Anthropic and General Catalyst: the market is mispricing resilience

In 2001, Amazon's stock had plummeted by nearly 95% from its peak, and many of its dotcom competitors were facing liquidation. However, Amazon had diligently constructed fulfillment capacity well ahead of market demand, expanding from 300,000 square feet to over five million square feet. This excess capacity would prove to be a valuable asset when the market crash occurred.

Rather than being a liability, it transformed Amazon's idle warehouse space into a platform business by opening its network to other retailers and independent sellers. In the past century, resilience was a valuable quality, often rewarded only in hindsight, as it meant withstanding disruption. But as disruptions become more frequent and diverse, the companies that can absorb shocks most rapidly are the ones that experience growth while competitors retract.

Artificial intelligence (AI) is accelerating this trend. While AI poses challenges to incumbent business models, it also provides companies with powerful tools to absorb shocks. Companies running on AI-enabled processes are more cost-effective and have greater flexibility when circumstances change. One of the most significant errors investors make is confusing "still growing" with "structurally sound."

Although recurrence can be a sign of a resilient business, it is not the same as durability. The challenge lies in identifying resilience before disruption becomes apparent, as its true value often becomes evident only after a company has already absorbed the shock. This gap in market perception can be observed in four key areas.

First, demand resilience. AI is revealing how contingent some recurring software revenue truly is. Software embedded in core processes and leveraging proprietary data has proven more resilient than software sold individually, with customers able to adapt as AI replaces traditional software roles. Investors, however, have traditionally evaluated companies based on sector labels rather than the depth of their market position.

Healthcare, defense, and energy industries fall into this category, where regulatory barriers and institutional trust make them less susceptible to digital disruption. These sectors also tend to experience increased demand during crises, as essential services become even more critical. Second, supply-chain resilience. While markets react swiftly to disruptions and often fail to recognize the value of companies investing in prevention, some firms recognize the importance of building flexibility into their systems.

For instance, Aramco spent years enhancing its system's flexibility, including constructing the East-West Pipeline and multiple storage options. This investment paid off within days of the Iran war and subsequent disruptions in the Strait of Hormuz, solidifying Aramco's position as one of the most dependable oil suppliers globally.

Third, capital-stack resilience. Private-equity sponsors invested heavily in software companies in 2021, often financing their acquisitions with substantial debt, assuming recurring revenue would remain predictable enough to service the loans. However, not all recurring revenue is created equal. The duration of that debt is fixed, whereas the revenue underlying it can be highly variable.

AI has exacerbated this disparity, shortening revenue cycles unevenly. Software deeply integrated into customers' operations has demonstrated greater durability. This difference has been evident from the outset but often overlooked by investors focused on recurring revenue rather than the underlying factors driving that revenue.

Lastly, system resilience. Companies can appear resilient while, in reality, relying on external systems to absorb shocks. Electrical grids, for example, are designed with ample capacity to handle disruptions, and AI data centers are increasingly relying on this capacity. This shift transfers risk from individual companies to the broader system.

The North American Electric Reliability Corporation has recently increased its forecast for peak electricity demand and issued its highest-severity alert due to a sudden surge in computing demand. The most resilient companies will invest in strengthening the systems they depend on rather than viewing spare capacity as their sole advantage.

This principle extends beyond infrastructure to organizational structure. JPMorgan Chase has built resilience through a well-capitalized balance sheet and a diversified business portfolio beyond investment banking, enabling it to weather periods of stress that left less diversified peers vulnerable. Building resilience, however, comes at a cost.

Redundant capacity, stronger balance sheets, and operational flexibility all require investment before they are needed. While these investments provide optionality and enable companies to continue operating when conditions change, the market often fails to recognize their value until after disruptions have occurred. Amazon in 2001 exemplifies a company that embodied these qualities, investing in capabilities that would later prove invaluable during the market crash.

Today, many more companies share this approach. The market's blind spot lies in its tendency to price resilience's cost before acknowledging its true value. Companies invest years in building resilience, which can negatively impact short-term returns while its full potential remains hidden until disruption strikes.