Historical Backtesting

Replaying real history

A backtest runs your plan through an actual, ordered sequence of past returns — say, the stock, bond, and inflation figures starting in 1966 — and reports whether the plan would have survived. Because the sequence is real, it preserves the way markets actually behaved: crashes that clustered, decades that went sideways, and inflation that compounded.

The classic stress tests are the hardest starting years in modern history:

• 1929 — the Great Depression crash.
• 1966 — a flat-to-negative real market through the 1970s, compounded by high inflation. Often the single hardest retirement to survive.
• 2000 — the dot-com bust followed by 2008.

Why the order matters

Two retirees can earn the same average return over 30 years and still end up in completely different places — because one hit a bad market early and the other hit it late. Early losses, combined with withdrawals, shrink the portfolio before it can recover. That is sequence-of-returns risk, and a backtest is the most concrete way to see it: a plan that thrives starting in 1982 can fail starting in 1966 on identical assumptions.

Backtest and Monte Carlo — not either/or

Neither tool is complete alone:

• History captures real autocorrelation, mean reversion, and crisis clustering — but it's a single path through one country's past.
• Monte Carlo explores combinations history never produced — but purely random draws can understate how returns cluster in real crises.

The honest approach shows both. If your plan survives the worst historical start years and clears a high Monte Carlo success rate, you can have real confidence in it.

How Horizons uses this

Horizons can replay your plan against historical return sequences alongside its Monte Carlo engine, so you can see how you'd have fared through real downturns — not only synthetic ones.

Check your understanding