, but in looking at it, people want to jump forward to AI, they feel like Alexa's speaking to you and it's not just -- that she's an actual person, particularly my 82-year-old father who does think Alexa is a real person. But I guess where we start is with the data set, right? In having the data set. And I think what's really unique about the company is that -- that I cover is that, that is an undiscovered value in what they bring to the market.
So maybe you could talk about the quality of the data set, how you get it, its power, its use, and then how you're starting to monetize after what you drive.
Richard Newman - FactSet Research Systems Inc. - Head of Content & Technology Solutions
To me, I run a business at FactSet, it's a business unit, and ultimately, it's not R&D, it's about monetization and how do you do that. Where the world has changed is it's all about data now in terms of analyzing that information. And what's happened with alternative data, I think, I'm not going to say it was a myth, it's true, there's a lot of alternative data that's out there firms are trying to analyze, but you still have to come back to the core content.
So what we do very well is, again, starting with the data we have, Fundamentals, Estimates, Ownership, integrating that with our own premium data, unstructured data, like Events, and then leveraging all that symbology is key. The growth in the market that we've seen now is around data science. I kind of, again, come to that full circle of my career. When I started the Quants, I'd like to say we're in the corner, now it's mainstream. And firms are hiring data scientists, hiring people who can code in Python, code in environments to test data, to come up with new strategies, to generate alpha, to minimize risk. And that's been the big change in the market right now. So our obligation, we, as a company, is to use our core strength, which is integrating all the third-party data in addition to our own data as fast as possible.
What's happened is there's statistics that show that each year, data is growing 10x faster than the previous 50 -- I mean that's not up a little bit, but it's similar to that. And we're just seeing there's more and more data. And our challenge is to help move as quickly as possible to make that happen. I feel like one of my strengths in FactSet is being an entrepreneur by background and what I did building my own company was the ability to be, in a sense, pioneering entrepreneurial and testing new ideas within a company that has $1.5 billion in revenue. And it's just an incredible balance that we can do with the client base we have as well.
Questions and Answers:
Joseph Dean Foresi - Cantor Fitzgerald & Co., Research Division - AnalystGot it. So is there a governor to growth? In other words, data scientists are in vogue, right, obviously we're both from Boston? I mean, outside MIT and Harvard, it seems very hard to find those types of skill sets, all in one individual that can also code and conform it. Is that the governor of the growth, the amount of talent that's available and getting them in the door? Or are there other areas that -- I mean, because you have unstructured, unrealized data that's part of the equation.
Richard Newman - FactSet Research Systems Inc. - Head of Content & Technology Solutions
That's a great question. And the other part about data science, I'd say, I think, firms or the market and the publications are overestimizing how quickly the organization is going to production with it. A lot of it is in the evaluation stage now. The example I gave is if you look at TV shows, you probably know what I'm talking about, is how satellite data is going to change the world. That's one example of alternative data. And that's actually a small subset that covers only about 50 companies typically. The Walmart parking lot example is just one example to fill out. It's all the other types of data, the foot traffic, the sentiment information, the ESG data that's critical.
What I find in data science is, again, it's the evaluation, it's building an environment so firms can test data, test hypotheses, because trying to integrate that data, get into production could take months if not years, and our obligation is to make them happen quickly.
To address the how to hire people, that's what's been interesting. It used to be, we, as a firm, and our clients were hiring CFA-based people. You have a finance background, you take a CFA exam, you do a price-to-earnings ratio analysis, you do earnings surprise, earnings momentum, there were key statistics. That piece of alpha has been just looked at over and over again, and that's not where the opportunity is. The opportunity is these new data sets that link to the core data sets, and that's where the challenge is.
So what I'm finding is there isn't a limit to how many people who can do data science, people can learn that, people can learn programming skills. The key actually now is really thinking differently. And I like to say it's a great opportunity for liberal arts majors and people who come from different background, because the idea to find alpha now is to gain new insights into data requires a different way of thinking. You're not just following Graham-Dodd, you're just not following traditional ways of doing analysis. It's how do you combine foot traffic data with GeoRevs, with Fundamentals, with Ownerships with other types of data and come up with a hypothesis that creates alpha. And with that constant testing and enabling, you need a different type of person to do that.
So I think we overestimate the technical skills needed, and we underestimate the creative skills that are needed.
Joseph Dean Foresi - Cantor Fitzgerald & Co., Research Division - Analyst
Got it. And so how do you -- just to kind of build on it, are you incubating sort of an environment within your team to sort of -- I think I'll say, you'll probably smile, I assume you are. But are you incubating an environment within your team to spur that type of thought process? I mean I'm already going through like a ping-pong table...
Richard Newman - FactSet Research Systems Inc. - Head of Content & Technology Solutions
Well, the reason I'm laughing is -- and if he's listening to this, the head of our, I call it our data science team, he doesn't like when I call him it a data science team. But let me kind of tell you where we've seen the opportunity right now. We've built -- we integrated data, not only FactSet content now, but what we've done with Open:FactSet, the Open:FactSet Marketplace is integrate third-party content from very small fintech firms to firms like MasterCard, and our competitors are even putting data into the marketplace, leveraging our symbology.
So we have all this data now. And our clients, initially, what we thought we'd be doing is sending them data feeds of that information. That's why I mentioned, it's -- clients aren't all ready for that, that's a big process to get IT involved, to build the servers, to get SQL loaded, to get R loaded, to get Python loaded. So we build in the cloud an environment called data exploration where, in SQL, links to R, links to Python, links to Tableau. All the data, FactSet content, third-party content, we have probably over 80 third-party content providers now integrated into this environment, are able to just evaluate and test information.
So that's the starting point. You basically come to us now, you want to trial it, you could try the data, and without having to wait for 3 to 6 months to get your environment build. That's the one thing. The other thing is we have built this incubated group, which I call our data science team, to basically provide templates, what you used to do in Excel. So when FactSet started, we would build -- our consultants would build Excel templates for clients. Excel has moved to Python now. Tableau has become the presentation layer. R and MATLAB becomes the statistical layers.
So we provide a very rich library of examples of Python code, SQL code, other code, as in a sense a starter kit for clients, and they work very closely with our clients. What I've seen in the market so far is, I think, there's been an excitement about alternative data, an excitement about data science. It's still early days. So our job is to coach clients, but we're still not the decision-maker. It's still for the clients to hire the data scientists, the quants and the traditional analysts who are using this as well. I don't want to underestimate. This is not again something in the corner being used by rocket scientists. Fundamental analysts are using data science. They're looking for new ways to evaluate companies, new statistics, new types of information. It's all encompassing at this point. But it's early and that's what's so exciting about, again, coming full circle in my own career is I kind of started here with big data, analytics, integrating data, and we're back there to no longer firms wanting to use us to GUI, they don't want us to say, here's how you do your analysis. The clients want to say, give us the data, give us the tools, give us some starting kits, we'll take it from here.
Joseph Dean Foresi - Cantor Fitzgerald & Co., Research Division - Analyst
And so what would be an example of an innovative or forward-looking data set that you're finding -- I mean, we'll obviously talk a little bit about ESG, what you're finding a demand for that, I certainly think of retail and being able to count bodies going in and out, or we should be up, right, but what is -- maybe you can just -- a product that you're working on that you think is on the forefront of finding a new data set that helps you make a decision?
Richard Newman - FactSet Research Systems Inc. - Head of Content & Technology Solutions
So I'd like to think I invented all this, Joe, I didn't. But I think what got me thinking initially was when we actually acquired Revere Data, so for Supply Chain and GeoRev and other information. That was kind of the start of it. And then we had our own Events data. So FactSet basically keeps track. We record every -- most public companies' earnings reports and keep transcripts, unstructured data. That was the start of it. What I've seen and I won't say I made mistakes, we try things in Open:FactSet, so satellite data, I thought, was going to be a big hit.
It actually wasn't. The big hits were things that were more related to big data sets, they had a lot of companies, and that related to core data like Fundamentals and Estimates. And that was ESG and Sentiment. I'd say new Sentiment would -- and ESG were the 2 big sets that have given us the most opportunity now in the marketplace.
The parts that are coming on now is foot traffic. So we see a lot of partners coming on with foot traffic at this point. That ties into data like MasterCard information, being able to look at sales trends within various types of organizations. A big area again is shipping and other information like that. Just trying to predict where sales are going to be. Our bet is that we're really good at data, at core content, but that there's a lot of firms out there that do a really good job of collecting data, collecting information and are really challenged about how to integrate that with other data. That's our subject matter expertise, that's our core competency at FactSet.
So what's happened in -- for us, is when we introduced Open:FactSet initially, we did try to pick a few winners in the space. And now, they're coming to us firms. And we have tremendous inbound requests to join the Open:FactSet Marketplace, different types of data. So you mentioned ESG. There's probably, again, 20, 30 different ESG providers who are either candidates or looking to join the Open:FactSet Marketplace at this point. Same thing with Sentiment. And we've had some -- we're not going to pick the winners and losers, our attitude is we should be a neutral integrator of this data, provide that symbology and link it together and let the market decide.
So back to our data science team. What they do a lot of is they work directly with clients who are evaluating data and we -- when a new data set comes on to FactSet and we're asked to integrate it, we do the best thing that we found out there, which is market research. We check with our clients, we check with the marketplace. We have a forum. It's private. It's not something anyone can get on and write on. But we have a user forum for institutional clients to basically say, "Hey, this is a data set we're really interested in." And so we created this whole ecosystem that's just -- it's new for us and it's just incredibly exciting.
Joseph Dean Foresi - Cantor Fitzgerald & Co., Research Division - Analyst
Yes. I guess as I was thinking about that as an investment professional, I would want exclusivity on the data set, right? Because if you're proliferating data around foot traffic at a company that I'm doing work on, I would want to be first mover and ahead of time. So within that discussion, how do you, I guess, navigate that challenge, right? And how much -- and do you get bespoke projects where people are like, I'd like to hear more and more about this particular foot traffic issue or this particular satellite data?
Richard Newman - FactSet Research Systems Inc. - Head of Content & Technology Solutions
I'd say there's 2 ways that's been dealt with. One way, you have data exploration now, you have in the cloud, a SQL box, a client can put their own data in there. So if they have something bespoke, something specific to them, they can load their own data, they have tools to integrate that data to the symbology linkages.
The other thing, again, this is an observation, is very rarely is 1 data set the answer a client's looking for. It's the way they integrate that data set with others and that's what becomes proprietary. So I think, again, this is -- my observation is, there's not this magical piece of data that's going to give you that alpha. It's that magical piece of data, how it integrates with other data. Again, MasterCard by itself, or (inaudible) by itself with foot traffic, or Alexandria by itself with Sentiment, or TruValue for ESG, or RepRisk for ESG by itself, well maybe if it's useful, but looking across those data sets with Fundamentals and Estimates is where firms are being very creative in generating the alpha. Ultimately, if you're going to do an analysis and you want [to generate] analyses, you're still going to have to have a 1-month forward price increase. What's the percent if you're doing a back test. That's still part of the analysis.
So that's where I find most of the work. But you're right, I mean, institutional investors, not as much, but I particularly find hedge funds. That was, I would say, a fear upfront or a concern upfront. Now I won't speak for them, but it's been amazing how many hedge funds are using Open:FactSet, involved, and looking at data exploration as a tool for themselves, but they do have that bespoke model, the idea that they can put up their own private data with the same linkages to other information.
Joseph Dean Foresi - Cantor Fitzgerald & Co., Research Division - Analyst
So this might be a little bit of a more difficult question. One innovation...
Richard Newman - FactSet Research Systems Inc. - Head of Content & Technology Solutions
Give it to me. I'm ready.
Joseph Dean Foresi - Cantor Fitzgerald & Co., Research Division - Analyst
Yes. We stopped each other. One innovation that you're working on that really got you fairly excited and 1 data set you really want to get your hands on?
Richard Newman - FactSet Research Systems Inc. - Head of Content & Technology Solutions
All right. Let me start with the innovation. So what's great about data exploration now is the openness to Python to tools like Tableau and all the data linked together. And it's great for a user who wants complete breadth of data starting from ground zero in a sense, we have examples, that's great. But what I'm seeing now is Open:FactSet is not just data, it's also partners, and it's certain workflow winners that basically define a workflow, integrate it with the data's real value.
Again, coming back to where I started, the value of our technology was really big, but it was nothing without the data integrated to it. So we always had to go to a client and then get third-party data in. We've partnered with a firm called Quantopian. I don't know if you've heard of them. Quantopian is a very rich Python-based workflow around -- it's around the quant workflow, data science workflow, rich data statistics, rich set of graphics around quants. It's a star company, very heavily funded by some key venture capital firms and they have a user base of over 200,000 where it's been used as a, in a sense, a free product for firms trying to do their own back -- coming up with back testing algorithms, but then Quantopian would use it in their asset management business.
We saw this tool and realized, boy, that tool combined with all the FactSet content can provide a whole next generation of quant workflow to the market. And the fact that they had over 200,000 users was incredible, a validation of the strength of the platform. So we've worked with them, we now -- we've released it where it's the Quantopian tool, which we call Quantopian Enterprise, with the FactSet content integrated out of the box and that provides out-of-the-box a really powerful quant workflow.
I see that same working in the regulatory, in reg tech, in other markets, in compliance and other areas where you have a great tool, it needs content, and the same way with Open:FactSet is it's become a data marketplace, it becomes an application marketplace. Not -- what we're trying to do has been done, it's a platform play versus a pipeline play. I mean you think about how our business used to be, or still is, you have to hire salespeople who have very deep knowledge of a specific workflow. Our approach now is to build a strength around integrating data, integrating applications and then again providing the platform that lets firms come. And what we do is we're basically the connection to a client. We have connections to almost every institutional client. We bill the client and then we send the royalty back to the partner.
Joseph Dean Foresi - Cantor Fitzgerald & Co., Research Division - Analyst
Got it. And the data set you're looking for, I'm assuming it's private companies.
Richard Newman - FactSet Research Systems Inc. - Head of Content & Technology Solutions
Well, you probably read my mind. Private company is a huge area right now. Private company data in several dimensions. One dimension is just Fundamentals, Estimates, basic information on private companies is really hard to come by. If you look at the institutional market, if you look at the sell-side market, private equity is where it's at. So the ability to have -- and there's a bunch of third-parties that are joining Open:FactSet around private company. Again, no one has like -- when you think about FactSet Fundamentals, we have hundreds of attributes on every public company because it's all available. You can get it through SEC filings.
But the private market is really hard. So we're now working with several third-parties, but that's a big area to hit in that area. And then deal side of private equity, being able to know what's out there, what companies, that's another area that's big.
We estimate, in the next few years, I mean, a lot of the big institutions are already doing this, you look at the portfolio and you want to price it, a huge percentage of those portfolios are private companies now. You need the Fundamentals, the Estimates, the Ownership, the alternative data as well as you need the pure deal information to figure out some -- about who's funding these private companies and things like that, what are the risks with them lasting and going forward. That's a big area.
The other area I'd say you mentioned before is foot traffic. It really is just that's new, it's how do you actually use that information. There's so much of it. The risk you have and then when you get into a lot of this data are the privacy issues you got to think about, and that's something that, again, is going to be an ongoing challenge for a lot of data firms in this market.
Joseph Dean Foresi - Cantor Fitzgerald & Co., Research Division - Analyst
How do you compel a private company or others to provide you with that unstructured data? What is their incentive? Because obviously, things can change around private companies, IPOs, seed rounds and stuff like that. How do you compel them to feed that database?
Richard Newman - FactSet Research Systems Inc. - Head of Content & Technology Solutions
It's a challenge. And it's a challenge, either there's several directions to it, some companies are doing it. You can -- sometimes the private equity firms themselves are doing it because it helps them to get -- they contribute, they get other back, and that's a traditional model, pay to play. If you want to get other private company data, you've got to put your data into it. So it's building up the same network is going to be ultimately the winner in that, I believe, for the private company game. Again, I think it is early. This is, I'd say, probably a 3- to 5-year horizon right now to -- there's going to be early adopters. There's going to be some winners in terms of data companies that really do well in private equity, and then there's ultimately going to be a consolidation.
Joseph Dean Foresi - Cantor Fitzgerald & Co., Research Division - Analyst
Got it. So as investment management has shifted from active buy-side, long only, to more of a hedge fund bend, to eventually kind of over to private brokerage house-type investment management side on the wealth side. How have you, or have you changed the way that you approach the data sets associated with those different shifts in the market?
Richard Newman - FactSet Research Systems Inc. - Head of Content & Technology Solutions
It really is interesting. So there is a passive shift. And I'll come back to PEs, earnings surprise, earnings momentum, price momentum. That's almost where it's passive, but there's always going to be -- the good thing about this market, people are always going to want to get an edge. There's always going to be ways to find alpha. That is the alternative game. And when you think about that market now, we're going with alternative data, it's never-ending. I mean, basically, it's just a market that's going to keep growing because the iterations and the -- if you encapsulate various types of queries and things that you're trying to do, there's just an infinite way to try to generate alpha now. That's been the big shift.
So to me, active management is going to require that creativity. You're absolutely right about passive management, it's gone. Again, there's only so much you can do with price-to-earnings ratios. I think it's been, been there, done that, and that -- and those types of analysts, I think, that's where the market is going to be. People say, what's the future for people in this business? We're going to need the creative people to come up with new ways to generate alpha, they're there. The traditional analyst, I think, is going to be tougher, just basically looking at balance sheets, financial statements, but hope there's not that many of you out there, sorry.
Joseph Dean Foresi - Cantor Fitzgerald & Co., Research Division - Analyst
So I mean you've been 20 years with FactSet. You've probably seen a tremendous amount of different ways and boom and busts of investment management cycle. Maybe you could frame it for us in the historical context, right? How much of what we hear around ESG and data management and because data seems open-ended, so it's probably not a time-stamped initiative, right? But from a broader perspective, how do you look at what's going on today versus what took place 5 years ago and 10 years ago? How much do you feel is more fad versus directional?
Richard Newman - FactSet Research Systems Inc. - Head of Content & Technology Solutions
It's a great question. So one thing I can say for people who integrate data, and I've been doing this a long time. I can say sadly for more than 19 years, I thought about CUSIP changes and corporate actions and how to link securities of different types. That's gotten more complicated, not less. So those opportunities -- again, AI is helping, machine learning is helping to link that data together, but still, sometimes you need people to do it. You still need that. So that area is definitely changing.
I think the big shift has been the types of users, types of people, investment firms are hiring. It went from, again, back in the '90s, you did have your programmers, your FORTRAN programmers. For people who know what FORTRAN is, back in the -- who would do custom development and things. Then it went to pure GUI-type products from vendors and maybe some custom programming now on the side, but not necessarily standards. What's moved now is the standardization of core capabilities.
So if you look at things like Python, you look at things like SQL, you look at various ways of using information, that standardization, the idea of open systems is here. That's the big change, that no longer is proprietary acceptable. And as someone who comes from -- we built some proprietary technology, that was my roots. That's been the change. If you're going to go out and build your own proprietary database technology, I'd probably recommend against that at this point. You got to look at standards. The big shift though is type of users. You do have to be technical. You have to be, again, creative yet be mathematical. I think that is important. You need to be able to write. I think people need to be able to express themselves. I think there's a lot of that now to explain the new ways of doing things.
The opportunities are bound -- are endless in terms of what we see in the marketplace right now. I do think alternative data is here to stay -- not I think. I know it's here to stay. I just don't want people to get overly excited, I guess, that every type of alternative data is going to find alpha. I think the big surprise to me having been on this journey for the last year now with Open:FactSet is it's more about evaluating than going into production at this point. People are still testing ideas. They're trying to come up with those hypotheses. That's still the stage we're at. I don't think -- hedge funds are definitely using it, but I think in the traditional investment management business, it's still early. I think the idea they're evaluating -- they're building data science teams. They're trying ideas, but to go to pure investment process to production, I still think we're early.
Joseph Dean Foresi - Cantor Fitzgerald & Co., Research Division - Analyst
Has the average end-user gotten -- or has the sophistication of the average end-user continued to go up over the last 19 years?
Richard Newman - FactSet Research Systems Inc. - Head of Content & Technology Solutions
So what's amazing is yes. And so I wasn't at the meeting, our CEO actually was recently at a big bank, and one of the comments that was made by the head of this group there was every junior banker is going to have to know Python. I mean it was Excel, and now it's Python. And I think that's true. I think you're going to have to not necessarily learn in college or anything, but you're going to take Python courses, and it's amazing how many people I know are taking Python courses. I don't care if they're 60 or they're 20, you have to know it. Just like we probably all knew Excel. That's the big shift. So that's -- it's probably -- I'm not sure if Excel users were considered technical, but Python is the analog right now to that.
Joseph Dean Foresi - Cantor Fitzgerald & Co., Research Division - Analyst
Got it. So your efforts are to get ahead of the pace of an increasingly sophisticated end-user who is finding it harder and harder, even though the market's up, not today, to generate alpha.
Richard Newman - FactSet Research Systems Inc. - Head of Content & Technology Solutions
So for FactSet as a whole, our job is to get as much data available, as quickly integrated as possible to give users the opportunity to do as much evaluation as possible and to give you a solution that you can very quickly test ideas using open technology.
So if you want to use Python, you want to use Java, you want to use C++, you want to Tableau, you want to use R, MATLAB, SAS, whatever, that's there. And to offer workflow solutions, whether we build them ourselves or third-parties build them. So there are a lot of third-party start-ups out there that are hungry for data. We feel that, that's our -- our strength is data. We are a content company and we are a data-driven company and the market now is data-driven. So I'd think about the keyword now is data-driven solutions. We're basically seeing it in the market now. It's about data, data, data. And we feel we're in an incredibly sweet spot because we're at our roots.
What's been interesting for my own journey and I think FactSet's journey is we're kind of doing what we've always done. And if I think about FactSet historically, it was integrating data, providing access to Excel. What we're doing is we're integrating data, and you can access it on the cloud, so you have infinite capacity. I mean the other big difference now because it's on the cloud, we don't -- clients -- we don't get happy anymore if clients bring our systems to its knees because it's just buy more processing, more discs, it's all there, it's so much more open. Openness is a key attribute of our strategy.
Joseph Dean Foresi - Cantor Fitzgerald & Co., Research Division - Analyst
How much is the unstructured nature of the data a bottleneck? Is that -- in addition to, sounds like you've kind of figured out a way culturally to overcome sort of the need for more and more data scientists. You still have all this new data, particularly with private companies, in an unstructured fashion. Is that a bottleneck...
Richard Newman - FactSet Research Systems Inc. - Head of Content & Technology Solutions
It is. It's hard. And to do it right requires a sophistication and a data knowledge. You can't just slap on unstructured data. So again, machine learning helps. So if you're getting an event, for example, and you come up with keywords, you look at the event and you -- ultimately, what you're doing is you're taking unstructured data, making it structured. One way we're addressing that, again, is with our partners. So if I look at some of the Sentiment feeds we're working with, they're pulling in Twitter feeds, they're doing stuff, a lot of the partners are doing the -- taking the unstructured and structuring it and sending it to us.
So that's where we find a lot of that. But that's a challenge. There's no doubt about it. The more data that comes out -- and it's picking again winners versus losers. I look back to my initial choices and some of the first partners we put into Open:FactSet, some are winners, some weren't. It's making us less that decision-maker, letting the market do it. But if we have a big partner, some of the big players out there, and they're providing data to us, it can be a lift to get it integrated. You got to think about those. The smaller sets are not, but sometimes the big sets requires saying, okay, if we're going to put this up, we've got to make a bet on this data set.
Joseph Dean Foresi - Cantor Fitzgerald & Co., Research Division - Analyst
Got it. So 2 final ones. First -- or one -- second will be a little bit more fun question, but on the decision-making process around investments, you mentioned sort of historically you look for the CFA types who can do the PEs, and I'm sure there's some element of forecasting involved as well. Does data become the third piece of an investment decision? Do you think that it provides the edge and is that where we're headed in the next 3 to 5 years?
Richard Newman - FactSet Research Systems Inc. - Head of Content & Technology Solutions
Yes. I think you're right. I mean I think it's no longer some -- and again, I hope I'm not offending anyone. I don't think it's professor coming up with the ideas. That's where machine learning and AI is going to come in. They're going help go through lots of data and look for trends. Only a person is making the decision. But looking at ways to make decisions is very data-driven. So absolutely, the theory is going to come from the data itself. I think data becomes the commodity.
What you said before is really interesting. I do think the way firms decide to use data or come up with some Web scraping or other ways to come up with bespoke data will give them an edge. But you still have to go back to the core. You still ultimately are investing in a commodity, in a company, in a security of some type and you're looking for a return.
The basics of investments and finance are the same. It's how you come up with that decision is different. It's still going to require though that, again, creativity to come up with the information. The computers and the data are there to help and there's horsepower, you can get that from computers. You have infinite data, unlimited capacity and hardware is fantastic. But ultimately, really, people make decisions.
Joseph Dean Foresi - Cantor Fitzgerald & Co., Research Division - Analyst
Got it. And the last one for me. We had an AI event with an MIT professor, and I always assumed, because of what I do for a vocation, that he wouldn't be able to beat computers. So he said, natural thought process, yes. Humans can take in more and they can understand more in a quicker fashion and computers are just linear. So where I'm going with this is that he said, as we look at AI, the function is to go through all the different nodes, and as we get more data, then we can feed more and more data into the nodes, essentially taking out the emotional aspect of investing. And theoretically, his conclusion was, yes, a computer could definitely invest better than you.
Now I know it's a little bit of a curveball at the very end, but how do you think about now you've got all this proliferation of data that you could theoretically send to all these different nodes that can be tweaked, but you can check it, recheck it and keep putting more and more information to it, and what a computer can do better than a human is process the data, right? So I just wanted to get your opinion on (inaudible)
Richard Newman - FactSet Research Systems Inc. - Head of Content & Technology Solutions
Yes, I'll come back with the black box theory and what's happened historically, and I always -- again, I'll go -- I promise not to go too rogue because I want to here. But it's always the funny thing is my investment strategy was perfect, the computer said this, but this 1 event, if the Black Swan or something happens because that's still there. And back to sentiment, there is emotion in investing. There is sentiment information. So no matter how much you let the computer do, there's greed, there's other attributes within an investment decision, those aren't going away. If it were completely true, then you won't have a market because everybody is going to make the same decision. And that's why I don't believe that.
I do believe what computers do and data science is help you come up with new ways to generate alpha, and you are going to, in a sense, crowd out the ones who are not thinking creatively because those are going to become very standard. And that's I think where passive investments come from. Really having an index or an enhanced index fund, probably not the big thing anymore. But it's coming up with funds that are ESG-oriented, sentiment-oriented, that's real. And that's the other part of this, too, in terms of investing. You mentioned ESG before. There is a huge opportunity there where people want to invest in companies for certain reasons. But the ESG, what's interesting in terms of the social part of it, I mean, does it make sense to invest in companies who are very socially aware. Does that attract to those companies, better employees, more loyal employees, employees who are going to stay there, employees who work harder? I don't know. But that's why ESG is so important. Those are the statistics and things that could be interesting, whether it's a carbon tax, other information or other types of information coming in there, that's where, again, I don't think a computer can do it now. People can come up with those hypotheses and then let the data decide if makes any sense.
That's what really interests me. I'm really fascinated by ESG because it was probably laughed at a while ago, but it's real. And there's so many firms addressing that. And it's not just the E, which is big, the environment is obviously the E. It's not just the G, governance, but it's going to be that social part. I think there's a lot -- especially with this new generation of people. I think there's a real hard belief -- and I don't think -- I can't tell you now what that is, but I think that's the next big area we'd be looking at.
Joseph Dean Foresi - Cantor Fitzgerald & Co., Research Division - Analyst
And then just the last piece of that is are you seeing -- I mean is the demand for ESG outstripping the infrastructure at this point?
Richard Newman - FactSet Research Systems Inc. - Head of Content & Technology Solutions
I'm trying to think about the infrastructure. I think there's so -- it's how to evaluate which ones are effective. So there's a lot of ESG data out there. We have a lot of inbound requests from ESG providers to provide information. The reason why ESG is so hot is not only the alpha side, but it's the risk side and the compliance side. So firms, when they build portfolios, have -- their mandates require certain diversity in companies they invest in, or pollution issues and so on. And now there's a test, is this data accurate? Is it good? So I'd say, it's really now in a stage of evaluating which ESG providers are doing it right and which ones aren't. But that's a really exciting area. And again, I would say there, again, I'm not going to bet on 1 because you can't just have 1, ESGs have so many different attributes. And I put -- so that's one where it gets -- the buzz is correct though for ESG. It's real. That's when I would -- I look at companies -- personally, I look at areas and say ESG is a place where it's just you have to, as an investor, look at as well.
Joseph Dean Foresi - Cantor Fitzgerald & Co., Research Division - Analyst
So the investment style would -- or the investment process would change. It's not just investment thesis, forecasting, value opportunity, learning the PE, now you've got the new emergence of data sets and how much it can impact it, followed by the ESG aspect.
Richard Newman - FactSet Research Systems Inc. - Head of Content & Technology Solutions
The ESG aspect of it, again, it's true. Environment, social and governance. It's real. I mean, governance is real. When you're investing, we have to make sure as a portfolio manager, you can explain why you made decisions. You have to explain why a company had some criminal offense against them or something, how did you not see that? I mean, again, there are so many examples of it. The examples always show that it would've called Volkswagen, they would have called it before hand, they're always after the fact. But as an investor, you don't want to be caught holding those companies that have a huge issue of one sort or the other.
Joseph Dean Foresi - Cantor Fitzgerald & Co., Research Division - Analyst
Got it. Richard, thank you.
Richard Newman - FactSet Research Systems Inc. - Head of Content & Technology Solutions
Great. Thank you so much. I had a great time.
Call participants:
Corporate ParticipantsRichard Newman, FactSet Research Systems Inc. - Head of Content & Technology Solutions
Conference Call Participants
Joseph Dean Foresi, Cantor Fitzgerald & Co., Research Division - Analyst
Refinitiv StreetEvents Transcript
FactSet Research Systems Inc at Cantor Fitzgerald Innovation Summit
Aug 05, 2019 / 05:45PM GMT
