Investment Thesis
Snowflake (NYSE:SNOW) is a globally renowned data technology company whose data transactions and management platform is very popular among customers who have anything to do with data, especially large enterprises. The company’s stock is also commonly held among many institutional clients, such as Berkshire Holdings and Bridgewater Associates.
Over the past 3 months, Snowflake has enjoyed a terrific run, rising almost 60% from November since a reacceleration in cloud spending forecasts turned around the growth prospects of many technology companies. On the back of such a massive rally, I believe valuations are stretched and premiums have expanded exuberantly, with investors being too optimistic about Snowflake’s earnings. I rate this stock as a hold for now and have explained my reasoning below.
Snowflake’s next chapter of the data platform growth
Snowflake was one of the first companies to offer database-as-a-service (DBaaS), a cloud-based industry that was quite nascent in the last ten years but has exponentially grown since then. In this period, Snowflake and other standalone data platform companies such as MongoDB (MDB) and Databricks disrupted the legacy database market that was traditionally on-premise.
Snowflake quickly pivoted from its on-premise offering to a hybrid cloud product suite, allowing enterprise users to securely connect, query, and transact with their own enterprise data in Snowflake’s cloud. The ease of use and rapid expansion of functionalities made the company’s data products extremely popular, which led to Snowflake’s meteoric growth.
However, with automation and AI now taking center stage, every company in the world has reviewed its long-term strategy to stay relevant in a new world. Snowflake appears to be swaying toward increasing the modalities of accessing data on its platform, with growth pillars I will discuss later.
Allowing users different modes to access data could be a genius move by the company while staying relevant in the age of AI, especially because the company uses a consumption-based revenue model that awards customers credits per future use. This appears to be the new model moving forward that most AI companies have adopted, including ChatGPT, etc.
Increasing modalities of data access increase consumption and revenue growth rates
I mentioned earlier how Snowflake is increasing the modalities of user access to its data with the goal of driving up data consumption. This strategy could be accretive to Snowflake’s future growth model, given that higher consumption leads to higher consumption fees, leading to stronger product revenue growth. Another way that Snowflake also earns revenue is through seat expansion. For every new user that an enterprise adds, Snowflake charges a fee and awards platform credits to the seat or new user for capped future use.
So far, the company has been very successful in rapidly penetrating an oversaturated on-premise database market whose disruption was overdue. Therefore, it made sense to me that a majority of Snowflake’s customers, whom the company had acquired in the last few years, would be focused on just migration workloads, i.e., migrating data from their old database systems, such as Teradata or Oracle, to Snowflake. But I suspect Snowflake had long planned for this moment where platform spending would flatline or, worse, decline. So the company worked towards building more features and functionalities for its customers to do more with their data than just migrate and transact with it. These trends can be seen in the dollar-based Net Retention Rates (NRR) which are still one of the strongest NRR trends seen in technology companies.
While the general trend in Snowflake’s NRR is decelerating, its last known NRR of 135% is still one of the strongest in the technology industry, showing customers are spending more than they did vs. the same period last year, just not as strongly as they did earlier. Another high-growth technology stock, Datadog, last reported its NRR at ~120%.
Although inflation and high interest rates have definitely dampened the growth rates for the entire technology industry, I believe the surge of AI and ML enterprise adoption has forced Snowflake to fast-forward its product roadmap or risk a de-rating. After all, for Snowflake to trade at ~205 times this year’s forward consensus EPS estimates or 138 times next year’s forward EPS estimates is not cheap, in my opinion.
Therefore, Snowflake has embarked on a couple of growth pillars in the post-AI era to allow customers to do more with their data and to spur adoption, consumption, and eventually revenue. In June 2022, Snowflake launched Snowpark, which allowed developers to unlock the power of their enterprise data using familiar programming languages such as Python in addition to the regular SQL-data language. In addition, Snowflake also added data sharing features, which appear to be an emerging segment within the DBaaS market, as per Gartner. Data sharing allows owners of data on Snowflake’s platform to securely, lawfully, and safely sell highly anonymized data to customers on the Snowflake network. Snowflake’s direct competitor, Databricks, has already moved quickly to launch its own data-sharing services. But Snowflake has wasted no time and is moving to form partnerships, most recently with Salesforce, to launch data sharing features using anonymized Salesforce data. I have attached a modified screenshot from Snowflake’s presentation, where I have also added y/y numbers below for Snowflake’s Data Sharing and Marketplace listings.
Moreover, the company appears to be doubling down on its strategy to launch data sharing by acquiring a young startup, Samooha, that operates in the data sharing space. So far, these strategies have not really moved the needle on re-accelerating growth in Snowflake’s revenue trends, as seen below. Worse, its Remaining Performance Obligations have flatlined for many months now, which is a sign that customers may not be spending more on the platform in, approximately, the next twelve months or so. I expect the company to update its outlook for investors in the full-year FY24 call when Snowflake reports later this month, on February 28.
Valuation suggests Snowflake is richly valued
Using Snowflake’s long-term operating model, I estimate that Snowflake will grow its revenue at a CAGR of 24% to reach its FY29 revenue target of $10 billion. This is also based on Snowflake’s own forward estimates of their FY24 guidance, as reported in their Q3 FY24 earnings presentation. However, in my view, the real threat in their guidance lies in their non-GAAP operating margin estimates, which are projected to expand from 7% in FY24 to ~25% in FY29. Per my calculations, non-GAAP operating income should be growing at a CAGR of an enormous 54% over the next five years. These kinds of forward growth rates are very impressive, given that the long-term earnings growth rate for the S&P 500 is 8%.
After applying a discount rate of 8% on future earnings and using their current share count, I arrive at a forward estimated share price of $230, assuming a forward PE of 45. I think a forward PE of 45 is fair given the impressive growth rates that Snowflake needs to maintain. However, it appears investors have already priced in all of the company’s future growth, since Snowflake is currently trading at its fair value.
Risks & Other Factors to consider
With no other speaking engagements from Snowflake’s management scheduled for the next two weeks, it appears investors will have to wait for more official guidance and outlook in the earnings call at the end of the month. I will be keenly looking for any additional information from management about adoption rates and trends in their Snowpark and data-sharing products. I will also be carefully listening for management’s perspectives on the broader macro outlook in terms of the economy as well as within tech enterprise spending. In November last year, Gartner projected cloud spending to rise ~20% y/y this year, which is slightly higher than the FY23 17.8% growth in projected cloud spending.
Conclusion
Snowflake has been able to ride the unprecedented AI hype cycle since November last year, but I think the rally to current levels has left no room for any further upside and makes the company vulnerable to a selloff since it is richly valued. There is no doubt about the popularity of Snowflake’s products or about Snowflake’s excellent management, which has a rich history of driving high-performance teams in the past, but valuation remains a big question here as things currently stand. I rate Snowflake as a HOLD.