Productivity-Driven Growth Masks Complex Economic and Market Dynamics

Key Quotes

"We have flat unemployment at 4 and a half percent but it's not going down as you might think when you're printing let's say 2 6 for real gdp in 2026 and a small part of the answer is that that 2 6 probably overstates the underlying trend a bit because we had the shutdown that depressed q4 it's going to add to q1 but the more important part of the answer is accelerating productivity growth and we've seen that over the last five years the five years since the pandemic have shown about 2 underlying trend productivity growth the prior cycle was at about 1 and a half percent and i think there's reason to believe that that acceleration is still ongoing because it probably doesn't have a lot of ai in it we expect more of a boost from ai going forward in the next five years than in the last five years and i think that's got important implications for the relationship between gdp and unemployment"

Jan Hatzius explains that while GDP growth is projected, unemployment may not decrease proportionally due to factors like the previous quarter's shutdown and, more significantly, accelerating productivity growth. Hatzius suggests this productivity acceleration, potentially boosted by AI in the future, could alter the traditional relationship between economic output and job creation.

"For three years we've been obsessed with ai as a market and for three years really the story has been increased capex extraordinarily growth in ai investment spending and although we're discussing a lot the eventual productivity benefits as jan mentioned really the trade in the equity market has been the companies with earnings that are benefiting from those capex dollars and what's happened this year especially in the last several months is it's become increasingly clear that a probably that growth will decelerate next year and b to continue that growth it's going to require more debt and both of those factors have made investors understandably uncomfortable"

Ben Snider highlights that the market's focus on AI has been driven by increased capital expenditure (capex) and investment spending. Snider notes that recent investor discomfort stems from the realization that this growth may decelerate and require more debt, shifting the market's attention from capex beneficiaries to the underlying productivity benefits.

"The impact of tariffs this is our concern too most of earnings variability is driven by margin variability when you have years with very large earnings growth or very poor earnings growth it's usually because of margins and earlier this year we were concerned that in part because of tariffs margins would get squeezed and that would weigh on earnings was it just you who was concerned and it and it really didn't materialize now i think part of what we have to keep in mind is the counterfactual so for the last couple of quarters s p 500 profit margins have basically been flat well in an environment of pretty healthy nominal gdp growth normally you would expect some operating leverage that would cause margins to expand and so i think part of the story here is you didn't get that counterfactual but we've seen consistently over the last few quarters companies across earnings calls really point to three levers they've been pulling to offset these pressures one is of course as john mentioned pushing through some in the form of prices second is both pushing back on suppliers to absorb some of the costs and restructuring supply chains where necessary and the third is cutting costs improving efficiency within companies and that ties back to the slightly better productivity story we've seen"

Ben Snider discusses how tariffs were a concern for compressing corporate profit margins, which are a key driver of earnings variability. Snider explains that despite expectations, margins have remained flat, and companies have offset pressures by raising prices, negotiating with suppliers, and improving internal efficiency, linking this resilience to productivity gains.

"The general consensus certainly that we have and that most of our clients have is that ai eventually will create a very large productivity boost to the economy that will create value for someone who that someone is is hard to answer and as i mentioned earlier what the market has been doing given that uncertainty over the last couple of years is really focusing on near term earnings it's been the semiconductors obviously it's been the hyper scalers to some extent power companies etc i think that is actually one of the key differences between this market and what happened 25 years ago during the dot com bubble where we saw valuations expand quite dramatically as investors tried to look forward and guess at the productivity gains and the economic benefits today investors are saying we saw what happened that time it's too hard and so what we're really going to focus on is the earnings today"

Ben Snider describes the market's current approach to AI, noting that while a significant productivity boost is expected, identifying who will benefit remains unclear. Snider contrasts this with the dot-com bubble, explaining that today's investors are prioritizing near-term earnings from sectors like semiconductors and hyper-scalers over speculative bets on future productivity gains, a more cautious strategy.

"I think the sentiment surveys have been getting less and less useful for predicting activity and that's true for the consumer sentiment surveys and the university of michigan in particular has been quite far out of line with what we've seen but it's actually true more broadly if you think back to 10 years ago 20 years ago just the importance of the ism print for markets and the importance now it's just nowhere close we still look at the business service and the consumer service because they have interesting detail they're very up to date but they just don't work as well as they used to or certainly were believed to so i would really focus more on the hard data the hard data would say that the consumer is doing okay consumer spending is certainly not super rapid maybe we'll get to you know 2 or so next year but i think consumer spending and real terms is likely to lag the overall economy"

Jan Hatzius points out that traditional sentiment surveys, like the University of Michigan's consumer sentiment index, are becoming less reliable indicators of economic activity. Hatzius advises focusing on "hard data" such as consumer spending, which, while not rapid, indicates the consumer is "doing okay" and is expected to lag overall economic growth.

"I view valuations through a physics lens as a measure of potential energy they tell you a lot about how much the market can move if there's a catalyst but they're not the catalyst themselves"

Ben Snider uses a physics analogy to describe market valuations, comparing them to potential energy. Snider explains that high valuations indicate the market's capacity for movement when a catalyst appears but do not themselves cause market shifts.