Goldman Sachs' trading materials mention that it recently launched a long-short portfolio for software, redefining investment logic for software companies based on the difficulty of AI replacement.
The positive side includes software requiring physical execution, facing high regulatory barriers, complex system integration, or requiring retained human accountability, as well as areas directly benefiting from increased AI usage, such as computing power, data infrastructure, observability, security, cloud platforms, and AI development tools.
The other side focuses on process-oriented software: if existing tasks can be gradually automated by AI, and companies can leverage AI to rebuild some internal processes, their reliance on external software and outsourcing services may decrease.
This means that the valuation logic of the software industry is further diverging. Software more resilient to AI shocks in the future often possesses moats such as regulatory compliance, data accessibility, integration complexity, and clear accountability boundaries; the more standardized and easily automated the workflow, the greater the pressure for replacement.
