{"paper_id":"2401.07183","title":"Optimal Investment with Herd Behaviour Using Rational Decision Decomposition","authors":["Huisheng Wang","H. Vicky Zhao"],"paper_date":"2024-01-14","paper_type":"empirical","paper_url":"https://arxiv.org/pdf/2401.07183v2","page_url":"https://thequant.space/flowcharts/optimal-investment-with-herd-behaviour-using-rational-decisi/","scores":{"empirical_rigor":5,"math_complexity":8.5,"quadrant":"Holy Grail"},"proposes":{"status":"available","statement":"The paper proposes a dual‑agent optimal investment model that incorporates herd behaviour via an average deviation term in the Merton objective, introduces a rational decision decomposition where the follower’s optimal decision is a convex combination of rational decisions, and defines an investment","quote":"In this paper, we study the optimal investment problem considering the herd behaviour between two agents, including one leading expert and one following agent whose decisions are influenced by those of the leading expert. To theoretically analyze the impact of the following agent's herd behaviour on his/her decision, we decompose his/her optimal decision into a convex linear combination of the two agents' rational","summary":"In this paper, we study the optimal investment problem considering the herd behaviour between two agents, including one leading expert and one following agent whose decisions are influenced by those of the leading expert. In the objective functional of the optimal investment problem, we introduce…","strategy":null},"assumptions":{"status":"available","items":[{"assumption":"Agent A2 (the leading expert) follows a rational decision path \\(\\{\\bar{P}_2(t)\\}_{t\\in T}\\) as given in equation (5).","type":"model","stated":"explicit","quote":"We assume that A2’s decision follows the form of the rational decision { ¯P2(t)}t∈T as in (5), and A1’s decision {P1(t)}t∈T is unilaterally influenced by A2’s decision.","support":"full"},{"assumption":"Agent A1 (the following agent) makes decisions that are unilaterally influenced by A2’s decision.","type":"model","stated":"explicit","quote":"We assume that A2’s decision follows the form of the rational decision { ¯P2(t)}t∈T as in (5), and A1’s decision {P1(t)}t∈T is unilaterally influenced by A2’s decision.","support":"full"},{"assumption":"A1 maximises expected utility of terminal wealth while minimising the distance to A2’s decision, with a trade‑off parameter (herd coefficient) \\(\\theta>0\\).","type":"model","stated":"explicit","quote":"Considering the herd behaviour, A1 aims to max-imize his/her expected utility of the terminal wealth while minimizing the distance between his/her decision and A2’s decision. With Definition 1, we define the following optimal invest-ment problem for A1’s decision: Problem 2. sup {P1(t)}t∈T ∈U Eϕ(X1(T)) −θD[P1∥¯P2] s.t. dX1(t) = [rX1(t) + vP1(t)]dt + σP1(t)dW(t), X1(0) = x1, where the herd coefficient θ > 0 is to","support":"full"},{"assumption":"The modified herd coefficient is defined as \\(\\vartheta:=\\theta\\alpha_1\\sigma^2\\) and the decay rate as \\(\\rho\\).","type":"model","stated":"explicit","quote":"For the convenience of expression, we define ϑ := θ α1σ2 and ϱ := 2 −ρ as the modified herd coefficient and decay rate, respectively. Theorem 1.","support":"full"},{"assumption":"Risk‑aversion coefficients are fixed at \\(\\alpha_1=0.2\\) for the follower and \\(\\alpha_2=0.4\\) for the leader.","type":"model","stated":"explicit","quote":"From the prior work in [31], we set α1 = 0.2 and α2 = 0.4 as the risk aversion coefficients for the following agent and leading expert, re-spectively.","support":"full"},{"assumption":"The risky asset price follows a geometric Brownian motion with estimated drift \\(\\mu=0.07\\) and volatility \\(\\sigma=0.17\\).","type":"market","stated":"explicit","quote":"Utilizing parameter estimation of the geometric Brow-nian motion, we determine the expected return µ to be 0.07 and volatility σ to be 0.17 for the risky asset.","support":"full"}]},"datasets":{"status":"available","items":[{"name":"Dow Jones Industrial Average daily closing prices","provider":"Dow Jones","kind":"market-prices","access":"public","period":"1972-01 to 2022-12","universe":"Dow Jones Industrial Average index","quote":"During this in-vestment period, we collect daily closing prices of the Dow Jones Industrial Average to represent the prices of the risky asset.","support":"partial"},{"name":"U.S. 1‑Year Treasury Bill daily average rates (2022‑2023)","provider":"U.S. Treasury","kind":"macro","access":"public","period":"2022‑2023","universe":null,"quote":"We use the daily average of the interest rates of U.S. 1-Year Treasury Bills in 2022 and 2023 to represent the risk-free interest rate r, which is approximately equal to 0.04.","support":"full"}],"known":[],"sample":{"start":null,"end":null,"market":null,"asset_class":null,"frequency":null},"data_resolution":"unclear"},"transaction_costs":{"status":"not_applicable","modeled":"n/a","headline_net_of_costs":null,"detail":null,"quote":null,"source":"evidence_card","rule":"R2","candidates":{"front_matter":"unclear","evidence_card":"n/a","evidence_card_quoted":false},"meaning":"the paper reports no trading or execution result"},"results":{"status":"available","evaluation_type":"unclear","findings":[{"finding":"The analytical optimal decision for the following agent is obtained via the variational method, but the resulting expression for the integral constant \\(\\eta\\) has no closed‑form solution and requires a numerical algorithm.","kind":"theoretical","quote":"(17) However, (17) has no closed-form solution. Thus, we pro-pose Algorithm 1 to calculate its numerical solution.","support":"full"},{"finding":"The investment opinion \\(Z_1(t)\\) is a decreasing function of the herd coefficient \\(\\theta\\) (i.e., \\(\\partial Z_1(t)/\\partial\\theta<0\\)).","kind":"theoretical","quote":"The investment opinion Z1(t) is a decreasing function of θ with ∂Z1(t) ∂θ < 0. Proof.","support":"full"},{"finding":"Numerical experiments show that the follower’s optimal decision stays within the region bounded by the two agents’ rational decisions and moves closer to the leader’s rational decision as the herd coefficient increases.","kind":"theoretical","quote":"It can be seen that A1’s optimal decision consistently re-mains within the region bounded by the rational decisions of the two agents { ¯P1(t)}t∈T and { ¯P2(t)}t∈T . Furthermore, the optimal decision progressively approaches A2’s rational decision { ¯P2(t)}t∈T as the herd coefficient increases.","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":9,"citations":{"cited_by_in_corpus":4,"by_independent_authors":0,"how_cited":{"used_as_baseline":0,"uses_method_or_data":0,"extends":0,"discusses_findings":0,"mentions":0,"unclassified":4},"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":"2026-11-14T19:59:39Z"},"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/"}