Triumph of the Optimists

Dimson, Marsh and Staunton assembled consistent annual returns for equities, bonds, bills, inflation and currencies for sixteen countries from 1900 to 2000, correcting for the survivorship bias that had distorted every prior long-run study by using only markets that survived and prospered.

The findings changed what could responsibly be claimed about long-run returns. Equities outperformed bonds and bills in every one of the sixteen countries over the century, which supports the equity risk premium. But the premium was substantially smaller than the American-only figures then in circulation, roughly four to five percent over bills globally rather than the six to seven percent commonly cited. Several markets suffered total or near-total losses at some point, including Russia and China where investors lost everything, and Germany, Japan and Italy where real returns were catastrophic for extended periods. The United States was among the best-performing markets of the century, which means using American data to estimate expected returns systematically overstates them.

They also document that dividend income rather than capital appreciation supplied most of the real return, that real bond returns over the century were poor in many countries, and that inflation destroyed fixed income wealth repeatedly.

The book is data-heavy with extensive tables and charts, and it functions as a reference. The database is updated annually in the Credit Suisse and now UBS Global Investment Returns Yearbook, which is where current figures should be taken from. The 2002 volume is the foundational statement of the method and the findings.