Why an 80% 'success rate' leaves millions unspent
Run a retirement plan through a Monte Carlo simulator and you get one headline number: an 80% (or 90%, or 95%) “success rate.” It reads like a grade, and the instinct is to push it higher by spending less. But the success rate hides what happens in the futures where the plan doesn’t fail — and that’s most of them.
What the other 80% looks like
A fixed-real spending rule withdraws the same inflation-adjusted amount every year no matter what markets do. Feed that rigidity into a range of market futures and the outcomes fan out asymmetrically: spending is capped, but wealth is not. In the bad sequences the portfolio grinds toward zero; in the ordinary and good ones, the withdrawals barely dent the compounding, and the portfolio ends far larger than it started. Historically, a retiree following the classic 4% rule was far more likely to multiply their starting wealth several times over than to run out — the median outcome of a “safe” fixed plan is dying with more real wealth than you retired with. An 80% success rate typically means something like: 20% of futures end badly, and a large share of the rest end with millions unspent.
The trap: buying safety with your own lifestyle
Here is the trap in slow motion. You see 80% and read “20% chance of ruin.” You cut planned spending until the dial says 95%. That extra safety was purchased in every future — including the 80–90% of them where it was never needed. The cost shows up as decades of forgone trips, gifts, and help to family, converted at death into an estate nobody planned for. The success rate is a one-sided metric: it prices failure and gives surplus away for free. It also overstates the risk it does measure — a simulated “failure” assumes a retiree who watches the portfolio collapse for fifteen straight years and never once trims spending. Real retirees adjust; rigid sims don’t.
Guardrails: pricing the surplus back in
Adaptive spending rules — the best known being Guyton–Klinger guardrails — fix the asymmetry by letting spending react in both directions: cut real spending a step when the withdrawal rate drifts too high after bad markets, and take a raise when markets run ahead. Because the plan defends itself, depletion becomes rare and the success rate saturates near 100% — which is precisely why it stops being the interesting number. The risk hasn’t vanished; it has moved into lifestyle volatility, and the honest questions become: how high can spending start, how deep could the cuts go, and is that floor one you could live with? In exchange, the median future stops stockpiling: money gets spent while you’re alive to spend it, and the unplanned legacy shrinks toward whatever legacy you actually intend.
Better questions than “what’s my success rate?”
None of this makes the success rate useless — it is a fine screening metric for a rigid plan. But a decision between plans needs the fuller picture: median lifetime spending (what you actually get to enjoy), the depth and duration of worst-case cuts, the median legacy versus the legacy you intend, and the failure age in the bad tails. Two plans with identical success rates can differ by hundreds of thousands of dollars in expected lived spending. If your simulator can only answer “what percent of paths survived,” it is answering the easiest question, not the most important one.
Try it in Deorbit Plan
The dashboard shows the trap directly: next to the success gauge sits the Terminal net worth (median) tile — the pile a fixed-real plan leaves behind in the typical future — and the Net worth across paths fan chart’s upper percentile bands show how top-heavy the outcomes are. Then open the Strategy Lab (Tool 1): Tool 1 — Max sustainable spending solves for the highest fixed-real spend that meets your success target and reports the median amount left over anyway. Switch the Spending panel to Guardrails (Guyton–Klinger, simplified) and the same tool becomes Tool 1b — Guardrails floor solve, reframing the question from “probability of success” to the spending floor you could actually tolerate.
Educational content only — not financial, tax, or investment advice.
Read this article in the interactive simulator →