{"paper_id":"2504.06028","title":"A Mean-Reverting Model of Exchange Rate Risk Premium Using Ornstein-Uhlenbeck Dynamics","authors":["SeungJae Hwang"],"paper_date":"2025-04-08","paper_type":"empirical","paper_url":"https://arxiv.org/pdf/2504.06028v1","page_url":"https://thequant.space/flowcharts/a-mean-reverting-model-of-exchange-rate-risk-premium-using-o/","scores":{"empirical_rigor":8,"math_complexity":7.5,"quadrant":"Holy Grail"},"proposes":{"status":"available","statement":"The paper proposes a mean‑reverting Ornstein‑Uhlenbeck model for the exchange‑rate risk premium and embeds it in a continuous‑time stochastic exchange‑rate equation.","quote":"a mean-reverting model of the exchange rate risk premium, motivated by persistent empirical deviations from the uncovered interest parity (UIP) condition.","summary":"This paper examines the empirical failure of uncovered interest parity (UIP) and proposes a structural explanation based on a mean-reverting risk premium. We define a realized premium as the deviation between observed exchange rate returns and the interest rate differential, and demonstrate its…","strategy":null},"assumptions":{"status":"available","items":[{"assumption":"The risk premium follows a single‑regime Ornstein‑Uhlenbeck process.","type":"model","stated":"explicit","quote":"Modeling the premium as an Ornstein‑Uhlenbeck (OU) process embedded within a stochastic differential equation for the exchange rate"},{"assumption":"The risk premium is assumed to vanish in the long run (initial hypothesis).","type":"model","stated":"explicit","quote":"Starting from the hypothesis that the premium should vanish in the long run—allowing exchange rates to eventually revert to parity"},{"assumption":"Exchange‑rate and interest‑rate data are daily and can be split into a training (80%) and validation (20%) set for out‑of‑sample backtesting.","type":"data","stated":"explicit","quote":"The dataset is divided into a training set (80%) and a validation set (20%), with the latter used for out-of-sample evaluation."},{"assumption":"Log‑exchange rates are approximately normally distributed, allowing log‑normal approximation for the spot rate.","type":"statistical","stated":"explicit","quote":"Since log St is approximately normal, we can approximate the probability density function (PDF) of St using the log‑normal distribution."},{"assumption":"Forecast confidence intervals are constructed using the approximate log‑normal distribution of future spot rates.","type":"model","stated":"explicit","quote":"Given the approximate log‑normality of St+h, confidence intervals for future spot rates are constructed as:"},{"assumption":"The single‑regime OU model cannot capture transitional dynamics around the 3‑month horizon.","type":"model","stated":"explicit","quote":"This highlights a potential limitation of the single‑regime OU model in capturing transitional dynamics."}]},"datasets":{"status":"available","items":[{"name":"USD/KRW exchange rate and interest rates","provider":null,"kind":"market-prices","access":"unknown","period":"January 2010 to April 2025","universe":null,"quote":"We use daily data on the USD/KRW exchange rate and interest rates from January 2010 to April 2025."}],"known":[],"sample":{"start":null,"end":null,"market":null,"asset_class":null,"frequency":null},"data_resolution":"daily"},"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":"forecast-accuracy","findings":[{"finding":"The model shows strong empirical coverage at short‑term horizons (2 weeks and 1 month), with coverage rates slightly exceeding nominal confidence levels.","kind":"performance","quote":"Coverage rates slightly exceed nominal levels, indicating a conservative but reliable forecast."},{"finding":"Coverage at the 6‑month horizon aligns closely with target confidence levels, indicating the mean‑reverting structure remains informative at medium‑to‑long horizons.","kind":"performance","quote":"The model also performs well at the 6‑month horizon, with coverage rates closely aligned with target confidence levels."},{"finding":"The 3‑month horizon consistently underperforms, with empirical coverage below nominal levels, suggesting a transitional volatility zone not captured by the single‑regime OU model.","kind":"negative","quote":"The 3‑month horizon consistently underperforms, with coverage falling below nominal levels across all intervals."},{"finding":"At the 1‑year horizon the model produces overly conservative tail forecasts, yielding near‑perfect coverage (≈99.98%–100%) that likely reflects over‑estimated long‑run uncertainty.","kind":"performance","quote":"near‑perfect coverage, reaching 99.98% and 100%, respectively. While this might seem desirable, it more likely reflects an overestimation of long‑term uncertainty due to the OU process’s asymptotically stabilizing variance."},{"finding":"Overall, the mean‑reverting risk‑premium model captures key exchange‑rate dynamics, confirming that deviations from UIP are structured and forecastable rather than pure noise.","kind":"theoretical","quote":"Our work contributes a tractable framework that connects short‑run fluctuations with long‑run equilibrium through a dynamic, mean‑reverting premium."}],"reported_performance":null,"out_of_sample":{"method":"holdout","headline_out_of_sample":null}},"independent_evidence":{"status":"none_found","supports":[],"contradicts":[],"qualifies":[],"compared_with":36,"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":1,"extracted_at":"2026-11-14T17:41:21Z"},"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/"}