Taxes and the Market--A Messy Relationship

These are some strange times for capitalists, with prominent political voices advocating for something we've never seen in this country--a tax on unrealized gains. Senior congressional Democrats like Senator Elizabeth Warren and Senator Ron Wyden argue that the current "realization" system allows asset owners to amass fortunes without ever paying income tax. They allege by borrowing against their appreciated assets rather than selling them, billionaires can fund their lifestyles while their wealth remains untaxed—a "loophole" they'd like to close with a "Billionaire Minimum Income Tax." The nebulous proposal targets households with a net worth exceeding $100 million, requiring them to pay an annual minimum tax of 25% on a combination of their ordinary income and the year's "paper gains."

I'm not a tax expert, but I think there are grievous problems with this proposal, both in theory and in practice. It's unlikely to get additional traction, but the fact that it is being discussed is worrisome.

However, this craziness did inspire me to take a deeper look at tax policy impacts on market returns--which is more in my wheelhouse.

More than a century of U.S. tax policy offers a natural experiment, which economists like: what actually happens to equities when Washington changes the rules? I mapped every major shift in individual capital gains rates, top ordinary income rates, and corporate income rates from 1913 through 2025 against the S&P 500's annualized performance under each regime. The dataset covers 16 completed tax regimes and one still in progress.

The story is not that lower taxes produce stronger markets and higher taxes kill them. The relationship is messier than that, but maybe more interesting.

The headline: rate direction is a weak predictor

Across the full sample, regimes that began with a capital gains rate hike averaged 7.6% annualized returns. Regimes that began with a cut averaged 12.8%. That's directionally consistent with the standard narrative, but two things muddy the picture.

First, sample sizes are tiny: seven hikes, five cuts, three periods where the rate held flat. No subset is large enough for statistical confidence. Second, the averages mask enormous variation within each group. The 1993 Clinton-era hike, which raised the top ordinary rate to 39.6%, was followed by 19.2% annualized returns, which is the single best regime in the dataset. The 1942 wartime regime imposed an 88% top ordinary rate and 25% capital gains rate, yet produced 16.8% annualized. Meanwhile, the 1997 capital gains cut from 28% to 20% delivered only 5.4% annualized as the dot-com bust dragged returns down.

The correlation between the magnitude of the cap gains rate change and the subsequent regime return is −0.41. That's the strongest signal in the data, and it points in the expected direction — hikes modestly associate with weaker returns. But at n=15, it falls well short of statistical significance. It's a tendency, not a predictor.

The corporate rate surprise (and its limits)

One observation that jumps off the page is the TCJA's 2018 corporate rate cut from 35% to 21%, which produced a 13.4% annualized regime return with a 14.2% trailing run-up as markets priced in the legislation. That seems to support the thesis that corporate rate changes drive immediate index repricing more effectively than individual rate changes.

But the systematic evidence doesn't back it up. Across all 15 transitions, the correlation between corporate rate changes and regime returns is essentially zero (r = −0.004). The TCJA effect appears to be real but idiosyncratic — a one-time repricing of after-tax corporate earnings that doesn't generalize to a reliable pattern across the full century of data.

The "CG flat" group tells a different story

Perhaps the most surprising finding involves the three regimes where the capital gains rate didn't change but other rates did: 1988 (ordinary income dropped to 28%, corporate to 34%), 1993 (ordinary hiked to 39.6%), and 2018 (corporate slashed to 21%). These "CG flat" regimes averaged 14.8% annualized — the best performance of any group, beating both the cut and hike cohorts.

This suggests that the capital gains rate, which dominates the political narrative around investment taxation, may be less important to equity returns than changes in ordinary income and corporate rates. Investors fixate on the rate applied to their gains, but the rates that shape corporate profitability and the broader tax burden on economic activity may do more of the actual work.

Mean reversion, not anticipation

I also examined whether markets anticipate tax changes, i.e., the "buy the rumor" hypothesis. If investors front-run rate cuts, we'd expect strong trailing returns before cuts and weak trailing returns before hikes. The data shows the opposite: regimes preceded by hikes averaged a 7.4% trailing six-month return, while those preceded by cuts averaged just 1.2%.

This inversion has a straightforward explanation. Many rate cuts may be responses to economic weakness--in other words, Congress cuts taxes because the economy and markets are already struggling. This is a tricky causality wrinkle.

The 1982 Reagan cuts arrived after a −4.8% trailing return; the 1978 cut came amid a −0.8% trailing return. The cuts didn't cause the prior weakness; the weakness prompted the cuts. Conversely, strong trailing returns before hikes often reflect gain-flushing behavior, i.e., investors accelerating the realization of gains to beat the higher rate, which temporarily inflates index returns even as capital is exiting.

The more analytically useful relationship is between the trailing return and the subsequent regime return, which shows a correlation of −0.33. This is a straightforward mean-reversion signal: regimes entered after weak markets tended to produce strong returns, and regimes entered after strong run-ups tended to disappoint. The 1982 regime entered at −4.8% trailing and delivered 19.9% annualized. The 1997 regime entered at +16.3% trailing and delivered only 5.4%.

Starting valuations and macroeconomic conditions (the state of the world when the tax change arrives) appear to matter at least as much as the tax change itself.

What the data can and can't tell us

This is not a dataset that allows for any sort of fancy model, even if you flog the data. Sixteen observations across 112 years, each shaped by world wars, technological revolutions, financial crises, and monetary policy shifts, cannot isolate a clean tax effect. There are no controls, the regime durations are unequal, and the individual and corporate rates are often multi-collinear, meaning they move together in the same legislation.

What the data can do is add some context. It shows that the relationship between tax policy and equity returns is real but modest, noisy, and frequently overwhelmed by other forces. The strongest bull markets in the sample occurred under high-tax regimes (1942, 1993) as often as low-tax ones (1982). The cleanest capital gains cuts (1997, 2003) produced middling returns. And the market's short-term reaction to a tax change often moves in the opposite direction of its long-term performance under that regime.

For investors, the practical implication is that tax policy changes are worth monitoring but dangerous to trade on in isolation. The macro context, e.g., where valuations stand, what the economy is doing, what the Fed is up to, has historically been a far more reliable guide than the direction of the tax rate alone. (See this recent article for a deeper dive.) All that said, going out on a limb, any serious momentum towards achieving anything like the "Billionaire Minimum Income Tax" would be more inimical to the American enterprise system and our capital markets than any of the incremental rate changes I looked at in this article. This sort of analysis couldn't begin to handicap those potential impacts.

A note on data nuances

Three caveats are worth flagging for anyone working with this data. First, the 23.8% capital gains rate shown for 2013 onward reflects the 20% statutory maximum plus the 3.8% Net Investment Income Tax that applies to high earners, so it's not a single legislative rate. Second, the 1970s regime's stated 30.2% capital gains rate understates the real burden because gains were not indexed for inflation; investors were taxed on nominal gains that often represented real losses in purchasing power. Third, the strong +13.2% S&P return in January 1987 (the month the capital gains rate rose from 20% to 28%) reflects a confluence of gain-flushing in late 1986 and optimism about the simultaneous drop in ordinary rates from 50% to 28%, which together neutralized the expected selling pressure.

The primary authority for historical tax rates is the Internal Revenue Service (IRS), supplemented by:

• Individual & Corporate Income: Internal Revenue Service (IRS), Statistics of Income (SOI) Bulletin: Historical Individual Income Tax Rates and Brackets.

• Capital Gains: Wolters Kluwer (Tax Research Division), Historical Capital Gains Rates, and the Tax Foundation, Historical Federal Individual Capital Gains Tax Rates & Collections.

• Legislative Summaries: U.S. Department of the Treasury, Office of Tax Analysis, and the Tax Policy Center (a joint venture of the Urban Institute and Brookings Institution).

Google Gemini was used to in aggregating the data from these various original sources.