Goldman likes defensives.

Wait, let me back up and be more specific. A model Goldman built to help assess the odds that a given equal-weighted sector will outperform likes defensives. Or as much as a model can “like” something. If you interrogate ChatGPT, it’ll concede that computer models can’t actually “like” anything, nor, by the way, can they be “sorry” (let alone “genuinely sorry,” as OpenAI’s image generator, DALL-E, occasionally pleads) when they screw up.

I digress. Goldman didn’t use any sort of advanced AI for their sector work, just a good old fashioned probit model. As the bank’s David Kostin noted, summarizing the output, the model “currently recommends overweights in Materials, Software & Services, Health Care, Utilities and Real Estate.”

Materials is in some respects the odd man out there. Thanks in part to lackluster RoW (i.e., ex-US) growth, the sector’s underperformed its relationship to commodities and, as a result, missed out on some of the cyclical rally. Of course, it carries significant tariff risk, but some of that may be offset by attractive valuations. In other words, “some pessimism is already priced in,” to quote Kostin.

The figure below gives you a hint as to why Goldman’s model currently favors some defensives.

Suffice to say cyclicals have overshot the growth impulse, which is really saying something considering how well the US economy’s performed.

“Our economists’ forecast for above-consensus US economic growth in 2025 coupled with potential fiscal policy changes generally support a cyclical posture [but] Cyclicals excluding commodities have outperformed Defensives by 5ppt since Election Day and is now consistent with real GDP growth well above even our economists’ optimistic growth forecasts,” Kostin went on.

The figure shows you the model’s “recommendations,” and do note the scare quotes. Not investment advice. For illustrative/entertainment purposes only. And all the usual caveats.

Why Software & Services? Doesn’t that space have at least some cyclical characteristics? Yes, is the answer to that latter question, but Goldman expects the space to “be a leading beneficiary of an AI ‘hand-off’ from infrastructure to AI-enabled revenues.” As for Utilities, they’ll benefit from AI-related power demand, Goldman reckons, and Health Care’s cheap (the stocks, not the care itself) at just 16x, even as Goldman was quick to add that the bank’s model isn’t capable of capturing “policy uncertainty.”

The overarching point is that the pro-cyclical nature of the rally may mean the Cyclical theme’s overcooked in the near-term, or even over the medium-term. “Current market pricing means risk/reward appears more attractive in some defensive sectors,” Kostin wrote.

For what they’re worth, the figures below show you the absolute and relative positioning of professional investors as polled by BofA for December’s installment of the bank’s popular monthly survey.

Investors are long financials, US equities and Utilities, relative to history, and underweight cash, Eurozone stocks and Energy.

Recall that the same survey suggested investors are the most Overweight US shares in at least a quarter century, consistent with expectations for continued outperformance in 2025, when US shares are the odds-on favorite to be the best asset (followed by global equities and, “naturally,” Bitcoin.)

As for Goldman’s regression, Kostin relied on the forecasts of Goldman’s economics team for the macro inputs the model required. There’s a “garbage in, garbage out” joke there, but I won’t tell it. (Oops, I suppose I just did.)

Oh, and if you’re wondering what you might’ve tried to best the S&P’s blockbuster 2024 performance, you’d have scored 7ppt of alpha with the ticker “YALL,” the “God Bless America ETF” which markets itself as an investment product for “God-fearing, flag-waving conservatives.” As Bloomberg dryly noted, the fund’s “biggest holding is Tesla.”