Snowflake Inc. (SNOW) Earnings Call Transcript

All right. Why don't we jump right into this. Hello, everyone. Welcome to today's webinar. My name is David Coluccio and I head up Distribution Solutions at S&P Global Market Intelligence. And I'm really excited to moderate and speak at today's webinar, Navigating The Storm: Innovating Asset Management in a Volatile Market. So as we get started, let me just kick it off by introducing our panelists. So first off, I'm just pleased to be joined by 2 leaders in the industry. David Murdock, he's Vice President, Product Management, Visible Alpha and S&P Global Market Intelligence. And Chris Napoli, he's responsible for Wealth & Asset Management strategy at Snowflake. So I want to welcome you both and thank you for taking the time to join us on today's webinar. I guess, I know we want to jump right into it but before we get started, there's just a few housekeeping reminders that I just want to kind of highlight. First off, all the engagement tools are resizable and movable. So definitely take the time, move them around to get the most of your monitor space. We're really hoping for an interactive session. So if there's any questions, the Q&A button is at the bottom of your screen. Just submit the question. We'll be the only ones that see the questions. So no one else will see what you asked and then we'll try our best to respond at the end of the webinar or we will definitely follow up with you afterwards. There's also a related content widget that includes links to our thought leadership resources. So -- and you could also find our webinar replay or other webinars as well on demand. We provide close caption. So just click on CC icon at the bottom. And then last but not least, there's going to be a survey that will appear at the end of the session. Please take the time. It's less than 1 minute to complete. It really does help us understand what you like, what you don't like and it'll just lets us improve these webinars going forward. And actually, before I get started, I do want to note that the activities of S&P Global Market Intelligence are independent and separate from S&P Global Ratings. S&P Global Ratings maintains a separate -- a separation of analytical and commercial activities. So with that, thank you all very much for joining us.

So let's get started. I guess to kick it off, I'm going to -- I'll start off with Dave, right? So Dave, S&P Global recently acquired Visible Alpha, which I know has some really unique content. Can you just kick us off and give us an overview of the Visible Alpha offering? And how Visible Alpha differentiates itself from other providers?

Yes. Sure. Dave, thanks for the intro here. So yes, looking at the slide here, if we think about it, kind of from left to right, it's almost building a data product. How do you kind of put it through the factory? And I'll start with the left-hand side, the raw materials that we have here. We partner with over 200 sell-side firms and we use that to bring in what we call their working model. So each analyst is in Excel modeling out the company, all the financial statements, even KPIs and other metrics. They're taking those and we're combining that with some of their research, some filings that we do, as well as market data. And we put all that into that kind of purple ribbon you see there, which is our machine. And that's a combination of really tech and human capital. And I want to make an important note here, too. I think it's important to have balance when you think about those 2 things. If you completely automate everything, I think you lose a lot of the nuance of some of these complex datasets. And of course, if you do everything manual, it's too costly to maintain. So I think we use a nice combination there. You'll see there's over 400 data analysts that really know their sector, their industry and are able to help bring this data to life. When I get to the middle of this, it starts to -- we start to build the foundation. So we think about analyst data, we take the data as is from those models and bring it into a database and then we start to layer on top of that. So at the company data level, we're then taking that and saying for each broker, the data items that we want out there and this is all forecast and as well as actual data. What are those datasets that they're cross comparable within that company. So if you're looking at Apple as an example, you can see iPhone units. We then take that another layer up for the standardized data level and say, now I want to do multi company analysis. I want to be able to compare and look at that data across companies. And that's what our standardized view does. I mentioned our actual datasets. We actually look at the historical as well. We combine filings and then also take that with our crowdsourced estimates and combine them into 2. At that point, we then look at the different delivery channels that we offer. So we've got a insights desktop. We also provide this data via an add-in and being able to pull formulas and refresh it within the Excel environment. And then the part that I focus on is really the 3 at the bottom there. So we have an API, RESTful API. We have a CSV data feed format. And we also have one of the partners who're obviously on the call today are Snowflake channel and doing more cloud sharing and delivery that way. And you package it all together and you think about these different channels and you -- really our target markets for this are the buy side, corporate IR departments as well as the sell side. And you'll see there's a number of use cases in terms of the types of analysis we see clients doing with this content. So what problem do we actually solve? I mean thinking about this, you may be saying in the audience, hey, I already have an estimates provider. What more does Visible Alpha brings to the table? We actually don't really view ourselves as a competitor to traditional forecast data. We're much more of what I would call a supplement for those that are looking to do deeper analysis. So these working cell models are giving us, a lot of times, hundreds of line items per company. And it really lets you to dig deep into that as well as having deep source counts on those critical line items that help move the markets. And the other thing I would bring up to, is our longer forecast data. So being able to go out sometimes 10, 15 years, doing that with pharmaceutical companies, mining industries, things that take a long time sometimes to develop and being able to utilize that to your advantage. So in a nutshell, we go beyond the financial metrics and bring you industry KPIs to really highlight what drives the company and brings you insights into future performance. This is just meant to give you guys a little bit of a visual of -- I was talking about the layer datasets and how we kind of build it out. So the idea here is we're taking, again, those broker models. We're finding the different line items that are relevant and then we're normalizing them at the company level. So now you can -- as my example said, iPhones for Apple, maybe Galaxy and other products for Samsung. But then the standardized layer allows you to do the real industry comparisons. Now for -- and this is where we get into really tight peer groups and being able to sort of look at like-for-like companies, the business lines that they do and this is something that you just don't see from typically traditional estimate providers in this space. So wrapping up, Visual Alpha really is, in my mind, the source of market expectations. We're going beyond EPS, net income and a lot of the high-level metrics that are out there. To just give you a complete view of the company, the markets that people compete in, the geographies and being able to really empower your analysts, like I showed on the last slide, being able to build those capabilities into the environment of your choice. So whether it's looking at some of the tools that we provide in our in our web app or it's doing some of the things that we have off platform and bringing the data into your own environment. So Dave, I'll turn it back over to you. I hope that was helpful for those, especially those who weren't familiar with Visible Alpha. And looking forward to connecting to people off-line if they need to go into more detail.

Yes. Thanks, Dave. I mean, really impressive in terms of full working broker models. I know that is something that isn't necessarily readily available. So definitely really unique content that individuals can do a lot more as they're bringing it together with all the other content that they utilize today. So as we think about that, right and we love to say that when individuals are analyzing data, they generate a lot more value in analyzing multiple datasets together versus 1 dataset in the silo. So since Visible Alpha is now part of the S&P Global family, can you walk us through the plans, let's just say, to integrate the Visible Alpha offering with the S&P Global content sets that we have today?

Sure. Yes. So I mean, starting off, the first thing that we've already done is integrate the symbology. So being able to connect and I'm going to show that in a demo later, being able to really connect from our ID to and find our company within the S&P ecosystem. So once you're there, you're able to combine our data with all the different data assets, which, of course, are very vast at the S&P level. What we're also doing and our content teams are hard at work is focusing on taking the CapIQ Estimates database into Visible Alpha database. Understanding what the differences in methodologies are, what are the different advantages that we can leverage from each one of these. So being able to say, Visible Alpha has longer forecast horizons, CapIQ has more depth of history in terms of how far back it goes. And just really taking those strengths and combining them into 1 future asset. And so the real vision of it is there's going to be 1 estimates database in the long run that will combine again the best of both worlds.

Really exciting. And then, I guess, for now, you've brought in the Visible Alpha data within the same Snowflake environment that we share within S&P Global, so our clients would be able to access the Visible Alpha data with the other S&P Global data through that single Snowflake share, is that accurate?

Yes. So you mean the same share. So that's coming out in the next month. We're going to have that available for customers and you'll be able to through [ Xpress ] cloud, get Visible Alpha data side by side with all of your other S&P content.

Yes, amazing. It's going to be great. So Chris, as you can see with this dataset, right, there's got to be some really interesting, I don't know, business use cases, personas that individuals can utilize this dataset, other datasets within Snowflake. So I'm just curious, what is Snowflake doing to really help clients with that whole persona business use case scenario with the data?

Yes. Thanks for the question, David. And conveniently, I happen to have a slide that goes into that. But first and foremost, I'd like to thank you both for the opportunity to be a part of this webinar. It always is full circle when being able to work with Standard & Poor's. And to be quite honest, I'm fortunate enough to have the role, in the seat that I sit in, kind of directing the wealth and asset management industry, how to leverage Snowflake but more importantly, how to take the datasets that are available through our collaboration feature in order to empower the business use cases that you actually see on the screen here. It truly was an executive MBA, I would always say, of learning how data flow through the financial services systems in particularly within wealth and asset management because of all the underlying content that Standard & Poor's Global Market Intelligence has and powers across not just wealth and asset management but insurance and banking and capital markets as well. So with that being said, to be a little bit more prescriptive to the question that was asked, what -- the 4 main drivers of use cases that we tend to see at Snowflake. So if we think about getting the data in for estimates, right, that's part of the quantitative research and investment process. Also part of that is the fundamental bottom-up approach, right, of leveraging Compustat CIQ fundamentals along those lines as well. So the main -- one of the main use cases that actually after data is moved into Snowflake and we will go into talking about Snowflake, not just as a method of transferring bulk, large big data files but also as the computation engine for advanced analytics, right, basically an advanced analytics operating system, which is how most of the industry truly uses Snowflake, right, to lever and orchestrate the complicated widgets, so most of the cloud providers to really start driving the time to decision, the security and governance needed in order to power a use case like quantitative research. But that's not all that we see. As mentioned, a lot of the, say, Standard & Poor's Ratings, the positions in holdings and things that come through, say, the RatingsXpress feed, they're tagged in regulatory reporting, things like [indiscernible], things like Form PF. So what we're really looking to say is that once you have this data in a single store, once it lands in the client Snowflake instance, so there's quite a few buy-side clients here, use that data once and then power these multiple use cases, right? Break down the fact that there's multiple copies of data all across your enterprise because at the end of the day, the regulators in this case, right, they look to see through all your funds, through all of your subsidiaries, right? What is the rating for IBM corporate debt, right, that you are holding? And it really needs to be the same across all of the portfolios or the regulators are then aware that there may not be a systematic way of reporting, right, into, say, FINRA or the SEC or whoever that may be. But as mentioned, one of the other main use cases that really drives and this is across all of financial services, is really risk management. And I'm fortunate to have learned so much about it again at my time being an alumni of Standard & Poor's, I guess, all of Standard & Poor's, not just Global Market Intelligence now, I think how long ago that has been. But really, we're talking about liquidity risk, counterparty risk, right, that most people generally just associate with Standard & Poor's but all of these data feeds actually go into various different types of risk management and even the client reporting and onboarding process. So once you truly have the opportunity and I've been fortunate to be in this industry long enough to really understand the interconnectivity of the data as it goes through each of these business functions, these enterprise functions, these are the things that really are exciting, that get us excited and more so lever the power of Standard & Poor's and all of the data that comes across through all of its content sets. If I may, David, just to move a little bit to the next slide that I have here because I really wanted to kind of reframe what Snowflake is for those in the audience, right? It's very challenging depending upon where you are over the journey and when you have had exposure to Snowflake to truly understand what it means to be an advanced analytics, the enterprise data cloud that we look to go to market to, right? How do we power enterprise AI, right, with what Snowflake is capable of doing and we'll get into that a little bit later. But if you take a look at this slide on the screen, the red box, right, the secure sharing, the distribution method that David had mentioned earlier, it is at most 5% of the platform, right? Most organizations truly use Snowflake to power their cloud data analytics and ensure the consistency, the breaking down of silos, the security and governance that comes with levering Snowflake, particularly through the collaboration and sharing feature. I don't want to get into too much of the rest of what is here. But truly, Snowflake is an advanced analytics development tool, right? It actually has the ability to run time series analytics, right, for tick data and things of that nature, it has the ability to bring in all of your Python packages, right and start running advanced analytics in different languages, not just SQL some people may think that is. So If you look at the bottom left here under the runtime piece, right, we're talking about bringing various different ways to compute data on pretty much every data type that exists, structured, semi-structured and unstructured, the ability to access data, not just in Snowflake but in, say, S3 buckets, in blob storage and things of that nature, bring whichever processor you need for that workload, i.e., as we get into artificial intelligence, not just the access to GPUs, the NVIDIA GPUs that can be levered throughout your cloud providers and then really keep building on to it to have more functions and features. And honestly, I believe there will be a demo actually that highlights a bit of it, leveraging one of our recent acquisitions, Streamlit. So with that, David and thank you again for the opportunity to present here again. I'll throw it back to you to see if there's any questions off of that.

No. I mean just a couple of things, right? When we started delivering our data to Snowflake, there was a ton of demand from the asset management space to access the data for that use case that you kind of highlighted. But since then, 5 years now, it has grown across all the different industry segments, personas in terms of utilizing Snowflake. And to your point, you're accurate. It started out with just data sharing to say, "Hey, simplify how you deliver data to us into the tools that we utilize." But looking at everything that you guys have done, because I do know that the competitors are constantly going after you, it seems like you're constantly innovating and trying to stay ahead of the competitors, which is great because everybody is trying to now copy a lot in terms of what you guys are doing.

Yes. David, yes, thanks for -- I was going to say thanks for highlighting that. And to your point, right, it's -- as mentioned, along the journey of Snowflake, right, post the IPO a couple of years back now, the development into the platform, which is basically to say, put -- once you have data in the cloud in a secure governed enterprise data cloud, of which Snowflake is the leader in that space, we're trying to then say, bring the compute to the data, instead of keep moving the data around your organization because that is ultimately what tends to lead to a lot of the challenges that we all are aware and experience and read on a daily basis, that occurs in this macroeconomic and political environment that we are all operating in, right? These challenging times that are the highlight of this webinar. And I unfortunately have what I call the cursive knowledge, which means I know the entire Snowflake platform. I develop in it every single day but not everyone, right, wakes up and reads all the marketing material, of which we have. So the opportunity to advise people, right, exactly, as you mentioned, this audience to say, Snowflake is more just than this, right? It's thinking about, okay, what processes do we have on-prem in a SQL server that are not governed, that are not GDPR compliant and things of that nature like it really winds up being a tall task, right, to truly understand how it all works. So the ability to stay paramount in the lead of not just data sharing, not just sharing code, right, our native application features and we'll get into that, if required. It's not just data, it's not just code but it's connecting the ecosystem, right? It's connecting Standard & Poor's to the wealth and asset management community without the need for SFTPs and APIs, right? And connecting to the regulators with reporting, so ultimately, what we call data liquidity, right? That's the main focus, is how do we ensure that more processes can be automated with better data quality and better breaks. It's easier if we're just looking into the Standard & Poor's database, which is really what data sharing does so that everyone is looking at the same data elements across the industry. It breaks down reconciliation processes for NAV reporting and custody reporting, everyone does know the last price as of the end of day of IBM's corporate debt. So that -- having been fortunate enough to have been around this industry long enough, those really wind up being a lot of the operational challenges that we all look to solve on a day-to-day after we do, say, the idea generation and the investments, it's everything post trade thereafter, that we also tend to help with. So thank you, David, again, for the ability to reframe that.

No. And you mentioned apps for a little bit, right? And I'll just jump right in. I know we're starting to get in a lot of feedback from clients on apps as well that can you deliver as your model, the app protects your IP, so we can't see that. You can see the data that we're incorporating into the model. So a lot of these new enhancements you guys are building, everyone is really excited both on the consumer end and on those that are delivering the data as well. So I love it. Dave, so as we start thinking about it, S&P Global and Visible Alpha had the same viewpoint, right? What can we do to take all our great content and deliver it to our clients seamlessly through the tools and the platforms that they want? And both firms, when we got together, we spoke and we said most of our clients are asking Snowflake, Snowflake. So what are you doing right now to try to really make it easy for clients to access the great Visible Alpha data through Snowflake?

Yes. Thanks, Dave. So I'm going to -- so I'm going to try to share my screen here and I have a number of different things I want to touch. It's going to be a little bit of back and forth but I hope to kind of build on a lot of what Chris is saying about how Snowflake is a lot more than just data sharing and a lot of things that we build on. I'm going to start with a really, a really simple example but it goes to the point about how important the power of linking your datasets is. And so I've got this really simple query here but we've done the work to get Visible Alpha into the S&P environment. And running this, we're going to bring back the -- and of course, there it is. So the idea here is, link the CapIQ company identifier to the VA identifier and be able to kind of move back and forth and be able to utilize the datasets across there. One thing I want to point out here, I mean a very simple worksheet but there's Notebooks, there's -- as Chris said earlier, Streamlit, there's a lot of different things, whether it's apps or tools that you can use on top of your data. So it's much more than a data warehouse. I'm going to quickly switch over to a different screen here. And so this is getting into a little bit more of what I was saying earlier about combining CapIQ Estimates as well as Visible Alpha Estimates. So I'm going to share my screen again. And so in this particular case, we have a product at S&P called Workbench. And Workbench is a way for us to showcase all the different datasets through Python Notebooks and all of this is built on Snowflake. So we have the Snowflake data sitting on the back end and then we're able to query that and put together all sorts of documentation. It's great for a product person like me because you can take use cases, sample queries and things like that and present them to the user to help them get familiar with our data. So I'm going to step through this really quickly. The idea here is, I was mentioning this earlier, CapIQ Estimates has a ton of history. So we brought back an example here, just pulling back 4 pharmaceutical companies, we're looking at revenue, EPS and EBITDA and we're bringing back that full time series for 20 years. So again, that's much deeper and further back than what VA is capable of. Switching over to VA, we have more forecasts and forward-looking data. So the idea here is, now I'm going to look at the next 15 years for those same 4 companies. What I did differently here, though is, I'm showing some of the depth of granularity that VA provides. So I'm actually looking at total revenue now for just things that help with oncology, just things that help with neurology and then finally, a pipeline view, which is saying all the drugs that are not in the market yet, what are we actually going to -- what are they forecasted to bring in down the road as they come to market. So bringing that data in, again, bringing that fully out for about 15 years. And then, of course, the last query here is just really bringing those 2 things together. So I just chose revenue here. But the idea here is now we have those 4 companies. We have a full time series going all the way back historically with actuals and then coming through from a forecast perspective. So again, simple but the idea was to sort of show what we're trying to do to combine and make these datasets very useful to you. My last example and this gets into app development is going to be around -- let's find it here.

David, compliments on doing live demos in a live webinar.

Yes, especially the way I had to go to kind of 3 different areas, but -- this is bringing up our Streamlit app. And this is -- it was totally an experiment that our data science team did. They came up with an algorithm, we wanted to highlight what we call significant revisions. So where was our outlier in sort of the revision time series. But it was really, what we wanted to prove was what are the capabilities of Streamlit. And so our team went through. And so I picked American Airlines here. I've got a number of KPIs in here from the Visible Alpha database, in this case we're -- I'm going to look at revenue passenger miles. And we're just looking at -- and this is just some of the capabilities of Streamlit when it comes to the kind of the widgets and things that you can build into it. So we've got a data frame here. I'll scroll down a little bit here -- sorry, on the wrong screen here. Some key performance indicators and then what I wanted to highlight here was this revisions chart. So the story this tells, as you go back to 2016, you've got revenue passenger miles growing over time but pretty consistent. You don't really see anything crazy in terms of outliers, COVID hits, travel comes to a grinding halt, this falls off a cliff and you've got these red dots that are highlighting what are some of the bigger negative revisions that were done in this. Things start to open back up, climbs back up and now you see some more green, some more positive outlier revisions. And then we steady state back here. And you can see even now that we're just catching up to pre-COVID levels, at least for American Airlines, with travel. And if I scroll down, I'm not going to go into more details here but we're doing regression analysis. So essentially, it's a tool, you've got dashboards, you've got all sorts of things but as well as analytics, depending on the use case and what you want to do here. My main takeaway here is really that when you modernize your tech stack, you can really unlock all sorts of possibilities. And the time to market for data analysis and app development shrinks significantly when you have that strong data foundation and then the tools on top of it to really empower your teams.