Skip to content

The Stock-Bond Correlation and Macroeconomic Conditions: One and a Half Centuries of Evidence: summary and key takeaways

The paper models the stock-bond correlation in US and UK monthly data from 1855 to 2001 as a function of the business cycle, short rates and inflation. The cycle pattern differs by country, and correlation is usually higher after short rates have been high, and somewhat higher after inflation has been high.

Yang, J.; Zhou, Y.; Wang, Z. · 2009 · Journal of Banking & FinanceSummary updated 20 November 2026
Published2009Journal article
Asset focusStocks+BondsStock-bond correlation
TypeResearchStock-Bond Correlation
Reading time4 minthis summary

The Stock-Bond Correlation and Macroeconomic Conditions: One and a Half Centuries of Evidence at a glance

The question
How does the correlation between stock and bond premia change with the business cycle, the short rate and the inflation rate over a very long history?
Dataset & period
US and UK monthly stock and government bond excess returns over short rates, January 1855 to November 2001 (about 150 years). Business cycle coverage: 32 expansions and 32 recessions for the US, 22 expansions and 21 recessions for the UK. Subsamples split at 1926 (US) and 1924 (UK) for data quality.
Method
The authors fit bivariate AR(1)-GARCH(1,1) models to stock and bond premia. They let the conditional correlation depend on dummies (recession, war, monetary regimes) in one version. In a second version it moves between a low-state and a high-state value through a logistic smooth transition driven by the lagged short rate or lagged inflation. They test constant against state-dependent correlation with likelihood-ratio tests and simulated critical values.

Key findings of The Stock-Bond Correlation and Macroeconomic Conditions: One and a Half Centuries of Evidence

  • In the US, the estimated stock-bond correlation is lower in recessions than in expansions. In the UK the pattern is the reverse, with higher correlation in recessions.
  • The authors conclude that US equity holders get a better hedge from bonds, and more diversification, than UK holders do.
  • Correlations tend to rise after periods of higher short rates. The transition between regimes sits around 0.053 to 0.069 for the US, above average short rates.
  • Higher inflation also tends to precede higher correlation, but more weakly. In the US this is clear in the post-1926 subperiod, while in the earlier subperiod the sign reverses and is not significant.
  • The general short-rate and inflation pattern holds across both countries in the post-1923 subperiod, and most clearly for the US over the whole sample.
  • The authors propose a contagion story for the short-rate result: at high rates both stock and bond premia fall and correlation rises. The evidence is weaker in the second subperiod.

Robustness, as reported in the paper

  • Because asymptotic likelihood-ratio critical values may be unreliable here, the authors simulate the test distribution with 200 replications. Constant correlation is still rejected, though the US first subperiod is insignificant at the 10 percent level.
  • With the short rate as transition variable, symmetry is rejected at the 5% level for the US in the whole sample and both subperiods on asymptotic values.
  • A 1972-1999 US subsample, which the authors say matches an earlier study, gives higher correlation after higher inflation but lower correlation after higher short rates.
  • The transition speed is fixed rather than freely estimated, and the authors vary that constraint as a check.

Caveats and limits

  • The authors state that the asymptotic properties of the estimators are unknown, so the results should be read with caution.
  • The fitted transition functions are close to abrupt rather than smooth, especially with inflation as the driver.
  • UK evidence is weaker than US evidence, and early data are spliced from several sources with gaps and an assumed dividend yield (inferred from the data description).
  • Only linear correlation is studied, with two countries, one bond maturity each, and data ending in 2001.

Items marked (inferred) are our reading of the design, not statements by the authors.

Why it matters

The stock-bond correlation is a core input to asset allocation and hedging. The paper shows its sign and size depend on the macro regime, and that the relation differs between countries and monetary eras. Practitioners can use it as a long-history reference when judging correlation assumptions.

Questions about this paper

Is the stock-bond correlation the same in recessions in the US and the UK?

No. The paper finds lower correlation in US recessions than in expansions, and higher correlation in UK recessions.

What predicts a higher stock-bond correlation?

Higher lagged short rates, and to a lesser extent higher lagged inflation, tend to be followed by higher correlation, mainly in the post-1923 subperiod and the US whole sample.

How long is the data sample?

Monthly data from 1855 to 2001 for both the US and the UK.

Where can I read the original paper?

The original is at its publisher: https://doi.org/10.1016/j.jbankfin.2008.11.010. We link to the paper and do not host the PDF.

Follow new quant research

Download MacroFinance market data as CSV or Excel, or pull it by API, and test the ideas in these papers yourself. New summaries are added as we review them.

Create a free account
Summaries are written by MacroFinance from the published papers, in our own words, for information only. They are not investment advice and may omit detail. Papers are not hosted here: each page links to the original at its publisher.