Canalyst is proud to announce our acquisition by Tegus. Learn more here.
Canalyst CEO talks about solving for inefficiency in the capital markets with renowned podcast host and investor, Patrick O’Shaughnessy

Last month, our CEO and co-founder Damir Hot sat down with Patrick O’Shaughnessy for the Invest Like the Best podcast. The full interview can be heard across the four episodes in April, but we thought you’d like to see it (and all in one place) too. Read on, to learn about the origins of Canalyst, the problems we solve for professional investors, and what the future of Canalyst looks like.


Patrick: So Damir when you and I met, I think we were having lunch in New York city, and the auspice of our introduction was that I was trying to learn about enterprise software distribution. I very quickly learned that your business — even though you did teach me about that — was also very investing specific. And it’s a good excuse to ask about the origin of Canalyst. So, already a very popular service with our audience and our listeners, and very, very focused on professional investors – what was the original insight or problem that led you to start the business?

Damir: The original thing was watching my co-founder James — a really smart guy, with a successful career on the buyside, building himself a tool he always wished he had, which was, at the time I met him, about 600 companies that he had modeled and a bunch of Excel spreadsheets and some junior analysts helping him keep things up to date. When you see an individual that smart going through something that looks like a pretty unique motion, you tend to dig a little bit more deeply into it. What I saw there was an opportunity to potentially commercialize something. 

And so I cold-called a number of funds — about 10 of them — and eight of them actually wanted to try it out. So I literally shared a Dropbox folder with them; had them take a kick. A couple of weeks later, I got some feedback. Of the eight of them who tried it, five asked how much to keep it.

The fact that real professional investors were willing to pay real dollars for a bunch of, at the time, fairly rough Excel spreadsheets built by a couple of people in a basement, there was obviously a pain there. 


Patrick: Talk a little bit about what the available solutions were when you started. I remember you saying Bloomberg and FactSet — and there’s other places that give you information on companies’ financials — but what I think is unique about what you built was not viewing what the SEC required the companies put out, but actually putting data in the context that an actual investor would think about a business. So that’s obviously important — that you’re in their world versus something else — but also that you encountered duplicative work that was just staggering. And so talk through duplicative work and framing the product around the actual user. 

Damir: Absolutely. You have some of the smartest folks with the highest opportunity costs, in basically most of the business world, having all these relatively rudimentary challenges, like ‘I need to go to six different filings to pull together a single row of a relevant KPI because none of the platforms that I subscribe to have that KPI and sellside models have it one quarter and not the other. I download a comp table and all the share counts are wildly off, because nobody does the correct treasury method on them and so my comp set is actually completely useless and I blink and I’ve wasted three days pulling PDF documents and manually entering numbers into very similar spreadsheets to what all of the rest of my colleagues are doing across the entire capital markets’. 

So that’s the pain point we solve. We want to be that first 80% of the fundamental model for anyone who’s looking at public equities. That first 80% of the model has to get done really, really well. It has to get done really accurately. There’s a right or wrong way to do it. It’s not art. It’s fairly repeatable. It’s a science. What if you could start from 80%? What are the other things that you can do with all that free time? You can have more conversations with management. You can get much more thorough on a broader shadow coverage universe.

We’re trying to give you back that most precious commodity in the capital markets: time.


Patrick: How does the fixed part of this work? So somebody’s gotta do the work. The first 80% of the exercise is being done by you and your team. What does that look like? Is it humans? Is it software and quantitative systems, is it some combination? How and why should a buyside analyst feel confident that the 80% is being done incredibly well?

Damir: Great question. We started out manually and we realized extremely early on that no one was ever going to use this or adopt it in any way, unless the data was unimpeachably good. Like, extremely accurate. And so what we did was all we could do with human resources – we did everything twice and then made sure that it diffed out perfectly. And basically where it’s evolved to now is the business is around 150 people, about half of whom are on our research team. They build and update everything primarily manually, and then, the majority of what we do build to automate this work, we’ve built as a check against the humans rather than vice versa. A big part of where our data accuracy comes from is the fact that when you model, you actually have to think about the business and things have to work in the forward periods. It’s amazing how often that helps you identify historical errors.


Patrick: Another important thing is how this actually feels to the customer. So people do this in Excel almost entirely. I’m not familiar with many people building company models that don’t do it in Excel, it’s that one tool to rule them all. How do you make sure that what you do doesn’t interrupt the normal workflow of the user – the analyst, or the PM in this case? What have you learned about meeting customers where they are, as you build the product?

Damir: It was not even a deliberate choice, it was an initial requirement. I came in as the B2B tech software person. And the first thing I said to James was, ‘This is great. We’re going to put it on a web portal and we’re going to make it really dynamic and slick’. And he said, ‘No, it has to be back compatible to Excel07, and I have to be able to hit F2 and trace everything back’. I wasn’t going to have a co-founder if that wasn’t part of the solution. Our design principle inside Canalyst for the product is to get out of the way. And what that means is: people are looking to refine a piece of their process. No one’s looking to do investment research in a completely new way.

And I think embracing the fact that we serve an extremely sophisticated clientele. They have great businesses. They’ve got great skill sets. They’ve got a ton of experience. There’s a bunch of other inputs and, at the end of the day, we’re solving for modeling and then solving for fundamental data and actuals as-reported, which is important, but there’s a whole bunch of other research that is happening that we’re not going to affect and we just need to be able to slot into somebody’s process. 

The opportunity costs of switching and learning how to model a different way, hypothetically, is just too high of a bar. So, we have to go and say, ‘that way that you do this part of your work, here’s like an insanely easier and actually higher quality, more efficient way to do it’.


Patrick: So just say a little bit more about the literal way this works in Excel. We talked about you meeting the customer where they are, but what does that mean in the literal sense? Like if I’ve got — which I do right now — have a company model open on my screen, where does Canalyst start integrating with Excel, and how do you think about that? 

Damir:  In the most literal sense, what we’ve now enabled clients to do is pull a Canalyst model. You can add charts, add tabs, rewrite things, change values, reformat it, do virtually anything and Excel is a pretty capable tool. Next earning season, you’ve now got a button in your Excel toolbar where you click ‘Update’, and it will roll in the updated actuals from next earning season, but will preserve all of your changes, and leave your re-drive as it is. And give you a full top to bottom model down comp of where you were, where the actuals landed. That’s where the rubber hits the road. We started with Excel models and then the tooling around it, and now we’re inside Excel, where you do all of your work. 

That Excel add-in and the Updater process that’s been transformational for our clients. If you save somebody 5 or 10 hours in between earning seasons, that’s one thing; if you save them 5 hours on the third Thursday of earnings, when they really need them, that’s a ton of value.


Patrick: How timely is this? I live in an area that’s sort of the hedge fund capital of the world. And I always know when it’s earning season every quarter, because none of my friends are around or want to hang out with me because they’re all busy doing this crap. How quickly does this happen? So a company releases its earnings, what happens next, and how does that flow through to someone that’s using the product? 

Damir: Our goal is: if you care about a name, it’s 30 minutes to an hour. And obviously if it’s a GE model, it might take a little bit longer, but that’s what our clients expect, and that’s what we tend to deliver. There’s a long tail of within 48 hours, we’ve updated every single thing on our entire coverage universe, and on hour 47 is probably the $100 million dollar market cap, rural regional bank, or REIT somewhere that no one has looked at in three years. We maintain that because one day it’ll be valuable. The practical aspect of it is, our clients get to identify what they care about in their portfolio. Anything that a client has actually taken the time to Watch List on our product, that’s [expedited].


Patrick: What is it like to see the capital markets evolve and change and then adapt the product? So I’m thinking specifically here, things like, SPACs in the most recent year, and IPOs. Obviously there’s always been IPOs, but we’ve seen a huge explosion of not just more of them, but huge, enormous companies that are finally becoming public after a long private to public drought. How do you adjust for things like SPACs and IPOs that seem to be not only outside the normal scope of a traditional company model, but present unique challenges? I’m sure there’s demand for those as well.

Damir: Yes, tons. We’ve adjusted by resourcing sufficiently to make sure that we keep up with the number of these new listings. We always budget for a certain number of spin-offs or mergers or IPOs from a team capacity perspective. With IPOs and SPACs, our perspective is they’re new things that are within the coverage that we’ve promised our clients. And so we do have a number of clients who highly value our IPO coverage and always have. Usage by those clients has gone up in the last six months. At the end of the day, IPOs and SPACs are really just an illustrative point of something that’s a principle of ours, which is that we are truly independent. Our reliance is exclusively on public filings. As long as there’s an EDGAR, we have all of the data that we need in order to provide our product. And so we don’t have to wait and model an IPO until we’re off of restriction.

IPO and SPAC models that are done right off the filings that are there in time for the road show, pre-pricing, is an extension of the same value proposition that we offer on the longest standing well-established public companies, which is that you’ve got the data — and you’ve got it formatted the way you want it — when you need it.


Patrick: Say a bit about the opportunity that you’ve seen this unlock in clients. Most specifically, I’m guessing that typical analysts will have a capacity to cover a certain number of companies. A lot of that is just the manual work required and keeping up with them. So if you remove that work, it would stand to reason that they could cover more companies, which I think generally would be a good thing. How does the use of Canalyst change the nature of an analyst’s work or the way that the firm invests?

Damir: We started with an assumption that this is a workflow efficiency time-saving tool, which it obviously very much still is at its core. What we’ve realized when speaking with our clients about why they signed up, why they bought, and then why they renewed, was that [it came down to opportunity cost]. All these times that a company, that otherwise you wouldn’t really have the time or the capacity or the clarity of mind to look at and think about. If you see a press release, you can just grab a [Canalyst] model and look at it. On the IPO side, there were a number of clients who said, ‘this was the first time I could really focus and ask smart questions of management during their roadshow process because I had a model that I really couldn’t have justified at the time to build’. 

So what we see as these opportunistic, ‘Oh, actually I have an ability to quickly look at what makes that company tick’. Having working models on your entire shadow coverage universe versus just on your core positions – these are the things that clients fall in love with. With the best funds in the world, the reason they’re looking at us is actually [an] opportunity to look at things that they just aren’t looking at because the markets move in real time, and that takes the best human a little while to make a model.


Patrick: What do you think the near term and longer term future looks like from a product roadmap standpoint? What are you most excited about in the next 12 to 18 months, and then what’s the dream big, 18 months and longer, vision for where this might go?

Damir: From the first day, we set out to be the fundamental dataset of record for all capital markets participants. That was true day one, when it was me and James and a couple of folks, and it’s true today. We believe we still have a lot of work to do in executing that goal, but that’s the grand vision. We feel very ambitious about it, very happy about it, and very confident about it. 

What that means near and midterm? There are three things that we’re focused on that our clients are already seeing. Those things are: increasing automation, global coverage, and what we call ‘freeing the data’ – basically enabling our clients to consume the content that we’ve created, however they need to consume it — whether it’s on the web, in Excel, in a custom view of their own, or even systematically via an API. We’re going to be launching global in the next little while. We have a true north to cover somewhere around 10,000 companies, up from the four or five thousand we have today.

And then on automation, one of the key things that we’ve launched in the last couple of years has been automating the quarterly earnings. There’s still a lot of investment there for us and a lot of value to provide saving our clients that most valuable, most precious time, which is right in the middle of earning season. Markets are open, things you care about [need to] have updated data. 

The most precious resource our clients have is time and everything that we build at Canalyst that we have built, that we are building, that we’re going to continue to build is about helping them reclaim some of that time – to pick better stocks, make their own clients happier, and ultimately have happier, more successful careers.