The software market is threatened by AI. An investor from The Big Short on stocks to avoid
The investor from The Downgrade Game analyzed 46 companies from the software and payments sector using his own methodology - and concluded that even after the so-called software apocalypse, there are no good buy options in the office software sector
AI creates two big risks for software developers at once - software replacement and job elimination, Burry writes / Photo: Shutterstock.com
The sell-off in the SaaS sector has not made software stocks attractive, says Michael Burry. Securities cannot be evaluated only through the scale of the fall: many of them still do not provide sufficient potential yield. To see where long-term opportunities remained after the collapse, the investor from "The Downgrade Game" selected 46 companies from the software and payments sector and evaluated them using his own system.
Burry describes his analysis in a series of six posts on the Cassandra Unchained blog on Substack. The first focuses on office SaaS - software developers for HR, finance, IT, sales and customer support. These companies have long traded at a premium as the beneficiaries of businesses moving to cloud solutions. Burry analyzes how sustainable this model is in a business where agent-based AI is already taking away some of the functions of enterprise software.
Three questions for each share
The investor checks what proportion of the reported profits are really attributable to shareholders. To do this, the investor first recalculates the company's reported earnings into owner's earnings: taking into account the real value of employee share-based payments, tax effects, interest on excess cash, lease payments, depreciation, amortization and other costs that affect the economic result for shareholders.
In valuing a business, Burry separately considers the debt load, future purchase commitments, and capital that the company actually puts at risk. He focuses on revenue and margins in his forecasts because he believes the other metrics are easier to distort.
Finally, Burry compares the current stock price to IV15 (Intrinsic Value 15), the estimated level at which a security can yield an investor 15% per annum over a horizon of 15 years or more. If a stock is trading above IV15, it's not a buy zone for Burry: even a strong drop from the peak doesn't mean that the security offers the right margin of return.All of the office SaaS companies Burry analyzes in the first note have a current share price above IV15: from 1.23 to 5.45 times.
Burry rates each company on a 100-point scale. And even the best in the first list - Paycom and Freshworks - are ranked 7th and 8th in the overall rating, but don't score 60 points. To him, this is an indicator of weakness in the entire office SaaS group.
"I wouldn't go short on any of these stocks. Right now, it is enough to play up where there is a probability of success, stay away from stocks that are known losers and perhaps look at those securities that are on the edge of attractiveness. Investing is a game of absolutes. Choosing from this sector is like looking for the cleanest pig in the pigsty. I haven't gotten into office SaaS developers yet, but I have clearly defined price targets for myself and know what metrics to look for in their quarterly reports"
AI threat
In the analysis, Burry sees AI as one of the key factors in evaluating office SaaS. He sees it as a risk that could change the economics of companies.
"The two big risks for software companies are software replacement and job elimination: the SaaS model was built on a per-job basis at many, if not all, software companies. AI threatens to dramatically reduce the number of jobs in some cases"
To assess this risk, Burry looks at how much control the company itself has over AI technology, whether it has a strong R&D team and whether it can withstand pressure from cheaper AI services. By this logic, he scales from the most secure businesses to the most vulnerable: the former are using AI themselves as a competitive weapon, the latter are dependent on other people's models and may lose their previous technology valuation premium over time.
Burry's Office SaaS Rankings
Paycom | 7th place
Burry singles out Paycom because of its treatment of shareholders. He writes that over 11 years, Paycom's picture of the real value of employee stock-based compensation is one of the cleanest among office SaaS companies. In addition, Burry believes Paycom is more resilient than its competitors because of the specifics of its payroll business. The company stores the data on which payroll, HR processes and interactions with regulators are based, and revenue depends on the number of employees at clients.
Freshworks | 8th place.
Freshworks, Burry writes, looks to be a cheaper and easier-to-install solution with its own AI products. The company is launching a Model Context Protocol Gateway that will allow customers to use Claude and ChatGPT with Freshworks data. Burry calls it a "delicate dance": the company gets to monetize that traffic, but it also lets external tools into its ecosystem that could weaken the SaaS model.
Salesforce is 17th.
Salesforce is betting on Agentforce, the agent-based AI inside its platform. According to data Burry cites, the product closed 29,000 deals in the first 15 months and generated about $800 million in revenue. But for Burry, Agentforce's scale isn't enough yet: on the back of Salesforce's above $40 billion in revenue, the product doesn't yet prove that the company can offset the pressure of AI on its core business. Burry isn't sure Salesforce can sustain growth and margins while competing with Microsoft, Oracle, ServiceNow and other AI services.
Paylocity | 20th place
Paylocity's revenue depends on the number of employees at customers, so the main AI risk here is downsizing, Burry says: the fewer employees at customers, the smaller Paylocity's revenue base. In addition, the company's competitive position is weaker than Paycom's, Burry estimates, and the stock is overpriced: Burry calculates that it trades at 3.54 times IV15, so a buyback does little to change the investment picture
Maunday.com | 27th place
Bjurri considers monday.com the absolute worst in the group for employee benefit stocks - and therefore puts the company in the middle of the list. A high expected yield of 12.8% keeps the stock on the waiting list, Burry writes, but it clearly falls short of unconditional buy status.
ServiceNow | 28th place.
ServiceNow, in Burry's estimation, is actively moving away from paying for user seats more than others: half of its business is already built on pay-per-use, and Now Assist allows the company to capitalize on the work of AI agents inside corporate processes, from IT and security to CRM and HR. But this realignment makes it harder for an investor to analyze ServiceNow. Burry believes margins now need to be monitored separately. If AI products grow and margins decline, it will show that the new revenue is costing the company more than before and not improving the economics of the business.
HubSpot 41 places.
For Burry, this is an example of a SaaS company that is particularly vulnerable to AI.HubSpot sells tools for CRM, marketing, sales, service and content to small and medium-sized businesses - the very functions that agent-based systems can already partially take over. Yet over the past ten years, the company has not created economic value for shareholders, Burry calculates, with owners' profits totaling minus $3.34 billion.
Workday is 45th place.
Burry attributes Workday's AI risk to its sales model: if agents can do some of the tasks in HR and finance processes, customers may need fewer paid users within the system. The company is trying to offset this risk through pay-as-you-use, but Burry thinks this defense is weak. In his estimation, Workday doesn't show that the new AI monetization will be able to absorb the potential loss of revenue and maintain the same margins.
This article was AI-translated and verified by a human editor
