Model Portfolio Update: Interpreting the Software-Led Market Downturn

Following Microsoft’s (MSFT) earnings release, the software sector (including Salesforce, ServiceNow, Workday, etc.) came under heavy selling pressure, with share prices falling across the board regardless of individual company fundamentals.

According to Bloomberg, the iShares Expanded Tech-Software Sector ETF has fallen more than 20% year-to-date, wiping out over $1 trillion in value. The problem is that the downward pressure initially concentrated in software is now spilling over into the broader information technology sector, driving a sharp market-wide decline.

One widely cited reason for the heavy selling in software is that over the weekend, Claude language model developer Anthropic announced a product specialized in legal advisory. The prevailing interpretation is that AI-driven productivity gains will slow the profitability of SaaS (Software as a Service) companies that have long enjoyed high margins, while in-house development by their customers will reduce demand for their products.

I already discussed this inevitable shift in the software industry early last year:

However, as I noted in that piece, it is an overreach to conclude that the software industry is facing imminent doom. It is true that AI has exponentially reduced the cost of building basic software, but the risk of replacing software that companies have relied on for decades with AI-generated code is far greater when you consider the levels of reliability, stability, and scalability they require. Publicly listed software companies, whose revenues are largely driven by large-scale B2B contracts, are unlikely to see their earnings take a major hit in the near term.

That said, corporate procurement teams are now far more likely to demand products that are fully featured from day one and come with proven stability. For startups that lack existing sales networks and must build products from scratch, the development costs required before launch, as well as marketing and sales expenses after launch, are likely to increase.

This is where the elephant in the room appears. Software companies can no longer lean on an “asset-light” business structure. In the early 2000s, Silicon Valley VCs justified their investments in software startups with the logic that while upfront development costs were high, once the customer base surpassed a certain threshold, operating leverage would improve dramatically, leading to high profit margins—and that thesis proved correct. Software development costs are akin to fixed costs (later capitalized as intangible assets), but once a product is built, it can be sold to countless customers. Unlike manufacturing, there is virtually no additional cost tied to higher sales volumes, which means the marginal profit (the extra profit from selling one more unit) is extremely high.

Recently, however, as barriers to entry in software development have fallen, product supply has surged, while the marketing and sales costs required to win increasingly demanding customers have risen. The model is no longer one where you build a product once and then simply sell it hard. Instead, it has become a capital-intensive structure in which costs are incurred continuously—from product planning and development through sales and ongoing maintenance. As a result, the valuation premium that SaaS companies enjoyed peaked during the liquidity euphoria of the 2021 pandemic era and has since fallen sharply.

In my view, the narrative that the software sector entered a steep downturn because of Anthropic’s launch of a legal advisory product is largely a post hoc explanation crafted to fit the price action. We have already seen the release of language models with strong coding capabilities, and there have been CEOs who claimed they boosted productivity by developing software in-house while laying off staff. The fact that this legal advisory product launch is being singled out as the main trigger suggests there was no more directly connected piece of news to point to. That is precisely why I see the indiscriminate sell-off within the sector as an opportunity.