When Amazon was hemorrhaging money in its early years, Wall Street kept buying. On paper, it didn’t add up. But, according to Lukas Kremens, an associate professor of finance and business economics at the University of Washington Foster School of Business, the market was doing exactly what markets do: looking ahead.
“Markets are forward-looking in a way that accounting statements are backward-looking — or, very generously, present-looking,” Kremens says. “Market power is something that a firm has without it necessarily being obvious from what it does.”
That gap between what a company earns today and what the market expects it to earn tomorrow lies at the heart of Kremens’s new research. In a paper published in the Review of Financial Studies, Kremens and his co-authors developed a way to read stock prices for signs of a company’s future market power — its ability to raise prices and earn higher profits down the road — often years before that power turns up in the company’s earnings.
Kremens likens those prices to a kind of prediction market: the aggregated bets of thousands of investors putting real money behind their read on a company’s future. When that many people with money at stake agree on a price, that price tends to mean something.
“Markets are forward-looking in a way that accounting statements are backward-looking — or, very generously, present-looking.”
Lukas Kremens finds that investors often recognize a company’s ability to raise prices and earn higher profits years before that power appears in financial statements.
Why high stock prices are about more than low interest rates
The paper begins with a deceptively simple question: Why have U.S. stock prices risen so dramatically since 1980?
The conventional answer points to interest rates. When rates fall, stock prices tend to rise, and rates did fall sharply over that period. But the data complicated that answer.
“The most surprising thing to me was how little of the rise in asset prices really goes to the obvious finance explanation, that interest rates have gone down a lot, so obviously asset prices are up,” Kremens says. “There’s a lot more to it.”
Interest rates still mattered, but cheaper money and the returns investors expected explained only about 29% of the increase in the value of U.S. companies between 1980 and 2020. The rest came from what investors expected those companies to do in the years ahead. Part of it was simple growth: Companies expected to sell more over time. But the part Kremens’s paper brought into focus was the expectation that companies would earn higher profit margins.
Those expected gains came with costs: R&D, technology, talent, and other investments companies needed to build and protect that profitability. But investors were betting the profits would more than cover it. The market’s four-decade run, in other words, was driven less by cheap money than by a broad bet on future corporate profits, including rising pricing power.
The difference between building market power and using it
One of the approach’s most useful features is its ability to detect market power that hasn’t yet appeared in a company’s financials. Current profit margins and future expected profit margins tell very different stories, especially over the long run.
Kremens illustrates the distinction with a pair of familiar examples. For years, Amazon and Uber operated at a loss or near break-even, deliberately forgoing profits to grow their customer bases and entrench their platforms. Nothing in their income statements suggested exceptional pricing power. But investors priced them as if that power was coming.
“Amazon probably had market power long before it chose to harvest it,” Kremens says.
Both companies were valuable, he notes, because investors were paying for profits they expected to arrive later.
What investors were really seeing, the research suggests, was the value of intangible investment. Amazon wasn’t building factories; it was building logistics software and Prime membership lock-in. Uber was building a network. Neither showed up as traditional capital expenditure, but both were laying the groundwork for the pricing power that would come later.
The retail sector offers a particularly striking validation of this idea. A 2018 study by economists Nicolas Crouzet and Janice Eberly examined rising market concentration in retail and noted something puzzling: Despite the growing dominance of a handful of players and heavy investment in technology and logistics, profit margins in the sector had barely moved. Kremens’s framework told a different story. When he and his co-authors applied their forward-looking measure to retail, they found that expected profit margins had been trending upward since the early 1980s, tracking the sector’s rising valuations even as realized margins remained flat.
“The forward-looking measure had been telling you that part of what’s going on in the retail sector is about market power,” Kremens says. “Firms were building market power with all their intangible investment; they just weren’t harvesting it yet. And then they started.”
The accounting data eventually confirmed it. Retail profit margins, flat for decades, began climbing after 2015, confirming what the market had been signaling all along.
Mergers tell a similar story: when two companies combine, expected future margins rise almost immediately, even when current margins show no meaningful change for at least five years. The market sees something in the deal that the income statement won’t reflect for years.
“Amazon probably had market power long before it chose to harvest it.”
In research published in the Review of Financial Studies, Lukas Kremens shows how stock prices can signal future profitability long before companies begin reporting strong earnings.
What this means for investors, regulators, and the classroom
That gap between what markets see and what financials show has real consequences, nowhere more so than in antitrust regulation. Regulators typically evaluate mergers based on current market share and existing margins, but Kremens’s research suggests that’s only part of the picture. When Facebook acquired WhatsApp, the platform was making almost no money, giving regulators little to flag. But the price Facebook paid encoded a clear expectation of future monetization and, arguably, future market power.
“If you’re betting on market power,” Kremens notes, “there are also regulators that might want to take a look at this.”
In his course on alternative investments, Kremens puts the same question to his students. But it applies well beyond the classroom, to investors, regulators, and business leaders alike.
“The key question I basically always end up asking them: What makes you think that the thing you’re predicting, the thing that makes this such a great investment opportunity, is not already in the price?”
Read the full paper: “The Present Value of Future Market Power” — Thummim Cho, Marco Grotteria, Lukas Kremens, and Howard Kung. Published in the Review of Financial Studies (2026).
Lukas Kremens is an associate professor of finance and business economics at the Foster School of Business. Learn more about Foster’s new MS in Finance program, which launches in 2027.