Availability Cascade

In the AI Story

The Death Cross chart that triggered the 2026 SaaS repricing functioned as a near-perfect catalyst for an availability cascade. It possessed every feature that makes claims cascade-prone: extreme vividness (a single image telling the entire story), strong emotional loading (the word 'death' applied to a three-trillion-dollar industry), radical simplicity (two lines crossing, requiring no analytical background). It also possessed the feature that distinguishes dangerous cascades from harmless memes: it was partly true. AI genuinely was commoditizing code; pandemic-era SaaS multiples genuinely were unsustainable. The cascade component was the gap between what the evidence supported — the specific observation that SaaS companies needed to demonstrate ecosystem value rather than code value — and what the market came to believe: that SaaS as a category was dying.

The reflexive property of financial cascades makes them particularly destructive. In most domains, cascade-driven belief distorts assessment without changing underlying reality. In financial markets, the belief changes the reality: cascade-driven conviction that valuations are collapsing produces selling pressure that reduces valuations, which is then interpreted as evidence that the conviction was correct. The circularity is invisible to participants, each of whom experiences the falling price as independent confirmation rather than as consequence of the same narrative operating through millions of simultaneously acting agents.

The affect heuristic compounds the availability heuristic. Vivid claims trigger emotional responses; emotional responses are then interpreted not as reactions to the presentation but as assessments of underlying probability. Investors responding to the Death Cross chart did not experience themselves as responding to an emotional trigger. They experienced themselves as rationally evaluating evidence. The substitution was invisible — and the invisibility is the mechanism's most dangerous feature.

The corrective to an availability cascade is not more information. Providing accurate, nuanced analysis to populations in the grip of a cascade typically has negligible effect, because cascade dynamics dismiss or reinterpret contrary evidence. The corrective is institutional: structures that insulate decision-making from cascade dynamics — circuit breakers, cooling-off periods, disclosure requirements that force granular analysis cascade narratives suppress, independent expert assessment insulated from public opinion.

Origin

The concept was developed jointly by legal scholar Timur Kuran and Sunstein in their 1999 Stanford Law Review article 'Availability Cascades and Risk Regulation,' which synthesized decades of research on the availability heuristic (Tversky and Kahneman) with Kuran's work on preference falsification to produce a framework for understanding how public beliefs form, spread, and persist independently of evidential support.

Key Ideas

Vividness substitutes for evidence. Claims that are easy to visualize and simple to communicate achieve truth-status through repetition rather than verification.

The mechanism is self-sustaining. Each repetition increases availability, which increases perceived credibility, which generates further repetition in a closed feedback loop.

Financial cascades are reflexive. In markets, cascade-driven belief changes the reality it purports to describe, creating circular confirmation invisible to participants.

Information alone cannot correct cascades. Counter-evidence is dismissed or reinterpreted; only institutional structures that insulate judgment from cascade dynamics provide reliable correction.

Debates & Critiques

Whether availability cascades are distinctive phenomena or merely special cases of broader conformity effects remains debated. Sunstein's view is that the specific mechanism — cognitive availability mistaken for evidential support — identifies a distinct causal pathway that other conformity frameworks miss. Critics note that the concept can be applied too liberally, labeling any widely held belief one disagrees with as 'cascade-driven.' The operational test — whether the belief would survive structured engagement with counter-evidence — provides some discipline but remains imperfect.