Path Dependency Recognition¶

Why It Matters¶

Many analyses focus on average returns, ignoring the path taken. But the path determines several critical outcomes: whether you run out of money during drawdowns, the psychological ability to stay invested, and the actual compounded returns when contributions/withdrawals occur. Early losses in retirement can be catastrophic even if long-term returns are good (sequence-of-returns risk).

The Key Move¶

When evaluating any multi-period outcome, don't just look at the average or final result. Ask: does the sequence matter? Specifically: Are there cash flows (contributions/withdrawals) along the way? Are there stopping rules (going broke ends the game)? Is there a psychological tolerance threshold? Model different orderings of the same events and check if outcomes differ materially. If yes, path dependency is significant.

Classic Application¶

Retirement sequence-of-returns risk. Two retirees with identical average returns but different sequences. Retiree A experiences -20%, -10%, +15%, +25% while withdrawing for living expenses. Retiree B experiences +25%, +15%, -10%, -20% with the same withdrawals. Retiree B ends with significantly more wealth because early gains compound before withdrawals, while Retiree A sells assets at depressed prices, amplifying losses.

Surprising Application¶

Skill acquisition sequencing. Learning programming: starting with high-level Python and later learning C has different outcomes than starting with C and later learning Python, even though you end with both skills. Starting with C builds low-level mental models but creates early frustration (steep learning curve when motivation is lowest). Starting with Python provides early wins (building things quickly) but potentially shallower understanding. The path matters because early experiences affect persistence, motivation, and mental model formation.

Failure Modes¶

Over-focusing on path can cause paralysis—you can't know the future sequence, so you can't perfectly optimize for it. May lead to market timing attempts (trying to avoid bad sequences). Ignores that most outcomes aren't path-dependent—average returns often do dominate. Difficult to model all possible paths—combinatorial explosion. Can confuse path dependency with causation (the sequence may be correlated with other factors).

Go Deeper¶

Milevsky & Robinson, "Self-Annuitization and Ruin in Retirement" (2000); Estrada, "Sequence Risk: Is It Really a Big Deal?" (2017)