{"paper_id":"2609.36405","title":"Finite-Horizon Reversible Investment under Multi-Factor Dynamics","authors":["Junkee Jeon","Takwon Kim","Jinwan Park","A. Max Reppen"],"paper_date":"2026-09-29","paper_type":"theoretical","paper_url":"https://arxiv.org/pdf/2609.36405","page_url":"https://thequant.space/flowcharts/finite-horizon-reversible-investment-under-multi-factor-dynamics/","scores":{"empirical_rigor":4,"math_complexity":9,"quadrant":"Lab Rats"},"proposes":{"status":"available","statement":"The paper proposes a finite‑horizon reversible investment model in which a risk‑neutral firm adjusts productive capacity via costly upward investment and lower‑value downward disinvestment, driven by an n‑dimensional geometric Brownian motion, and reduces the marginal capacity problem to a parabolic","quote":"We study a finite-horizon reversible investment problem in which a risk-neutral firm adjusts capacity at a proportional purchase cost and a lower salvage value under multi-factor geometric Brownian motion. This paper studies a finite-horizon reversible investment problem in which a risk-neutral firm adjusts productive capacity through two finite-variation controls representing investment and disinvestment.","summary":"We study a finite-horizon reversible investment problem in which a risk-neutral firm adjusts capacity at a proportional purchase cost and a lower salvage value under multi-factor geometric Brownian motion. Via the singular control–optimal switching correspondence, the marginal value of capacity…","strategy":null},"assumptions":{"status":"available","items":[{"assumption":"The firm is risk‑neutral and faces a constant proportional purchase cost that exceeds the proportional salvage value for disinvestment.","type":"model","stated":"explicit","quote":"Upward adjustment incurs a constant proportional cost, while downward adjustment produces a smaller proportional salvage value.","support":"full"},{"assumption":"The economic environment follows an n‑dimensional geometric Brownian motion with each state coordinate interpreted as a favourable economic index, implying marginal revenue is nondecreasing in every coordinate.","type":"model","stated":"explicit","quote":"The economic environment is driven by an n-dimensional geometric Brownian motion, and the operating revenue may depend jointly on all of its components. Because the free-boundary analysis assumes that marginal revenue is nondecreasing in every state coordinate, each coordinate is interpreted as a favourable economic index.","support":"full"},{"assumption":"Marginal revenue is locally Lipschitz and satisfies monotonicity and growth conditions required for the free‑boundary analysis.","type":"model","stated":"explicit","quote":"Because the free-boundary analysis assumes that marginal revenue is nondecreasing in every state coordinate, each coordinate is interpreted as a favourable economic index. Exact Leontief and production-switching specifications are nonsmooth benchmarks, and are covered as well: the free-boundary assumption is stated in terms of a comparable smooth majorant of marginal revenue, precisely so that marginal revenue","support":"full"},{"assumption":"Capacity adjustments are represented by two finite‑variation controls (investment and disinvestment) with singular control–optimal switching correspondence.","type":"model","stated":"explicit","quote":"This paper studies a finite-horizon reversible investment problem in which a risk-neutral firm adjusts productive capacity through two finite-variation controls representing investment and disinvestment. Each τn is an F-stopping time taking values in [0, T].","support":"full"},{"assumption":"The revenue function Π(x,u) is bounded by a linear function of capacity and a state‑dependent term, ensuring integrable discounted revenue for any fixed capacity level.","type":"data","stated":"explicit","quote":"By construction, 0 ≤Π(x, u) ≤ρK−u + Ψ(x), (x, u) ∈(0, ∞)n × (0, ∞), (2.5) so every fixed capacity level has integrable discounted revenue, Ex \u0014Z T 0 e−ρtΠ(Xt, u) dt \u0015 < ∞, u > 0.","support":"full"},{"assumption":"The free‑boundary analysis assumes the existence of a smooth majorant of marginal revenue, allowing the marginal revenue itself to be only locally Lipschitz.","type":"model","stated":"explicit","quote":"Exact Leontief and production-switching specifications are nonsmooth benchmarks, and are covered as well: the free-boundary assumption is stated in terms of a comparable smooth majorant of marginal revenue, precisely so that marginal revenue itself need only be locally Lipschitz.","support":"full"}]},"datasets":{"status":"none_found","items":[],"known":[],"sample":{"start":null,"end":null,"market":null,"asset_class":null,"frequency":null},"data_resolution":"none"},"transaction_costs":{"status":"not_applicable","modeled":"n/a","headline_net_of_costs":null,"detail":null,"quote":null,"source":"front_matter","rule":"R4a","candidates":{"front_matter":"n/a","evidence_card":"yes","evidence_card_quoted":true},"card_detail":"Upward investment incurs a constant proportional cost K+; downward adjustment yields a smaller proportional salvage value K−.","card_quote":"Upward adjustment incurs a constant proportional cost, while downward adjustment produces a smaller proportional salvage value.","meaning":"the paper reports no trading or execution result; the cost the extractor found is a model parameter"},"results":{"status":"available","evaluation_type":"none","findings":[{"finding":"The marginal value of capacity solves a family of parabolic double‑obstacle problems; existence, uniqueness and local Sobolev regularity of the strong solution are proved.","kind":"theoretical","quote":"We prove existence, uniqueness and local Sobolev regularity of the strong solution, characterize investment, waiting and disinvestment regions by continuous, strictly separated free boundaries, and verify optimality of the reflected capacity process.","support":"full"},{"finding":"Investment, waiting and disinvestment regions are characterized by continuous, strictly separated free boundaries, and the reflected capacity process is shown to be optimal.","kind":"theoretical","quote":"We prove existence, uniqueness and local Sobolev regularity of the strong solution, characterize investment, waiting and disinvestment regions by continuous, strictly separated free boundaries, and verify optimality of the reflected capacity process. Now, we can decompose the domain DT × (0, ∞) into three parts as follows:        Increasing Region : IR = {(t, x, z) ∈DT × (0, ∞) | Q(t, x, z) = K+}, Decreasing","support":"full"},{"finding":"Numerical experiments show that joint demand improvements shift both free boundaries super‑additively, with the disinvestment boundary moving 1.5–2.7 times as strongly depending on factor correlation.","kind":"theoretical","quote":"Numerically, joint demand improvements shift both boundaries super-additively, 1.5--2.7 times as strongly at the disinvestment boundary, depending on factor correlation.","support":"full"},{"finding":"The state space decomposes into three regions (investment, waiting, disinvestment) as formalized in the free‑boundary analysis (Theorem 4.1).","kind":"theoretical","quote":"5 Free Boundary Analysis In Theorem 4.1, we obtain the unique solution Q to the double obstacle problem (3.27). Now, we can decompose the domain DT × (0, ∞) into three parts as follows:        Increasing Region : IR = {(t, x, z) ∈DT × (0, ∞) | Q(t, x, z) = K+}, Decreasing Region : DR = {(t, x, z) ∈DT × (0, ∞) | Q(t, x, z) = K−}, Waiting Region : WR = {(t, x, z) ∈DT × (0, ∞) | K−< Q(t, x, z) < K+}.","support":"full"},{"finding":"The paper demonstrates that the multi‑factor structure materially affects policy regions, with the investment region narrowing near expiry while the waiting region expands.","kind":"theoretical","quote":"As t →T, the investment region shrinks markedly and the waiting region expands to absorb it: buying a unit near expiry must recover the round-trip friction K+ −K−over a vanishing remaining horizon, whereas retaining one is governed by the user cost ρK−alone. The disinvestment region is largely unchanged.","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":16,"citations":{"cited_by_in_corpus":0,"by_independent_authors":0,"how_cited":{"used_as_baseline":0,"uses_method_or_data":0,"extends":0,"discusses_findings":0,"mentions":0,"unclassified":0},"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":"2026-11-14T19:39:18Z"},"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/"}