Mispricing Software, Misreading Ethereum

Last week’s selloff in SaaS stocks wasn’t subtle. But it also wasn’t a collapse.

It was a repricing driven by confusion, not fundamentals.

For years, SaaS was treated as untouchable. Recurring revenue meant durability. Switching costs meant safety. High margins meant inevitability. AI disrupted that comfort fast.

The market is now asking a question it conveniently avoided for a decade:
who actually survives the next three to five years?

From a corporate venture studio and VC perspective, this moment matters because fear is flattening judgement. Everything is getting thrown into the same bucket. That’s usually when mistakes get made.

Yes, AI makes code cheaper. Everyone knows that by now. But code was never the moat.

Context is the moat.

Context is proprietary data, embedded workflows, regulatory history, permissions, audit trails, and ten years of messy edge cases. You don’t LLM your way past that. You earn it slowly, or you don’t get it at all.

This is especially true in software tied to financial infrastructure. Salesforce remains critical not because of features, but because it sits directly inside revenue workflows and customer data. The same applies to companies closer to the future of finance like Intuit, Fiserv, Block, and PayPal.

Their advantage isn’t innovation theatre. It’s trust, distribution, and regulatory depth. What is breaking is the pricing model. Seat-based pricing is dying.

We’re moving toward labour replacement and outcome-based economics. A company that used to pay twelve hundred dollars a year for software supporting an eighty-thousand-dollar employee will pay several thousand dollars a year for an AI agent that replaces that role outright.

That’s massively deflationary for customers.
And structurally expansionary for software companies that get it right.

Software spend will increasingly replace payroll spend. If public markets want to punish incumbents in the short term, fine. Long term, software only becomes more valuable.

The mistake is modelling margin pressure without recognizing how much bigger the revenue pool just became. This isn’t the end of SaaS. It’s the end of lazy assumptions.

The same misunderstanding driving SaaS volatility is spilling into crypto.

ETH and BTC sold off hard despite continued institutional progress. That disconnect matters.

Price is reacting to headlines. Adoption is moving quietly in the background.

Vitalik Buterin recently acknowledged what many builders already understood, saying that “the rollup-centric roadmap is no longer sufficient on its own” and that Ethereum will focus more on execution, usability, and real throughput. That wasn’t a reaction to markets. It was a technical admission that the next phase is about shipping, not architecture diagrams. That clarification matters. It forces Ethereum and Layer-2 assets to be judged on real usage, security, and economic alignment rather than abstract narratives.

At the same time, AI capital expenditure is accelerating aggressively. GPUs. Data centres. Inference infrastructure. This always happens before application value shows up.

Fibre optics came before the internet.
Compute always comes before agents.

Jensen Huang addressed the software selloff directly, calling it a “category error” and noting that “AI doesn’t eliminate software, it makes software dramatically more valuable.” His point was simple. AI expands what software can do. It does not shrink the software economy. Nvidia isn’t selling into a declining market. It’s supplying the foundation of a much larger one.

Crypto markets missed another obvious signal. Fidelity launched a stablecoin on Ethereum. BlackRock continues expanding BUIDL. Franklin Templeton’s on-chain funds keep growing. Visa is settling real transactions on-chain. These aren’t pilots. They’re operating decisions by institutions that don’t move early or loudly. TradFi isn’t debating Ethereum anymore.

It’s integrating it into production systems. When adoption moves forward and price moves backward, CIOs should pay attention.

On Tuesday, Feb 3rd, 2026, I spoke at VNTR Toronto to a closed room of global investors about where AI, Ethereum, and finance collide. The framing was intentional.

What sounds like finance is actually software.

Money still runs on systems designed decades ago. Slow. Fragmented. Manual. Expensive. What’s changing isn’t money itself. It’s the software underneath it. For most of the last decade, on-chain finance, payments infrastructure, and regulated fintech were bad venture categories. Regulation was unclear. Procurement cycles dragged on for years. Buyers didn’t exist in meaningful numbers. Exits were questionable. Even great teams struggled because institutions simply weren’t ready. What changed wasn’t ambition. It was timing.

Regulatory clarity reached a point where institutions could act without career risk. Banks and asset managers started allocating real operating budgets to modernization. And incumbents finally accepted that rebuilding core infrastructure internally is slow, risky, and expensive.

That’s the shift from experimentation to procurement. AI fits here as the decision layer. It sits on top of financial infrastructure where money, risk, and compliance intersect. Monitoring transactions. Detecting fraud. Automating approvals. Optimizing liquidity in real time.

Ethereum defines what is allowed to happen.
AI decides when and how it should happen.

From a CIO perspective, this is the inflection point. The timing is finally right to invest in on-chain infrastructure, evolved workflows, and modern payment-rail ecosystems that institutions are now structurally prepared to buy.

For high-net-worth individuals, family offices, and strategic investors, the question isn’t whether this transition happens.

It’s how exposure gets constructed across private and public markets.

Ethereum should be viewed as shared financial infrastructure, not a speculative trade. Institutions don’t care about decentralization ideology. They care about standardization, interoperability, auditability, and control. Ethereum delivers that.

Tokenization follows naturally. Not new assets. Better plumbing. Existing financial assets turned into software objects that move faster and reconcile automatically.

Stablecoins come first because they solve a real problem. Faster digital cash enables instant settlement, automation, and software-to-software transactions. That’s why adoption is steady and understated.

Returns in this space won’t come from hype-driven IPOs. They’ll come from acquisitions, structured secondaries, balance-sheet integrations, and new liquidity off-ramps. Strategic buyers want capability. Not noise.

The opportunity sits at the intersection of private-market innovation and public-market infrastructure. Software is being repriced. Ethereum is being misunderstood. And the future of finance remains intact for investors who can separate signal from noise.

Our Future of Finance report is now live:
www.highlinebeta.com/eth