{"paper_id":"2311.06665","title":"Withdrawal Success Optimization","authors":["Hayden Brown"],"paper_date":"2023-11-11","paper_type":"empirical","paper_url":"https://arxiv.org/pdf/2311.06665v1","page_url":"https://thequant.space/flowcharts/withdrawal-success-optimization/","scores":{"empirical_rigor":8,"math_complexity":7.5,"quadrant":"Holy Grail"},"proposes":{"status":"available","statement":"The paper proposes to maximize the probability of completing a given schedule of investments and withdrawals by optimizing progressively measurable portfolio weight functions.","quote":"For $n$ assets and discrete-time rebalancing, the probability to complete a given schedule of investments and withdrawals is maximized over progressively measurable portfolio weight functions. Given n assets available for investment, the goal is to maximize the probability to complete the schedule, with the maximization occurring over time-adapted portfolio weights.","summary":"For n assets and discrete-time rebalancing, the probability to complete a given schedule of investments and withdrawals is maximized over progressively measurable portfolio weight functions. Applications consider two assets, namely the S&P Composite Index and an inflation-protected bond. The…","strategy":null},"assumptions":{"status":"available","items":[{"assumption":"The number of investments and withdrawals is finite and completely determined at time 0.","type":"execution","stated":"explicit","quote":"Assume the number of investments and withdrawals are finite and completely determined at time 0.","support":"full"},{"assumption":"Rebalancing times are finite and contain all investment and withdrawal times.","type":"execution","stated":"explicit","quote":"Assume the set of rebalancing times is finite and contains the set of investment and withdrawal times.","support":"full"},{"assumption":"Annual S&P returns are independent and identically distributed.","type":"statistical","stated":"explicit","quote":"Observing that X2(t) is deterministic, it suffices to justify the assumption that annual S&P returns are iid. The p-value of .078 is slightly concerning, but not enough to reject independence with the usual 95% confidence.","support":"full"},{"assumption":"The inflation‑adjusted bond return X2(t) is deterministic.","type":"model","stated":"explicit","quote":"Table 2 p-values of Ljung-Box test on annual S&P returns returns log-returns absolute log-returns lag = 1 .928 .833 .555 lag = 5 .113 .078 .975 inflation-adjusted value of the S&P Composite Index at time t, and let X2(t) = (1 + r)t, meaning X2(t) denotes the inflation-adjusted value of an inflation-protected bond, with interest r, at time t.","support":"full"},{"assumption":"When a mortality distribution is present, death rates are taken from the 2017 US Social Security female life table.","type":"data","stated":"explicit","quote":"Death rates are taken from https://www.ssa.gov, the official website of the Social Security Administration. In particular, the female per-age death rates of the US Social Security area population are taken from the 2017 period life table.","support":"full"},{"assumption":"We assume the assets’ price processes are inflation‑adjusted, so cash flows ck and wealth Wk are also inflation‑adjusted.","type":"model","stated":"explicit","quote":"When a mortality distribution is present, adjust the previous F as follows F(t) ←σ \u0010 F(t) ∪ [ i−1≤t σ(τi) \u0011 , t ∈T. Since X1(t) and X2(t) are inflation-adjusted, it follows that the ck and Wk are also inflation-adjusted.","support":"full"}]},"datasets":{"status":"available","items":[{"name":"S&P Composite Index annual data (price, dividend, CPI)","provider":"Robert Shiller / Yale","kind":"market-prices","access":"public","period":"1871-2020","universe":"S&P Composite Index (Cowles, S&P 90, S&P 500)","quote":"4.1 Data Annual data from the S&P Composite Index and Comsumer Price Index is taken from http://www.econ.yale.edu/∼shiller/data.htm, collected for easy access at https://github.com/HaydenBrown/Investing. Note that S&P Composite Index refers to Cowles and Associates from 1871 to 1926, Standard & Poor 90 from 1926 to 1957 and Standard & Poor 500 from 1957 to 2020.","support":"full"},{"name":"Female per‑age death rates (2017 life table)","provider":"Social Security Administration","kind":"macro","access":"public","period":"2017","universe":"US Social Security area population (female)","quote":"Death rates are taken from https://www.ssa.gov, the official website of the Social Security Administration. In particular, the female per-age death rates of the US Social Security area population are taken from the 2017 period life table.","support":"full"}],"known":[],"sample":{"start":null,"end":null,"market":null,"asset_class":null,"frequency":null},"data_resolution":"weekly-monthly"},"transaction_costs":{"status":"available","modeled":"no","headline_net_of_costs":null,"detail":null,"quote":null,"source":"front_matter","rule":"R1","candidates":{"front_matter":"no","evidence_card":"no","evidence_card_quoted":false},"meaning":"a trading or execution result is reported without costs"},"results":{"status":"available","evaluation_type":"none","findings":[{"finding":"Using optimal portfolio weights yields noticeable improvements in the probability of completing a schedule compared with constant weights such as 90% S&P / 10% bond.","kind":"performance","quote":"Results indicate noticeable improvements in the probability to complete a given schedule when optimal portfolio weights are used instead of constant portfolio weights like the standard of keeping 90% in the S&P Composite Index and 10% in inflation-protected bonds.","support":"full"},{"finding":"Constant weight (1,0) (100% S&P) produces probabilities close to the optimum, but the gap can be several tenths.","kind":"performance","quote":"Of the constant port-folio weights, πi = (1, 0) appears to produce P(Wk ≥0) and P(W k ≥0) that are closest to their maximum versions; however, the difference can be several tenths.","support":"full"},{"finding":"Optimal weights can offer investors a worthwhile improvement in the probability of completing a given schedule.","kind":"theoretical","quote":"So optimal portfolio weights can offer investors a worthwhile improve-ment in their probability to complete a given schedule of investments and withdrawals.","support":"full"},{"finding":"Overall maximization of P(Wk ≥ 0) over time‑adapted weights provides a noticeable, though sometimes small, improvement over constant‑weight strategies.","kind":"theoretical","quote":"5 Conclusion In general, maximization of P(Wk ≥0) and P(W k ≥0) over the πi offers a noticeable, but sometimes small, improvement over the P(Wk ≥0) and P(W k ≥0) produced when the πi are constant over i.","support":"full"}],"reported_performance":null,"out_of_sample":{"method":"none","headline_out_of_sample":null}},"independent_evidence":{"status":"none_found","supports":[],"contradicts":[],"qualifies":[],"compared_with":13,"citations":{"cited_by_in_corpus":1,"by_independent_authors":0,"how_cited":{"used_as_baseline":0,"uses_method_or_data":0,"extends":0,"discusses_findings":0,"mentions":0,"unclassified":1},"independent_papers":[],"scope":"Citations between papers in this corpus only; citations from outside it are not counted."}},"implementation":{"status":"available","public":true,"code_url":"https://github.com/HaydenBrown/Investing","note":null},"replication":{"status":"none_found","card":null,"note":"No replication of this paper has been run by thequant.space yet."},"extraction":{"text_source":"fulltext","model":"openai/gpt-oss-120b","version":2,"extracted_at":"2026-11-14T20:00:30Z"},"provenance":{"claims":"Everything here is what the authors state in the paper. Nothing was re-run or reproduced by the platform.","quotes":"Assumptions, datasets, cost detail and findings were extracted by a language model. Every quote is verbatim paper text. Where an item has `support`, \"full\" means every figure in it appears in the quoted sentences and \"partial\" means a figure is stated elsewhere in the paper; items with a figure found nowhere in the paper were discarded.","datasets_known":"datasets.known lists well-known vendors and public datasets by a stable id, found by a dictionary over the full text where the sentence says the data is used; it is independent of the model and covers only names in the dictionary.","flags":"transaction_costs.modeled, results.evaluation_type and results.out_of_sample.method are platform classifications (structured-decision models); see https://thequant.space/score-guide/#penalty-flags","comparisons":"supports / contradicts / qualifies are a language model's reading of two papers' quoted findings, restricted to closely related papers that share no author. They are pointers to read, not verified replications, and coverage is partial.","implementation":"code_url is a repository the paper itself presents as its own (rule-scored, ambiguous cases judged by a model), checked for liveness when it was recorded."},"docs":"https://thequant.space/agents/"}