BCA Research’s chief global strategist Peter Berezin has lowered his recommendation on equities to a "slight underweight" following a fresh warning from the firm's MacroQuant framework.
The decision came after MacroQuant’s equity z-score fell to -1.01 on Friday, moving below the -1 threshold that, Berezin says, "has historically predicted the start of equity bear markets." MacroQuant combines a range of measures - spanning the economy, monetary conditions, sentiment and positioning, and valuations - to compare current readings with past patterns.
"Simply put, the model is saying that today’s configuration of macro and financial data has been typically associated with negative future equity returns," Berezin wrote, summarizing what the composite signal implies for forward returns.
Within the model, the sentiment component stands out as the most alarming input. Berezin notes it is signaling "extreme investor complacency," a stance he finds particularly worrying given the uncertain macro backdrop.
Inflation-related measures inside MacroQuant have also weighed on the equity score. Berezin acknowledged, however, that a resolution to the Iran conflict could ease price pressures and would likely make the model "marginally more constructive" in subsequent months.
At the same time, Berezin cautioned that MacroQuant has limitations in the current policy environment. He argued that "the model is not well equipped to see the rapid shifts in policy in the Age of Trump," and added that if AI amounts to a genuine structural break, "the historical correlations underpinning MacroQuant may cease to be reliable guides."
The strategist also offered a specific take on AI dynamics. He said he is "bullish on AI as a technology but much more skeptical on AI as a business," suggesting that companies providing models could evolve into low-margin, capital-intensive utilities. In his analogy, such providers might end up resembling airlines - "critical infrastructure but commoditized, capital-intensive and low-margin."
One of MacroQuant’s sub-indicators has already shifted to an amber status. Token pricing has fallen in recent weeks as cost-conscious users appear to step back from the priciest frontier models. Berezin warned that if this trend persists, it could reignite concerns about the monetization of AI-related capital spending.
Putting these pieces together, Berezin said he is downgrading stocks to a slight underweight across both a three-month and a 12-month horizon. He set clear conditional actions: he would revert to a neutral stance should MacroQuant’s outlook improve, but would move to a full underweight if the signal worsens or if additional AI indicators in BCA’s framework begin to deteriorate.
Market context notes appearing alongside commentary referenced major equity gauges, including NDX, US500 and the Nasdaq 100, as background markers of market activity.
Investors and market participants will likely watch MacroQuant’s next prints closely to see whether sentiment readings and token pricing stabilize, and whether inflation and geopolitical developments ease pressures on the model’s equity score.