M2, GDP, and the stock market

One chart explains nearly everything about modern finance, and almost no one will ever show it to you. Plot three things from 1980 to today: the US M2 money supply, the S&P 500 total return index, and US nominal GDP. Normalize all three to 100 in 1980 and watch what happens. M2 grows roughly 13×. Nominal GDP grows about 10×. The S&P 500 grows 53×. Real GDP — the actual quantity of stuff America produces, with inflation stripped out — grows only about 3.3×. The economy did not get 53× better. The number on the index moved 53× because the unit it is denominated in shifted underneath it.

This is the thesis: stock market returns over the last four decades are more tightly correlated with the growth of the money supply than with the growth of corporate earnings or the size of the underlying economy. Earnings matter. Productivity matters. But they are not the dominant driver. The dominant driver is liquidity — the quantity of dollars in the system and the cost of borrowing against them. When the marginal dollar expands, the price of every claim on the future expands with it.

The mechanism is not mysterious. A stock is a claim on a future stream of cash flows, discounted back to today at some rate. Lower the discount rate and the present value rises, even if the cash flows themselves do nothing. Expand the money supply and you do two things simultaneously: you push more dollars toward a fixed pile of productive assets, and you suppress the rate at which those assets' future cash flows are discounted. The same productive economy is now being priced in more dollars, and each of those dollars is being charged less to wait. Both effects push asset prices up. Neither requires earnings to grow. The instrument below makes this concrete.

Set the slider on the chart to 2008 and then slowly drag forward. Watch what M2 does. The Federal Reserve had a roughly $900 billion balance sheet on the eve of the financial crisis. By 2014 it was $4.5 trillion. By 2022 it was nearly $9 trillion. Every major drawdown in the S&P 500 since 2008 — late 2008, August 2011, December 2018, March 2020, December 2022 — has been met with monetary easing in some form: rate cuts, balance-sheet expansion, emergency lending facilities, or simply a public communication that easing is coming. This is what market participants call the "Fed put." It is not a written policy, but it is an empirical regularity precise enough to trade. The central bank does not have an asset-price target on paper. It has one in practice.

The consequences of this are not evenly distributed. The most-quoted statistic for political theater is that "half of Americans own stocks." It is technically true and substantively misleading. Only about 14% of US households own more than $10,000 in equities — directly or in retirement accounts combined. The top 10% of households owns roughly 89% of all stocks. The top 1% owns more stock wealth than the bottom 90% put together. When the central bank expands the money supply to support asset prices, the wealth it creates lands almost entirely on existing asset holders. The mechanism is not redistributive; it is concentrating by construction. A renter watching grocery prices climb does not benefit when the S&P recovers. The same easing that lifted the index lifted the rent.

Look at the Buffett indicator — the ratio of total US stock market capitalization to nominal GDP. Warren Buffett famously called it "probably the best single measure of where valuations stand at any given moment." Historically it ran between 60% and 80%. It crossed 100% in the late 1990s, returned to roughly 130% before 2008, and has been above 200% — twice the long-run norm — for most of the post-2020 period. There is no plausible story in which the cash-flow base of American business has doubled relative to GDP. Companies have not become structurally twice as profitable as a share of the economy. What has happened is that the price applied to those companies' cash flows has roughly doubled. That is not earnings. That is repricing. And repricing of that magnitude does not happen without a corresponding expansion of the monetary base.

The corollary is uncomfortable: when M2 contracts, risk assets correct hard. In 2022 the US M2 money supply contracted year-over-year for the first time since the 1930s — a roughly 4% drawdown. Stocks fell about 25% peak to trough. Long-duration bonds had their worst year in modern history. Cryptocurrency, the longest-duration asset of all, fell 70%+. These markets did not all crack on the same earnings news, because there was no earnings news that explained all of them simultaneously. What they shared was sensitivity to the marginal dollar. When the dollar got scarce, every asset whose value comes from discounted future promises got re-discounted at a higher rate. The episode is the cleanest natural experiment in the modern liquidity-asset link, and most coverage of it never connected the two.

The implications for an ordinary investor are unusually direct. First, if asset prices are heavily liquidity-driven, then asset allocation is partly a bet on monetary policy — knowing whether the central bank is expanding or contracting matters more than projecting next quarter's earnings. Second, retirement planning that assumes 7-10% real returns indefinitely is implicitly assuming this regime continues; if M2 ever stops expanding at multiples of real GDP, the regime ends. Third, the wealth gap that everyone laments is not separable from the monetary system that produced it: the same machine that supports your 401(k) on the downside also concentrates the upside in the hands of the people who already own everything. You cannot fix the inequality without confronting the mechanism. And the mechanism — central-bank support of asset prices — is the load-bearing wall of the modern political economy. Removing it would crash the housing market, the pension system, and the federal government's ability to service its own debt. So it does not get removed. It only gets expanded.