AI is coming for software stocks

AI is changing the economics of software development, and the market is starting to price in what that means for SaaS stocks.

---

There’s a good article on Livewire from Matt Joass about the risk to software stocks – the revelation is that coding skills are now unnecessary, rendering a generation of programmers effectively redundant.

From a stock market point of view, whilst you can argue the toss that companies like Xero (ASX: XRO) have much more to offer than “just” a software application, the reality is that even for the companies that cannot be “replaced” by AI coding, AI is going to compress margins, lower barriers to entry, invite new (cheaper) entrants to compete, and require significant capex from incumbent software companies as they incorporate AI within existing products, possibly for no extra revenue.

It’s not great, there’s a lot of uncertainty, and it’s a long-term thing, not a “moment” of fear. It’s real.

We have already seen significant repricing in software stocks, and whilst there is a trade in buying the bottom, the likelihood is that the industry is forever changed and threatened, and in the reshuffling, there will be significant losers, and “winning” will be little more than surviving.

TechnologyOne (ASX: TNE) CEO commentary this morning offers hope there will be significant winners. We can see the BetaShares S&P/ASX Australian Technology ETF (ASX: ATEC) in a sentiment hole that for some companies is almost certainly overdone.

 

How AI is changing the software industry

AI has made writing code cheap and abundant. Since OpenAI released ChatGPT in 2022, and later reasoning models like o1 plus tools like Anthropic’s Claude Code, software development has shifted from slow, manual effort to AI-assisted (and increasingly AI-driven) production.

Coding is no longer the bottleneck. What once took months can now be built in days or even hours using AI agents. Software is not just infinitely replicable at near-zero cost – it is now increasingly producible at near-zero cost.

Stack Overflow’s decline symbolises the shift. Developers once relied heavily on forums like Stack Overflow to solve problems. AI coding agents have largely replaced that workflow, dramatically reducing traditional friction in software development.

 

AI threats to SaaS business models

Threat #1: Customers building their own software

As AI agents improve, more businesses may choose to “build” rather than “buy” software. What started with simple “point solutions” could move up the complexity ladder over time.

The build-vs-buy line is moving. AI raises the “water level” – increasingly complex tools could become feasible to self-code internally, particularly for large enterprises with in-house tech teams.

Threat #2: Ultra-lean startups undercutting incumbents

Small teams using AI agents can now replicate large SaaS products at far lower cost, reducing pricing power and making it harder for incumbents to push through 10–15% annual price hikes.

Threat #3: Horizontal expansion from tech giants

Large platforms could more easily expand into adjacent SaaS markets. If software is cheaper to build, companies like brokers, fintechs, or payment platforms can bundle new products into existing customer bases.

Threat #4: AI-native competitors

Instead of just building cheaper SaaS, new entrants may redesign the entire “job to be done” using AI-first models – automating workflows so thoroughly that traditional SaaS interfaces become secondary.

Threat #5: “Pure agentification” (the omega threat)

A future “Super Agent” (e.g., via OpenAI, Anthropic, Google, or xAI) could sit above all applications, directly managing tasks across banking, accounting, payroll, and more – potentially disintermediating SaaS providers entirely.

 

What this means for software stocks and ATEC

SaaS isn’t dead, but AI adoption is going to cost existing companies (pressuring margins) and lower barriers to entry for competitors of companies with no moat. Competitive intensity will rise. Some businesses will be structurally weakened, new leaders will emerge – we are in the “disruption cycle”.

It’s not yet clear which companies will emerge as winners (survivors) and which will fold. For now, ALL software companies are being lumped in together, and the loss of value has been seismic and, doubtless, for some companies with moats, overdone.

We’ll wait for the sentiment change and see if we can catch the bottom.

 

ATEC chart

The Australian All Tech Sector, which includes our software stocks (and many others).

 

ATEC stocks