Goldman says the stock market has already priced in the AI boom, with $19 trillion of market value running ahead of actual economic impact so far

The US stock market may have already priced in much of the potential long-term value generated by artificial intelligence, according to a new analysis by the investment bank. Analysts Dominic Wilson and Vicky Chang point out that some “simple math” suggests that the market pricing of AI gains is running “well ahead of the overall impact,” with rising valuations in AI-related companies approaching the upper bounds of plausible economy-wide benefits.

Although Goldman’s portfolio strategy team insists the company’s valuations are high but have not yet reached “bubble levels”, the macro approach helps put limits on “what is collectively possible”.

What is a few trillion dollars, anyway?

The report estimates that the present discounted value (PDV) of capital revenues generated by generative AI to the U.S. economy amounts to a base estimate of $8 trillion. Although these calculations are inherently uncertain, a reasonable range for these future capital revenues is between $5 trillion and $19 trillion. It is worth noting that these expected benefits are sufficient to justify current and projected levels of investment spending on AI-related capital expenditure, a major concern in the financial media recently. On the other hand, market enthusiasm appears to have overtaken basic macro calculations.

Since introducing ChatGPT in November 2022, Goldman calculates that the value of companies directly involved in or adjacent to the AI ​​boom has risen by more than $19 trillion. This boom includes big gains in semiconductors and among “hyperscalers,” as well as nearly $1 trillion for the latest valuations of the three largest private companies providing AI models.

This overall valuation increase places market gains at the “upper bound of expected aggregate benefits” ($19 trillion) and far exceeds the baseline estimate of $8 trillion. Specifically, the change in the value of AI-related companies in semiconductor and private AI model providers – which is more plausibly attributable to the AI ​​boom – already exceeds the baseline estimate of $8 trillion for capital appreciation.

Goldman Sachs notes forward-looking markets He should Price gains are ahead of time, describing it as a “feature, not a bug”, but analysts have identified two key risks that may reinforce the tendency to “overpay” future earnings, citing two ominous precedents: “Previous booms that were driven by innovation – such as the 1920s and 1990s – led the market to overpay future earnings even though the underlying innovations were real.” (Goldman did not comment directly on the crashes of 1929 or 2000, which accompanied the famous booms of US history.)

The two main risks highlighted are:

1. Aggregation fallacy: Investors may imply excessive total revenue and earnings gains by extrapolating the impressive earnings growth that can be achieved by individual companies across all potential winners. This risks the shared value attributed to chip designers, model builders, and hyperscalers beyond what they can ultimately capture together.

2. Fallacy of induction: Competition often erodes the initial profitability gains from innovation over time. Markets may overestimate the long-term path of earnings growth if they treat short-term temporary increases in earnings as persistent.

The fundamental productivity promise of AI remains strong: It is estimated that AI could boost US productivity by about 1.5 percentage points for a decade, ultimately raising US GDP and profits by about 15%. As long as the broader economy and the AI ​​investment boom remain “on track,” markets are likely to maintain an optimistic view. But outside of hardware, AI’s current gains remain limited, which could pose risks if predictions don’t materialize quickly.