{"paper_id":"2508.02691","title":"Statistical modeling of SOFR term structure","authors":["Teemu Pennanen","Waleed Taoum"],"paper_date":"2025-07-23","paper_type":"empirical","paper_url":"https://arxiv.org/pdf/2508.02691v2","page_url":"https://thequant.space/flowcharts/statistical-modeling-of-sofr-term-structure/","scores":{"empirical_rigor":4,"math_complexity":6.5,"quadrant":"Lab Rats"},"proposes":{"status":"available","statement":"The paper proposes a statistical SOFR term structure model for risk management and derivatives pricing in incomplete markets.","quote":"This paper develops a statistical SOFR term structure model that is well-suited for risk management and derivatives pricing within the incomplete markets paradigm.","summary":"SOFR derivatives market remains illiquid and incomplete so it is not amenable to classical risk-neutral term structure models which are based on the assumption of perfect liquidity and completeness. This paper develops a statistical SOFR term structure model that is well-suited for risk management…","strategy":null},"assumptions":{"status":"available","items":[{"assumption":"SOFR derivatives market is illiquid and incomplete, so classical risk‑neutral arbitrage does not hold.","type":"market","stated":"explicit","quote":"SOFR derivatives market remains illiquid and incomplete so it is not amenable to classical risk-neutral term structure models which are based on the assumption of perfect liquidity and completeness."},{"assumption":"Futures contracts are quoted with bid‑ask spreads and are not perfectly liquid.","type":"market","stated":"explicit","quote":"The exchange-traded three-month futures rates come with bid-ask spreads."},{"assumption":"SOFR jumps whenever the Fed changes its policy rate at FOMC meetings.","type":"model","stated":"explicit","quote":"This causes jumps in SOFR whenever the FED changes its policy rates at the Federal Open Market Committee (FOMC) meeting dates."},{"assumption":"Macro‑economic factors (policy rate L, inflation I, GDP G) follow a linear vector autoregressive process.","type":"model","stated":"explicit","quote":"We model the vector y := [ln(L + cL), I, G] by the linear stochastic difference equation (vector autoregressive model) ∆yt = Ayt−1 + at + ϵt."},{"assumption":"Zero‑coupon bond prices are approximated via the overnight forward curve despite bid‑ask spreads.","type":"model","stated":"implied","quote":"If the ZCB market was perfectly liquid without bid-ask spreads, this would give a one-to-one correspondence between the zero curve …"},{"assumption":"The forward curve can be represented by piecewise constant basis functions fitted to observed futures quotes.","type":"model","stated":"explicit","quote":"to the observed futures quotes as well as to the value of SOFR observed the morning after."}]},"datasets":{"status":"available","items":[{"name":"SOFR futures bid‑ask quotes","provider":"Bloomberg Finance L.P.","kind":"options","access":"commercial","period":"1 April 2020 onward","universe":"3‑month, 1‑month SOFR futures contracts","quote":"These tenors are available from Bloomberg Finance L.P. from 1 April 2020 onwards.","known_id":"bloomberg"},{"name":"Daily SOFR rates","provider":"Federal Reserve","kind":"market-prices","access":"public","period":"historical (e.g., 2023‑2024 shown)","universe":"overnight secured financing rate","quote":"Figure 3: Daily values of SOFR, ON RRP, and the limits of federal funds target range [L, H] (source: FRED)."},{"name":"FED monetary policy measures (target range, ON RRP, SRF)","provider":"Federal Reserve","kind":"macro","access":"public","period":"as of 4 January 2024","universe":null,"quote":"Table 1: A glimpse of FED monetary policy measures as of 4 January 2024. SOFR was 5.32% (source: FED)."},{"name":"Macro‑economic risk factors (L, I, G)","provider":"FRED and Bloomberg Finance L.P.","kind":"macro","access":"public","period":"January 1972 to January 2024","universe":null,"quote":"We use monthly observations of y from January 1972 to January 2024.","known_id":"bloomberg"}],"known":[{"id":"bloomberg","name":"Bloomberg","provider":"Bloomberg L.P.","kind":"market-prices","access":"commercial","quote":"The futures rate quotes are computed using Remark 1 and are plotted in the middle of the reference quarter (data source: Bloomberg Finance L.P.)","mentions":6,"used_in":3}],"sample":{"start":null,"end":null,"market":null,"asset_class":null,"frequency":null},"data_resolution":"daily"},"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":"yes","evidence_card_quoted":true},"conflict_kind":"applicability","card_detail":"Bid‑ask spreads of SOFR futures are incorporated as lower and upper bounds in the convex optimization calibration.","card_quote":"The components of the vector e quantify the violation of the lower and upper bounds in (13) for each of the quoted forward contracts.","meaning":"the platform classifier and the paper-text extraction disagree, so no value is asserted"},"results":{"status":"available","evaluation_type":"unclear","findings":[{"finding":"Calibration of the convex optimization problem is extremely fast, taking less than 0.4 seconds on a standard laptop.","kind":"performance","quote":"The solution is obtained in less than 0.4 seconds using a Dell Latitude running Linux with 16 GB of RAM and a 4-core Intel i7 CPU at 3.00 GHz; see Example 4."},{"finding":"Solving 1300 instances of the calibration problem takes under 8 minutes, demonstrating scalability.","kind":"performance","quote":"Similarly, solving 1300 instances of problem (11) takes less than 8 minutes; see Section 4.2."},{"finding":"The model reproduces the discrete jumps in the forward curve observed on FOMC meeting dates.","kind":"theoretical","quote":"The model operates on a daily frequency, and it is consistent with the discrete jumps often observed on FMOC meeting dates."},{"finding":"The first component of the forward curve (ξ1) closely tracks SOFR and shows small variations except when the Fed target range changes.","kind":"statistical","quote":"ξ1 tends to have only small variations between consecutive FOMC meeting dates but it jumps whenever the target is changed."}],"reported_performance":null,"out_of_sample":{"method":"none","headline_out_of_sample":null}},"independent_evidence":{"status":"none_found","supports":[],"contradicts":[],"qualifies":[],"compared_with":24,"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":"2026-11-14T17:31:37Z"},"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/"}