The prevailing inflation and fiscal narrative currently dominating the US Treasury market is facing a potential takeover by a new, powerful theme: Artificial Intelligence. Torsten Slok, Chief Economist at Apollo Global Management, has warned that the risk of a significant decline in long-term interest rates over the next six months is rising, a scenario he believes is difficult to avoid whether AI ultimately succeeds or fails.

In his latest report from August 31, Torsten Slok argues that AI will become the central narrative driving interest rates in 2027. If AI commercialization is successful, tech companies creating trillions of dollars in revenue would trigger a massive deflationary effect, pushing long-term rates down. Conversely, if the AI bubble bursts, the Nasdaq could plummet by 50%, prompting investors to flee equities en masse into the safety of US Treasuries, which would also drive long-term yields sharply lower.

The core logic behind this projection is that financial markets are driven by narratives. Currently, the interest rate narrative is centered on high inflation and fiscal deficits. However, as we approach 2027, Torsten Slok believes this storyline will be replaced by the success or failure of AI—and both potential endings point in the same direction for rates: downwards.

The AI Success Scenario: Deflation Lowers Rates

Torsten Slok points out that the current high level of long-term rates is fundamentally due to two structural pressures: persistent inflation and the deteriorating US fiscal position. These two forces together support market expectations of prolonged high interest rates and dominate current bond market pricing logic.

However, Torsten Slok argues that this framework is not permanent. He explicitly states that these forces could be overtaken by the trajectory of AI development in early 2027, at which point market attention will undergo a fundamental shift.

In the scenario where AI succeeds, Torsten Slok paints a clear transmission path: if tech companies generate trillions of dollars in revenue through AI technology, the technology will have a large-scale deflationary impact on the overall economy. Deflation means a systematic decline in price pressures, which would directly weaken long-term inflation expectations and subsequently push long-term interest rates lower. In this scenario, the productivity dividends of AI translate into downward price pressure, fundamentally altering the basis for rate pricing in the bond market.

The AI Failure Scenario: Bubble Burst Triggers a Flight to Safety

If AI development fails to meet expectations, Torsten Slok predicts the market will face the shock of a bubble bursting. He specifically notes that the Nasdaq could fall 50% in this scenario, with investors taking large-scale withdrawals from the stock market and rotating into US Treasuries for safety. This massive capital shift would directly push bond prices up and yields down, leading to a "sharp decline" in long-term rates.

This logic aligns closely with the historical pattern of Treasuries acting as a safe haven during risk asset selloffs—a stock market crash actually becomes a powerful catalyst for the bond market.

A Six-Month Window: The Market Prepares to Make a Judgment

Torsten Slok emphasizes that the next six months are a critical observation window. During this period, the market will gradually form a judgment on the AI development path—whether it successfully delivers commercial value or whether the bubble begins to loosen.

He notes that financial markets are, by nature, narrative-driven. The current rate narrative focuses on inflation and fiscal issues, but entering 2027, the dominant narrative will shift to the success or failure of AI. Whichever narrative ultimately prevails, the direction of its impact on long-term rates is consistent: downward. This implies that for bond investors, the current high-rate environment may represent the eve of a narrative shift, and the logic for positioning in long-end Treasuries is quietly building.