{"paper_id":"2410.20060","title":"Constrained portfolio optimization in a life-cycle model","authors":["Wenyuan Li","Pengyu Wei"],"paper_date":"2024-10-26","paper_type":"theoretical","paper_url":"https://arxiv.org/pdf/2410.20060v1","page_url":"https://thequant.space/flowcharts/constrained-portfolio-optimization-in-a-life-cycle-model/","scores":{"empirical_rigor":3.5,"math_complexity":9.2,"quadrant":"Lab Rats"},"proposes":{"status":"available","statement":"The paper proposes a dual control neural network approach to compute tight lower and upper bounds for constrained life‑cycle portfolio optimization, extending the artificial markets simulation method.","quote":"Moreover, we propose a dual control neural network approach to compute tight lower and upper bounds for the original problem, which can be utilized in more general cases than the simulation of artificial markets strategies (SAMS) approach in Bick et al. (2013).","summary":"This paper considers the constrained portfolio optimization in a generalized life-cycle model. The individual with a stochastic income manages a portfolio consisting of stocks, a bond, and life insurance to maximize his or her consumption level, death benefit, and terminal wealth. Meanwhile, the…","strategy":null},"assumptions":{"status":"available","items":[{"assumption":"The interest rate process satisfies E[exp(∫0^T |r_t| dt)] < ∞.","type":"model","stated":"explicit","quote":"The interest rate process rt satisﬁes E \u0014 exp \u0012Z T 0 |rt|dt \u0013\u0015 < ∞, where | · | means the absolute value. Remark 2.1.","support":"full"},{"assumption":"The income process satisfies a bounded‑expectation condition sup_v E_Qv[∫0^T e^{-∫0^t r_s ds} Y_t dt] ≤ K_y.","type":"data","stated":"explicit","quote":"To satisfy the lower boundedness property (6) of wealth process Wt, we add the following assumption to the income process Yt Assumption 3.1. sup v∈N∗EQv \u0014Z T 0 e− R t 0 rs+λx+sdsYtdt \u0015 ≤Ky, (13) for some positive constant Ky > 0.","support":"full"},{"assumption":"Trading strategies are restricted to a non‑empty closed convex set that includes non‑tradeable asset, no‑short‑selling and no‑borrowing constraints as special cases.","type":"execution","stated":"explicit","quote":"Meanwhile, the individual faces a convex-set trading constraint, of which the non-tradeable asset constraint, no short-selling constraint, and no borrowing constraint are special cases.","support":"full"},{"assumption":"The financial market consists of n+1 assets: a locally risk‑free bond and n risky stocks driven by an n‑dimensional Brownian motion.","type":"market","stated":"explicit","quote":"In the ﬁnancial market, there are n+1 assets. The ﬁrst asset is the bond which is locally risk free and pays no dividends.","support":"full"},{"assumption":"The stochastic income and asset dynamics are defined on a filtered complete probability space.","type":"model","stated":"explicit","quote":"2.1 Financial market Let (Ω, F, P) be a ﬁltered complete probability space.","support":"full"}]},"datasets":{"status":"none_found","items":[],"known":[],"sample":{"start":null,"end":null,"market":null,"asset_class":null,"frequency":null},"data_resolution":"synthetic"},"transaction_costs":{"status":"not_applicable","modeled":"n/a","headline_net_of_costs":null,"detail":null,"quote":null,"source":"front_matter","rule":"R3","candidates":{"front_matter":"n/a","evidence_card":"n/a","evidence_card_quoted":false},"meaning":"the paper reports no trading or execution result"},"results":{"status":"available","evaluation_type":"none","findings":[{"finding":"When trading constraints are imposed, the individual reduces demand for life insurance.","kind":"theoretical","quote":"Finally, we conclude that when considering the trading constraints, the individual will reduce his or her demand for life insurance. Similarly to Figure 4, Figure 7 also shows that when considering the trading constraint, the individual reduces their demand for life insurance.","support":"full"},{"finding":"The dual control neural network yields tight lower and upper bounds for the primal problem, with the lower bound J(0,…) ≤ J(c,M,W_T) ≤ eJ(0,…).","kind":"theoretical","quote":"From all things above, we obtain the tight lower and upper bounds for the primal Problem (P) J(0, W v∗,0, Y0; v∗) ≤J(c, M, WT) ≤eJ(0, Wv∗,0, Y0; v∗). Remark 6.1.","support":"full"},{"finding":"When risk‑free rate, stock appreciation, and volatility are constant, the neural network method performs similarly to the SAMS method.","kind":"performance","quote":"We ﬁnd that when the risk-free interest rate, stock appreciation rate, and volatility are all constant, the SAMS method and neural network performance are very close.","support":"full"},{"finding":"With a time‑perturbed stock appreciation rate, the SAMS method shows a large duality gap (0.1663950) and welfare loss (3.9263%), while the neural network with snake activation reduces the gap to 0.0230592 and welfare loss to 0.5516%.","kind":"performance","quote":"From Table 2, we see that Method 1 generates a big duality gap of 0.1663950, a relative gap of 1.9828%, and suﬀers from a large welfare loss of 3.9263%. More speciﬁcally, the duality gap shrinks from 0.1663950 to only 0.0230592, the relative gap re-duces from 1.9828% to 0.2762%, and the welfare loss decreases from 3.9263% to 0.5516%.","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":false,"code_url":null,"note":"No repository presented by the authors as their own was found in the paper. Absence here is not proof that no code exists."},"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":"2026-11-14T19:54:23Z"},"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/"}