Why This Cloud Slowdown Is Temporary and Why This Presents an Opportunity

Cloud infrastructure stocks have been among the worst performers this year despite all the excitement around artificial intelligence (AI). In this report we dive deeper into the drivers of the current underperformance and highlight the medium term upside potential. 

Contents:

• Enterprises are cutting spending and optimizing cloud usage
• Not all cloud business models are the same - consumption vs. project based  
• AI breakthroughs will drive strong medium term cloud demand 

Cloud spend optimization 

There are two effects that are driving a temporary slowdown in enterprise spending (1) fear of recession is forcing companies to cut spending (2) after two years of robust growth, companies are optimizing their spend. 

Fortune 500 decision makers expect a dramatic slowdown in IT spend. Per a recent ETR survey, respondents expect 2023 IT spend to grow only 1.4% vs. 4.4% yoy only 2-3 months ago (based on data from over half of the Fortune 500).

Interestingly, cloud spending is expected to fare worse than the average IT spending. In addition to broader slowdown in IT spend, companies are optimizing their cloud usage and delaying new projects.

Per Microsoft's F2Q23 , the company expects the optimization to last several quarters (but not years). Microsoft noted that in addition to the delays, "when the new projects start they don't start at peak usage...They start and they scale. So the two cycles are creating an air-pocket in demand."

Not all cloud companies are the same 

The global spending on public cloud services totaled $490.3bn in 2022 according to Gartner.

• Cloud infrastructure (PaaS) and (IaaS) represented ~50% of the total ($227bn). The hyperscalers (AWS, Azure, GCP) represented 70% of that.
• SaaS applications represented 39% ($167bn). The use cases for SaaS are very broad (from CRM systems to design software). Within SaaS there are several new entrants that sit on top of the cloud infrastructure platforms and provide ways of using the platforms more efficiently and securely e.g., separating storage from compute, sharing data without having to store it twice, etc. 

Consumption based businesses models were the first ones to go into the downturn and noted deteriorating demand since 1H22 (e.g., Snowflake pointed to optimization efforts, Datadog noted weakness in their logs business). As we went through the year, the consumption side stabilized at lower levels, but the headwinds broadened to the project side where recession fears are influencing investment decisions (e.g., hyperscalers just started noting demand softening in 2H22). 

We believe that as the hyperscalers may be facing 1-2 more quarters of weaker demand, the consumption based models will start facing easier comps.

Interestingly this cyclical downturn in IT spend is coming at a time when Artificial Intelligence (AI) adoption is reaching an inflection point, creating a meaningful medium/long term tailwind for these companies. 

Strong medium term outlook driven by AI inflection 

AI is undergoing a significant breakthrough driven by several successful applications of a new foundational model (Transformer) in large language model training. The Transformer model is a neural network that learns context and can therefore be used for "generative" rather than "predictive" applications. 

We believe that this will be transformational for many industries and has already generated a fair amount of hype. But what investors are not appreciating is the amount of data and compute power that these applications will require which will likely be provided in the cloud.

According to an ETR report surveying companies and budgets, the cloud spending CAGR (for PaaS and IaaS) may be as high as 67%. Among all respondents, the median annual spend on IaaS/PaaS is currently $375,000 and is expected to increase to $1,750,000 in 3 years. Looking at large organizations, spend is currently $750,000 today but is expected to increase to $3,750,000 over that same time period. 

These spend projections tie to commentary that we are hearing from companies, such as Snowflake. Once their costumers see what they can do with their data, the spend grows exponentially. While we are using a more conservative ~40% growth estimate, we believe that there is room for upside.

Another interesting observation from the same survey is that respondents expect that while AWS and Azure will remain the top platforms, there will be an opportunity for new entrants and smaller players to gain market share.

The views and opinions expressed herein are the views and opinions of the author and do not necessarily reflect those of Nasdaq, Inc.

Other Topics

Cloud Computing Companies