AI Boosts Market Efficiency but Heightens 'Tail Risk,' Bernstein Says

Artificial intelligence is making financial markets more efficient in many respects but is simultaneously introducing new vulnerabilities that could amplify volatility and sharp market downturns, according to a new research note from Bernstein.

The findings suggest a dual impact where AI reduces average, day-to-day market inefficiencies while increasing the probability of more severe, infrequent market shocks, a phenomenon known as tail risk.

The Efficiency Gains

At the company level, analysts at Bernstein noted that AI is improving market efficiency by rapidly processing vast amounts of data. This includes earnings reports, regulatory filings, and alternative datasets, allowing for faster analysis than humanly possible. This acceleration helps narrow information gaps, improves price discovery, and reduces the magnitude of earnings surprises as forecasts converge more quickly with actual results.

AI is also transforming the research process itself. Key benefits highlighted in the note include:

• Automated workflows that cut the time needed for earnings reviews and financial model updates.
• Expanded analyst coverage, particularly for previously under-researched stocks like small-cap companies in emerging markets.
• A potential reduction in the long-standing valuation premium often associated with stocks that receive less analyst attention.

Ad

New Vulnerabilities and Systemic Risks

Despite the benefits, the widespread adoption of similar AI models presents new dangers. Bernstein warns that as more investors rely on comparable tools trained on overlapping data, trading signals could become increasingly synchronized. This can lead to crowded positions and make market reversals more severe during periods of stress.

The note pointed to recent market events as potential examples of these risks, including the unwind of the yen carry trade in August 2024 and instances where AI-generated misinformation briefly moved U.S. equities. These events illustrate how automated strategies can accelerate volatility before markets have time to verify new information.

Another concern is the "reflexivity problem," a feedback loop where AI-generated recommendations influence investor behavior, pushing prices in a direction that then reinforces the model's future outputs. This could strengthen momentum trades, increase market concentration in a few assets, and produce larger valuation extremes than would otherwise occur.