Loading calculator...
How this works — and why it's “optimal”
Most glide-path rules (e.g. “100 minus your age”, target-date funds) are heuristics — they have no explicit link to your income target, risk tolerance, or guaranteed income. This tool derives an allocation from first principles instead.
What “optimal” means here
The optimizer finds the equity weight at each age that maximizes your expected utility of retirement consumption, scored with a CRRA (constant relative risk aversion) function — so a dollar of spending counts for more when you'd otherwise fall short than when you're already comfortable. Two preferences shape the curve: γ sets how much you dislike swings in retirement spending (higher → safer, lower-equity plans), and β sets how much you front-load spending into your earlier, more active retirement years. The result is the best risk-adjusted outcome for your situation — not the highest expected return, and not the least volatile.
It maximizes welfare, not success rate: where no stock/bond mix can close a funding gap, the fix lives in your inputs (retire later, spend less, save more), not the allocation. Guaranteed income is assumed to start at retirement and be paid every year — a pre-pension “bridge” is out of scope (use the retirement tool for that funding question). Income shortfall — a year the portfolio can't fund your targeted spending — is reported two ways: drawdown-only (from the expected retirement balance) and full-path (which also includes pre-retirement market luck). See the analysis note below for how γ and β enter the objective and typical values for each.
How it optimizes
The optimizer runs Monte Carlo coordinate ascent: it holds every 5-year step's weight fixed but one, scans all candidate equity weights for that step on a shared set of simulated market paths (common random numbers), keeps the best, then repeats for every other step — cycling until nothing improves. No parametric shape (flat, rising, falling) is assumed; the shape emerges from the math. By default, returns are independent (iid) Monte Carlo draws from our PWL / FP-Canada capital-market assumptions in real (today's) dollars — the model whose recommendation holds up across historical regimes. A sequence-aware mode (the Simulation toggle) instead replays historical stock and bond sequences (the JST Macrohistory cross-country panel) rescaled to the same assumptions and block-bootstrapped, preserving history's multi-year structure — crashes followed by recoveries, persistent inflation decades.
Key findings
The return model and the spending rule set the shape. Under the default independent (iid) mode, rigid constant-$ spending derisks into a moderate “bond tent” near retirement — an interior answer that is never the worst case in any historical era or country cut we tested (it always sits inside the range the sequence-aware mode sweeps across regimes). Under the sequence-aware mode, history's joint pattern (equities recover, while long nominal bonds quietly fail in inflation decades) pushes the optimum to ~100% equity — but that result leans on pre-1990 monetary regimes, surviving developed markets, and long nominal bonds being the only alternative: swap in a real, short-duration holding (e.g. an inflation-linked bond fund) and the optimum drops well below 100%, so we treat the ~100% figure as a scenario rather than the default. Flexible spending (income moves with the market) stays near 100% equity in both — the two modes bracket the answer (see the analysis note's caveats).
The shape is worth little; the level matters. Out of sample, the full per-age glide path beats the best single flat weight by only ~0.5% of certainty-equivalent income under the default iid mode, and essentially nothing under the sequence-aware mode. Getting the equity level right matters far more than optimizing the curve. That is why we also report the best constant equity weight: since its outcome is nearly identical and a single fixed allocation is far easier to hold through market swings, it is often the more practical choice.
Guaranteed income is your bond floor. A larger guaranteed income (CPP + OAS + DB) lets the portfolio take more equity risk because it already covers some downside. With no guaranteed income, rare depleted years dominate CRRA utility, so the web app requires at least $10,000 per year for a meaningful recommendation.
Full methodology: glide-path analysis note. Runs in your browser — nothing is sent anywhere. This is an illustration, not financial advice.