Never Go Full Macro: Five Lessons from Aahan Menon
The Prometheus founder on the market that isn’t the economy, the only kind of inflation that sticks, and the rare privilege of being paid to diversify.
Jack introduced Aahan Menon as the most feared person in systematic macro, on the strength of the boxing videos circulating on Twitter, and Aahan promised he gets “it all out in the ring so we don’t have to take it anywhere else.” Then he shared a deck of more than 90 slides, of which we managed to cover about a dozen, with a six-hour follow-up episode duly threatened. Aahan runs Prometheus, where he builds daily, systematic reads on the economy, nowcasts of growth and inflation, business cycle monitors, and regime probabilities stretching back to 1965. One of the deck’s opening titles was “Not a Good Time to Make Macro Bets,” and it became the conversation’s through-line.
Lesson 1: The Market Is Not the Economy
We asked Aahan about the popular claim that AI capex is basically the whole economy now, and he called it “a very gross overstatement.” Tech investment is contributing more to GDP than it ever has, he said, and that contribution is somewhere between ten and thirty basis points against headline growth near three percent. It feels bigger because it’s the only part of the investment complex that’s moving, with residential and industrial investment both lackluster. But US GDP is mostly US consumption, and end consumption of AI barely registers next to healthcare, retail services, and the same expenses that have powered the economy for two decades.
In markets, AI CapEx is driving almost everything. But extrapolating that to saying that AI is just powering the entire economy is an overstatement that really deviates from the actual reality that at the end of the day, the economy is mostly just consumption.
Prometheus runs several versions of its nowcast, one that reconstructs the official GDP numbers monthly, plus weekly and daily gauges, with the daily one built on a proprietary income-based approach. “We’re not trying to get exactly on the money what the next official GDP number print is gonna be,” he said. The point of a gauge that updates every day is “to get ahead of the big muscle movement.” What it’s said all year is that incomes, which feed spending, which is what GDP is, have been really stable. So nominal growth chugged along while asset prices swung violently around it. Confusing the two leads investors to project the market’s drama onto an economy that mostly isn’t participating in it.
Lesson 2: Only One Kind of Inflation Sticks
While his growth gauge has been boring all year, Aahan’s inflation readings have been absolutely all over the place, swinging from some of the hottest prints in decades to outright deflationary ones and back. Oil is the culprit, and it is generating enough volatility to drown out everything else. Every sell side and buy side shop runs its own version of the CPI calculation, so each big oil move feeds a reflexive loop where estimates change, Fed expectations change, and market pricing chases all of it. His response was to take the number apart. Borrowing an idea from Fed research and rebuilding it with actual nominal demand data, Prometheus splits inflation product by product into demand-driven and supply-driven components.
The persistent inflation comes from a services-driven, demand-driven inflation. The really volatile kind of back-and-forth inflation tends to come from a supply shock type of inflation.
The demand kind, he said, has much stronger autocorrelations. “Rising inflation leads to more rising inflation. Falling inflation leads to more falling inflation.” Supply shocks behave differently. “You have a sudden shortage. People rush in to start producing because prices are high, and then that neutralizes itself over time.” By his math, roughly 75 percent of the current reading is demand-driven, so even stripping out every supply force, oil included, “you’re still pretty well removed from target.”
Lesson 3: A Risk Can Be Real and Untradeable
We brought up Andy Constan’s concern from a recent episode, that Americans are dissaving, and Aahan, a friend of Andy’s, agreed with the worry while drawing a sharp line about what to do with it. Every unit of dissaving flows somewhere, and his extension of the Kalecki-Levy profits equation shows where. Consumers are simultaneously corporate America’s biggest expense and biggest revenue source, and since COVID, the household has become the dominant driver of corporate profits, with falling savings rates mechanically pushing profits higher.
When you go out and say that you’re worried about the savings rate being too low, and you start trying to make timing decisions based on that, you’re basically trying to fade the most dominant driver of corporate profits. And that’s a really hard thing to do.
“I would be on the lookout for when the impulse from this saving begins to turn,” he said, “rather than to try to take action beforehand.” As for what’s behind the dissaving, his favorite theory “is just that you have a really, really large wealth effect in the United States,” people spending more of their income because asset prices hold up their net worth, which means the economy depends on the stock market more than it ever has. When Jack wondered whether that changes Fed behavior, Aahan pointed out that every recession that led to easing came with falling equity prices, so “that sensitivity has always existed.” It fit a theme the two kept circling, that we treat old phenomena as unprecedented, right down to presidents pressuring the Fed, which history has seen plenty of times before.
Lesson 4: Indicators Expire When the Economy Changes Shape
Classic leading indicators, the yield curve, housing permits, PMIs, have been flashing recession since 2022 and been wrong the whole time. “I’m sure I’m not supposed to do this as a research provider,” Aahan said.
In 2022, Prometheus was using the classic models too.
We were like, “Oh, all the indications are pointing to recession,” and markets began to look nothing like it.
The miss pushed them to ask, in his words, “how do we update this framework to take into account how the economy is changing?” The old indicators were designed for an economy dominated by manufacturing, with its big, lagging inventory cycles, and for a housing sector at its former scale. Manufacturing shrank for decades as production moved offshore, housing’s base never recovered its pre-2008 weight, and services, which now dominate, are simply less cyclical. So Prometheus added information processing and equipment investment to its business cycle monitor, and the expanded aggregate has correctly said expansion through three years of recession calls. An indicator’s track record belongs to the economy that produced it. The same skepticism governs his approach to shiny new datasets. He wants refinements of series he already understands, like faster reads on lagged consumption data, because a brand-new source with three years of history contains no cycles to evaluate, and “the higher the edge in the thing that you find, the more likely that it is that it’s gonna get eroded,” the way the satellite photos of retail parking lots were.
“Whenever I find myself wanting to really go out there and, as they say, go full macro,” Aahan started, and Jack jumped in, “They say never go full macro, right?” They do. “When I find myself having that urge, I always look to the case of Australia, where, barring COVID, I think you haven’t had a recession in over 40 years.” Some economies are just much less cyclical.
Lesson 5: Sometimes the Market Pays You to Diversify
Aahan’s regime framework descends from the Dalio four-quadrant idea, growth and inflation each rising or falling, made investable by using asset classes themselves as the anchors, equities for rising growth, treasuries for falling everything, commodities for rising nominal growth, TIPS for stagflation. Every day since 1965 gets scored for similarity to today, and expected returns get built by overweighting history’s look-alike days. Concentrated macro bets pay when those probabilities are decisive, since “most macro portfolios benefit from persistent trends in a direction.”
About 80% of all bond returns basically come during easing cycles. The rest of the time, bonds are just fine.
Timing those cycles well, he argued, comes from hard macro data rather than parsing speeches, because “it is very, very hard for the Fed to deviate from the path of nominal GDP,” and strategies that treat deviations from the hard data as opportunities have tended to work. Right now, though, the distribution is nearly flat, with no regime clearly favored. That is precisely why the deck says it’s not a good time for big macro bets. Trends keep starting and getting disrupted by the next headline before they can compound. Jack noted that Twitter seems awfully negative on stocks and this framework isn’t. Aahan walked through the reasons: monetary policy pressures neutral, nominal GDP stable, the business cycle expanding, and “all of those things are fairly constructive on equities.” Stretched valuations don’t change that, because “valuations are a good way to allocate over time, but they aren’t a good way to time the markets in the near term.” The big sell-offs come when rich prices finally meet deteriorating fundamentals, and his gauges will say so when they do. Until then, a flat distribution means the portfolio that owns a bit of everything is, for once, the well-priced one.
The Bottom Line: The Hardest Thing Always Ends
Jack said humility seems really important right now, and that if you follow Twitter you’re seeing maybe the opposite of it. Aahan agreed. Even investors who nailed this year’s winners can look back on an amazing total P&L, but “the amount of volatility you had to endure to get there is something generational.” When Jack brought up semiconductors, no longer cyclical in January and a completely different tune by the pullback, Aahan called it “just a classic investment cycle.” “There’s always gonna be the hardest thing, and that hardest thing is gonna look like it solves all investment problems to come. But invariably, that always ends.” Systematic investing, as he practices it, is asking what your expected return is across the whole distribution of outcomes rather than the path you hope for. Near the end, Jack noted they’d come full circle to the deck’s title, with oil driving so many things and oil so unpredictable.
It’s one of the few times in history where you can actually get paid to diversify.
After years of hearing that diversification is dead, we’ll take it.
Watch the full episode here: