Asset performance by macro regime
Median monthly return of equities, bonds, gold and energy by U.S. macro regime. Interactive, descriptive heatmap.
An asset’s return is best read not in isolation but against the macro regime around it: the same equity index does not behave the same way in a balanced expansion and under inflationary pressure.
The matrix below crosses four asset classes with the seven cyclical states identified by Eco3min’s regime classification, over January 2003 – May 2026. Each cell shows the median monthly return observed during the months classified in that regime. It is a backward-looking measure: it describes what happened, it forecasts nothing.
Colour = median monthly return observed in this regime. Blue = positive, terracotta = negative. This describes past performance — not a forecast and not an allocation recommendation.
Grey cells: insufficient sample (fewer than 12 months, or a single episode) — not displayed.
Sources: Kenneth French Data Library (equities, North America, Mkt-RF + RF) · U.S. Treasury (bonds, 10Y CMT, total-return proxy) · World Bank — Pink Sheet (gold; energy, spot index). Computation: Eco3min Research.
Monthly median, January 2003 – May 2026 · classification v1.0.0 · values rounded.
Past performance, measured after the fact within each regime. It does not predict future returns and is neither advice nor an allocation recommendation.
How to read the matrix
Colour encodes a continuous dimension: blue for a positive median, terracotta for a negative one, the intensity rising with magnitude. The statistic is the monthly median, more robust than the mean to the extreme months that punctuate crisis regimes. On hover, each cell shows its interquartile range, its month count and its number of distinct episodes.
Three columns appear in grey. Overheating, Stagflation and Disinflationary Contraction were each traversed only once over the period, on too few months for a median to be interpretable. They read “n/a” rather than implying a regularity drawn from a single episode.
Why the regime conditions returns
Three channels connect the macro backdrop to returns. The discount rate first: under inflationary pressure, rising rates mechanically weigh on the valuation of long-duration assets. The risk premium next: in a stressed regime, investors demand higher compensation, which depresses prices before letting them rebound. Inflation pass-through last: a real underlying such as energy reacts differently than a nominal bond to the same price shock.
The matrix does not theorise these channels, it shows their numerical imprint. On the sufficiently populated columns, equities post their highest median outside crisis under inflationary pressure (+2.54% per month, essentially 2022-2024), while bonds stay close to zero whatever the regime — consistent with a total-return proxy dominated by rate moves that roughly net out over the median month.
The “Slowdown” column trap
Reading “+3% per month for equities in a slowdown” as a sign that markets weather slowdowns inverts the chronology: the classification registers the slowdown with a lag, once the equity drawdown has already happened. The window therefore captures mostly the rebound phases.
The highest cell in the entire matrix is equities in the Slowdown regime: a median monthly return of +3.03%, with a wide interquartile range, from −2.10% to +5.72%. The figure surprises, then explains itself. This regime has been active only twice since 2003: from April 2008 to November 2009, then from April 2020 to April 2021.
In both cases, the macro diagnosis of slowdown arrives after the market shock. The February-March 2020 crash predates the regime switch, which came in April: the COVID window contains almost nothing but the recovery. The 2008-2009 window does include the October 2008 collapse, but also the entire 2009 rebound, throughout which the regime stayed labelled a slowdown. The high median and the very wide range say the same thing: the cell adds up two episodes of opposite natures, not a regularity.
The lesson goes beyond this cell. A dated regime state is not a leading signal: it is a time-lagged description. The matrix measures what happened during officially classified months, not what one ought to do when one believes a regime is beginning.
Limits
Twenty-three years of data hold only a limited number of regimes, and the most distinctive ones are the rarest — hence the three greyed columns. The matrix is therefore more solid on the ordinary, frequent regimes than on the extremes. The classification lag described above applies to every column, not only the Slowdown. Two series also rest on acknowledged proxies: bonds are approximated by the total return of a U.S. 10-year sovereign (U.S. Treasury, CMT rate), and energy is a spot index, not investable as such. The figures are frozen in a dated version and will be revised explicitly, never silently.
A dated macro regime describes the recent past, it does not run ahead of it: an asset’s median by regime measures an observed reaction, never an entry point.
- The same asset class posts very different medians across regimes: the macro backdrop is a reading variable, not a contextual detail.
- A median by regime is a reaction measured after the fact; a dated state describes the recent past and is not a leading signal.
- The most striking regimes — overheating, stagflation, contraction — are also the rarest over the period, and the matrix is deliberately silent on them.
To anchor the starting point of any read, you can consult the macro regime in force today. The inflationary column maps in turn to the inflationary macro regime and its mechanisms.
Last updated — 1 June 2026
Disclaimer – Financial Information: The analyses, commentary, and content published on eco3min.fr are provided for informational and educational purposes only. They do not constitute investment advice or a solicitation to buy or sell financial instruments. Past performance is not indicative of future results. All investment decisions involve risk and are the sole responsibility of the reader.