{"paper_id":"2412.14144","title":"Application of the Kelly Criterion to Prediction Markets","authors":["Bernhard K Meister"],"paper_date":"2024-12-18","paper_type":"methodological","paper_url":"https://arxiv.org/pdf/2412.14144v1","page_url":"https://thequant.space/flowcharts/application-of-the-kelly-criterion-to-prediction-markets/","scores":{"empirical_rigor":2,"math_complexity":7,"quadrant":"Lab Rats"},"proposes":{"status":"available","statement":"The paper proposes a modified payout structure and investigates a simple asset price model based on flipping biased coins.","quote":"A modified payout structure is proposed. A simple asset price model based on flipping biased coins is investigated.","summary":"Betting markets are gaining in popularity. Mean beliefs generally differ from prices in prediction markets. Logarithmic utility is employed to study the risk and return adjustments to prices. Some consequences are described. A modified payout structure is proposed. A simple asset price model based…","strategy":null},"assumptions":{"status":"available","items":[{"assumption":"All investors are logarithmic utility maximizers who view the prediction market in isolation.","type":"market","stated":"explicit","quote":"Here we assume that all investors are logarithmic utility maximizers, who view the particular prediction market in isolation from other activities, and the contribution of the ith investor to the expectation value, i.e. mean belief, is their subjective probability qi times their capital Ci.","support":"full"},{"assumption":"The market is treated as a black box that ‘works’, ignoring mirror markets, fees and other costs.","type":"execution","stated":"explicit","quote":"In the following analysis, the existence of mirror markets as well as the existence of fees and other costs is ignored. Instead we regard the market as a black box that ‘works’.","support":"full"},{"assumption":"There is only an initial margin and no variation margin, reducing capital efficiency.","type":"market","stated":"explicit","quote":"There is therefore only an initial but no variation margin.","support":"full"},{"assumption":"Borrowing is prohibited in the optimal Kelly fraction calculation for an ‘all-or-nothing’ contract.","type":"model","stated":"explicit","quote":"The optimal Kelly fraction for an ‘all-or-nothing’ contract with a prohibition on borrowing follows from ﬁnding the extremum of the logarithmic utility function U: U(q, p, f) = (1 −q) log(1 −f) + q log \u0010 1 + f 1 −p p \u0011 ∂U(q, p, f) ∂f = q 1−p p 1 + f 1−p p −1 −q 1 −f = 0 ⇒f = q −p1 −q 1 −p = Q −P 1 + Q . with P = p 1−p and Q = q 1−q.","support":"full"},{"assumption":"Investors cannot hold naked short positions, leading to trading of two linked contracts.","type":"market","stated":"explicit","quote":"The inability to have naked short positions leads in binary outcome markets to the trading of two linked contracts.","support":"full"},{"assumption":"The analysis assumes a simplified market of just two investors for bounding the price‑belief gap.","type":"model","stated":"explicit","quote":"For the market to clear the sum P i Cifi(p, qi) has to be zero at the market price p. This leads to a simple 3 scaling relations, for example increasing the capital of all participants by the same factor will not aﬀect p. To get a feeling for possible bounds we assume a simpliﬁed market of just two investors. One being certain of the outcome and the other having the belief q.","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":"available","modeled":"unclear","headline_net_of_costs":null,"detail":null,"quote":null,"source":"conflict","rule":"R4b","candidates":{"front_matter":"n/a","evidence_card":"no","evidence_card_quoted":true},"conflict_kind":"applicability","card_detail":null,"card_quote":"In the following analysis, the existence of mirror markets as well as the existence of fees and other costs is ignored.","meaning":"the platform classifier and the paper-text extraction disagree, so no value is asserted"},"results":{"status":"available","evaluation_type":"none","findings":[{"finding":"The gap between observed price and mean belief can be wide, ranging from 0 to 1/2 for p<1/2 and from 1/2 to 1 for p>1/2 in the two‑investor toy example.","kind":"theoretical","quote":"(2) The diﬀerent possible choices of q− 1 give in this case for E−a range of outcomes, which are [0, 1 2] for p < 1/2, [1 2, 1] for p > 1/2, and 1/2 for p = 1/2.","support":"full"},{"finding":"Introducing a parameter α in the payout structure smooths and reduces the price‑belief gap near the boundaries for 0≤α≤1, and reverses behavior for α≥1.","kind":"theoretical","quote":"Instead of trying to ﬁnd an optimal model, we just show how a simple change in the payout can change the utility and the optimal fraction: Uα(q, p, f) = (1 −q) log(1 −f) + q log \u0010 1 + f \u0010 p 1 −p \u0011α\u0011 ∂Uα(q, p, f) ∂f = \u0010 p 1−p \u0011α 1 + f \u0010 p 1−p \u0011α −1 −q 1 −f = 0 ⇒f = ˆPα −Q Q + 1 4 with ˆPα = \u0010 p 1−p \u0011α , where for α = 1 one gets the conventional result, and for 0 ≤α ≤1 one gets a smoothing and reduction of the gap","support":"full"},{"finding":"Using Kullback‑Leibler divergence, the paper shows how misjudgment of bias and miscalculation of the investment fraction affect portfolio growth rate.","kind":"theoretical","quote":"It is shown using the Kullback-Leibler divergence how the misjudgment of the bias and the miscalculation of the investment fraction influence the portfolio growth rate.","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":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":"2026-11-14T19:53:20Z"},"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/"}