QuantCalcDataSequence of Returns 2026

Sequence-of-Returns Risk by Retirement Cohort 2026

For every U.S. retirement start year from 1871 to 1993, the realised first-5-year and first-10-year annualized real return sequence and the terminal outcome of the 4% rule — the empirical record behind sequence-of-returns risk.

QuantCalc Research · Generated 2026-07-22 · v2026.1 · CC-BY-4.0 dataset

Two retirees can follow the identical 4% plan and end 30 years apart in wealth — not because one earned a higher average return, but because of when the good and bad years landed. This dataset makes that concrete with history rather than simulation: for all 123 rolling 30-year retirement cohorts in Robert Shiller's long-horizon record, it reports the annualized REAL return of the first 5 and first 10 years and the terminal wealth of a 60/40 portfolio drawn down at 4% real. The single number that summarises sequence risk is the rank correlation between the first five years and the 30-year outcome: 0.633.

0.633
First-5yr vs. terminal (Spearman)
0.795
First-10yr vs. terminal (Spearman)
123
30-year cohorts (1871–1993)
1.31×
Median 4% terminal wealth

The headline: Across all 123 historical cohorts, the first five years' annualized real return has a Spearman rank correlation of 0.633 with 30-year terminal wealth under the 4% rule — and the first ten years, 0.795. The full 30-year average return correlates 0.799. In plain terms, where a retirement started tells you most of where it ended: the worst cohort (1965, first-5-year real -1.18%) finished at 0.000× its starting balance, while the best (1982, first-5-year real 17.87%) finished at 5.630×. This is the citable dataset companion to our 50,000-path Monte Carlo study, which isolates the same effect by reordering the first decade.

Per-cohort record (123 cohorts)

Click any column header to sort. Terminal wealth is a multiple of the starting balance after 30 years of 4% real withdrawals (0.000× = the plan depleted before year 30). "Max sustainable rate" is the highest constant real withdrawal rate that would have survived the full 30 years — the same figure published in the safe-withdrawal-by-cohort study, computed by the same imported code.

Retirement start year First 5yr real (annualized %) First 10yr real (annualized %) Full 30yr real (annualized %) 4% rule terminal wealth (x start) 4% rule survived Max sustainable rate (%)
18719.65%11.74%7.88%5.359×Yes8.89%
18726.07%10.06%7.89%5.327×Yes8.81%
18738.82%10.03%7.56%4.962×Yes9.04%
187412.48%10.09%7.18%4.555×Yes9.30%
187511.52%9.20%7.29%4.532×Yes8.87%
187613.88%10.22%7.32%4.411×Yes8.50%
187714.20%11.89%7.46%5.161×Yes9.89%
187811.25%9.39%6.36%3.534×Yes9.01%
18797.75%7.94%6.27%3.025×Yes7.81%
18806.93%8.11%5.96%2.710×Yes7.65%
18816.67%5.09%5.37%1.849×Yes6.50%
18829.63%7.68%5.74%2.402×Yes7.27%
18837.57%6.78%5.47%2.238×Yes7.32%
18848.13%6.57%5.17%2.057×Yes7.31%
18859.30%6.94%4.90%1.919×Yes7.37%
18863.53%4.90%4.70%1.445×Yes6.29%
18875.75%4.80%4.22%1.194×Yes6.11%
18886.00%6.29%3.36%1.075×Yes6.65%
18895.02%7.13%2.92%0.880×Yes6.36%
18904.63%4.71%2.42%0.672×Yes5.95%
18916.29%6.92%2.42%0.785×Yes6.48%
18923.86%5.97%2.53%0.600×Yes5.58%
18936.58%5.92%3.25%0.847×Yes5.92%
18949.27%4.94%3.21%0.830×Yes5.90%
18954.79%5.75%3.50%0.807×Yes5.62%
18967.55%6.92%3.85%0.923×Yes5.69%
18978.13%5.83%4.01%0.898×Yes5.53%
18985.26%3.48%4.44%0.850×Yes5.20%
18990.77%3.79%4.79%0.516×Yes4.58%
19006.72%5.09%5.11%1.137×Yes5.37%
19016.28%4.13%4.36%0.600×Yes4.80%
19023.58%3.62%3.34%0.311×Yes4.52%
19031.73%3.72%3.79%0.520×Yes4.82%
19046.89%4.03%4.96%1.017×Yes5.25%
19053.48%2.06%4.20%0.478×Yes4.65%
19062.02%2.34%4.71%0.193×Yes4.20%
19073.66%2.07%5.41%0.529×Yes4.49%
19085.75%0.40%5.13%1.086×Yes5.28%
19091.25%-1.95%4.90%0.488×Yes4.53%
19100.66%-2.37%5.04%0.637×Yes4.68%
19112.66%-3.50%4.64%0.480×Yes4.56%
19120.51%-1.84%3.88%0.355×Yes4.51%
1913-4.67%0.19%4.16%0.550×Yes4.77%
1914-5.06%0.70%4.63%0.808×Yes5.05%
1915-5.30%2.71%5.08%1.145×Yes5.40%
1916-9.29%2.37%5.22%0.764×Yes4.79%
1917-4.13%4.15%4.61%0.943×Yes5.29%
19185.31%9.64%5.35%2.181×Yes7.36%
19196.80%13.08%5.70%2.685×Yes8.15%
192011.40%13.17%6.30%3.500×Yes9.09%
192115.52%13.09%6.88%4.532×Yes10.39%
192213.15%8.51%6.42%3.659×Yes9.19%
192314.15%7.58%6.09%3.065×Yes8.33%
192419.73%10.37%6.10%3.188×Yes8.70%
192514.97%7.92%6.41%3.194×Yes7.92%
192610.72%9.59%6.52%3.087×Yes7.47%
19274.05%10.17%6.14%2.620×Yes7.12%
19281.39%5.57%5.19%1.550×Yes6.05%
19291.75%4.12%4.93%0.754×Yes4.87%
19301.30%4.88%5.15%1.059×Yes5.23%
19318.48%4.97%5.59%1.579×Yes5.79%
193216.65%5.24%6.67%3.192×Yes7.40%
19339.91%4.85%6.32%2.791×Yes7.19%
19346.55%3.06%5.70%1.611×Yes5.76%
19358.58%4.68%6.24%2.294×Yes6.39%
19361.58%3.85%5.43%1.044×Yes5.09%
1937-5.05%-0.22%4.66%0.434×Yes4.50%
19380.03%1.03%5.55%1.507×Yes5.70%
1939-0.32%0.29%5.19%1.086×Yes5.25%
19400.92%1.20%4.58%0.923×Yes5.27%
19416.17%2.85%4.97%1.314×Yes5.77%
19424.86%5.56%5.85%2.437×Yes7.17%
19432.04%5.86%5.92%2.490×Yes7.19%
19440.91%4.99%4.89%1.740×Yes6.84%
19451.47%6.67%3.74%1.165×Yes6.52%
1946-0.36%6.19%3.62%0.820×Yes5.57%
19476.26%8.77%4.62%1.757×Yes7.31%
19489.84%9.14%4.51%1.961×Yes8.37%
19499.23%10.62%4.35%1.883×Yes8.43%
195012.13%9.54%3.78%1.501×Yes7.89%
195113.17%9.03%3.53%1.313×Yes7.46%
195211.34%9.27%2.84%1.027×Yes7.18%
19538.46%8.28%3.41%1.141×Yes6.87%
195412.03%9.14%3.61%1.238×Yes6.98%
19557.02%7.38%2.93%0.693×Yes5.64%
19565.04%6.26%3.04%0.485×Yes4.98%
19577.24%5.63%3.82%0.668×Yes5.11%
19588.10%6.63%3.76%0.838×Yes5.53%
19596.33%4.90%3.37%0.426×Yes4.75%
19607.74%3.19%3.70%0.470×Yes4.75%
19617.49%3.14%3.40%0.355×Yes4.60%
19624.05%2.83%3.71%0.173×Yes4.25%
19635.19%3.66%3.90%0.281×Yes4.39%
19643.48%0.72%3.83%0.049×Yes4.06%
1965-1.18%-2.51%3.33%0.000×No3.79%
1966-1.03%-1.41%3.97%0.000×No3.69%
19671.62%-0.33%4.57%0.000×No3.99%
19682.15%-1.92%5.05%0.000×No3.92%
1969-1.97%-2.09%5.63%0.000×No3.93%
1970-3.83%-1.11%6.22%0.847×Yes4.64%
1971-1.79%-1.32%6.01%0.695×Yes4.55%
1972-2.24%-3.20%5.49%0.420×Yes4.37%
1973-5.82%-1.48%4.88%0.199×Yes4.20%
1974-2.22%1.19%6.03%1.331×Yes5.19%
19751.69%4.18%6.92%3.083×Yes6.83%
1976-0.85%4.43%6.44%2.273×Yes6.15%
1977-4.14%6.30%6.47%2.285×Yes6.14%
19783.06%6.79%6.86%3.275×Yes7.24%
19794.71%7.54%6.21%2.912×Yes7.65%
19806.72%9.30%6.81%3.803×Yes8.46%
19819.99%8.61%7.03%4.177×Yes8.78%
198217.87%12.05%7.61%5.630×Yes10.64%
198310.66%9.83%7.02%4.305×Yes9.15%
198410.45%9.84%7.09%4.372×Yes9.10%
198511.95%8.63%7.29%4.707×Yes9.30%
19867.24%9.14%6.49%3.451×Yes8.38%
19876.53%7.93%6.00%2.529×Yes7.15%
19889.01%10.66%6.67%3.509×Yes8.10%
19899.22%11.95%6.25%3.103×Yes8.05%
19905.42%10.88%6.39%3.034×Yes7.59%
199111.08%11.16%6.88%3.826×Yes8.32%
19929.35%8.21%6.31%2.855×Yes7.35%
199312.34%6.60%5.47%2.284×Yes7.44%
Download CSV (123 cohorts) Download JSON (rows + aggregates)

CC-BY-4.0 — free for any use including republication and journalism, with attribution to QuantCalc Research.

Methodology

Data: Every figure is computed from the committed snapshot research/data/shiller-annual-2026-07.csv (Robert J. Shiller, "Irrational Exuberance" long-horizon dataset (ie_data.xls, Data sheet), retrieved 2026-07-21). This is HISTORY, not a forecast or a simulation.

Returns: Annual REAL (CPI-deflated) total returns on a 60/40 stock/bond portfolio, rebalanced yearly. The bond leg is a 10-year constant-maturity par-bond total return derived from the GS10 yield move. Annualized figures are geometric.

4% rule terminal wealth: Starting from 1.000×, withdraw 4% of the initial balance (a set real amount) at the start of each year, then apply that year's real return, for 30 years. A cohort whose balance is driven negative is depleted and reports 0.000×. This withdrawal loop is IMPORTED verbatim from scripts/gen_cohort_swr_study.py (its survives() convention), so terminal wealth > 0 exactly when that study reports survival, and the max-sustainable-rate column here equals its max_swr column cohort-for-cohort.

Headline correlation: Spearman rank correlation (average-rank tie handling) between the first-5-year annualized real return and the 30-year terminal wealth, across all 123 cohorts: 0.633. Rank correlation, not Pearson, so it measures monotone ordering and is not distorted by the pile-up of depleted cohorts at 0.000×.

Scope: A rolling-cohort historical description, not a probability. It reports what happened to each start date; it does not forecast any future cohort. For a forward-looking, distributional treatment see the Monte Carlo study.

Reproducibility: Generated by scripts/gen_sequence_risk_dataset.py; re-running against the same snapshot reproduces byte-identical rows.

Not advice: Reference data for research and planning, not financial advice.

Stress-test your own start date

Sequence risk is about your first decade, which you can't choose. The QuantCalc planner runs your plan through thousands of ordered return paths — including bad-first-decade sequences — so you can see the spread of outcomes, not just the average.

Open the planner →

Changelog

Dataset license: CC-BY-4.0. QuantCalc is an independent retirement-planning research project. Not financial, tax, or legal advice.

Cite this dataset

QuantCalc Research (2026). Sequence-of-Returns Risk by Retirement Cohort 2026. https://quantcalc.app/data/sequence-of-returns-2026/ (accessed <date>).

BibTeX
@misc{quantcalc2026sequenceofreturnsriskbyretirementcohort2,
  title  = {Sequence-of-Returns Risk by Retirement Cohort 2026},
  author = {{QuantCalc Research}},
  year   = {2026},
  url    = {https://quantcalc.app/data/sequence-of-returns-2026/},
  note   = {Accessed <date>}
}

Machine-readable citation metadata (schema.org identifier and citation fields) is embedded in this page's JSON-LD, at the stable identifier https://quantcalc.app/data/sequence-of-returns-2026/.