Paper: arXiv 2412.15986

Authors: Michele Leonardo Bianchi, Dario Ruzzi, Anatoli Segura

Abstract

We use granular regulatory data on euro interest rate swap trades between January 2021 and June 2023 to assess whether derivative positions of Italian banks can offset losses on their debt securities holdings should interest rates rise unexpectedly. At the aggregate level of the banking system, we find that a 100-basis-point upward shift of the yield curve increases on average the value of swaps by 3.65% of Common Equity Tier 1 (CET1), compensating in part for the losses of 2.64% and 5.98% of CET1 recorded on debt securities valued at fair value and amortised cost. Variation exists across institutions, with some bank swap positions playing an offsetting role and some exacerbating bond market exposures to interest rate risk. Nevertheless, we conclude that, on aggregate, Italian banks use swaps as hedging instruments to reduce their interest rate exposures, which improves their ability to cope with the recent tightening of monetary policy. Finally, we draw on our swap pricing model to conduct an extensive data quality analysis of the transaction-level information available to authorities, and we show that the errors in fitting value changes over time are significantly lower compared to those in fitting the values themselves.

Complexity vs Empirical Score

  • Math Complexity: 5.0/10
  • Empirical Rigor: 8.0/10
  • Quadrant: Holy Grail — high math complexity, high empirical rigor

Why this score: The paper employs advanced financial mathematics for full revaluation of swap contracts and uses statistical models for data quality assessment, placing it in the medium-high complexity range. Empirical rigor is high due to the use of granular regulatory data (EMIR), extensive backtesting over a multi-year period, and rigorous data quality checks.

Research Flowchart

  flowchart TD
  A["Research Goal: Assess if Italian banks'<br>derivative positions offset debt security<br>losses during rate hikes"] --> B["Data & Methodology<br>Granular regulatory data<br>Jan 2021 - Jun 2023"]
  
  B --> C["Computational Process<br>100bps upward yield curve shift<br>Simulation applied to:<br>- Interest Rate Swaps<br>- Debt Securities (Fair Value & Amortised Cost)"]
  
  C --> D["Analysis & Pricing<br>Swap valuation modeling<br>Data quality analysis<br>Value change fitting"]
  
  D --> E["Key Findings / Outcomes"]
  
  E --> E1["Aggregate: Swaps offset bond losses<br>Swap gain: +3.65% CET1<br>Bond loss: -8.62% CET1 (combined)"]
  E --> E2["Institutional Variation<br>Swaps hedge for some,<br>exacerbate risk for others"]
  E --> E3["Data Quality<br>Value-change fitting errors<br>significantly lower than<br>value-fitting errors"]
  E --> E4["Conclusion<br>Swaps reduce interest rate exposure<br>Improves resilience to monetary tightening"]

Evidence Card

What the authors claim, on what data, and who has checked it. Read from the paper's own text by a language model: every quote is verbatim, nothing was re-run. How to read this card

Proposes

The paper proposes a swap pricing model to assess hedging effectiveness of Italian banks and conducts an extensive data quality analysis of EMIR transaction data.

Source sentence
Finally, we draw on our swap pricing model to conduct an extensive data quality analysis of the transaction-level information available to authorities, and we show that the errors in fitting value changes over time are significantly lower compared to those in fitting the values themselves.

Assumes 5

  • Portfolio allocations are held constant until the next month‑end date. data
    Source sentence
    we rely on the assumption that portfolio allocations are held constant until the next month-end date.
  • A parallel upward shift of 100 basis points in the yield curve is used as the stress scenario. market
    Source sentence
    a 100‑basis‑point upward shift of the yield curve
  • Discount rates for swaps equal risk‑free rates, while bond discount rates include an extra credit spread component. model
    Source sentence
    In the case of derivatives, including swaps, discount rates and risk‑free rates are the same, whereas the discount rates of bonds have an extra component (spread) that reflects the credit risk of the issuer.
  • Swap contract rates are cleaned by replacing outliers (below –2.5% or above 25%) with market par swap rates. data
    Source sentence
    when is an outlier (below -2.5% and above 25%) that needs replacing with the market par swap rate.
Show 1 more
  • Valuations dated more than one week prior to the reporting date are discarded. data
    Source sentence
    discarding trades whose valuation is dated more than one week prior to the reporting date

Data

  • Euro interest rate swap trades (EMIR) · European Market Infrastructure Regulation (EMIR) trade repos market-prices proprietary
    January 2021 to June 2023 · 218,717 unique swap contracts traded by 54 banks
    Source sentence
    our sample includes 218,717 unique swap contracts traded by 54 banks, for a total of 14,664,747 swap‑week observations covering the sample period from January 2021 to June 2023
  • Bond holdings data · Statistical and supervisory reports of the Bank of Italy fundamentals proprietary
    December 2020 to June 2023 · Italian banks’ fixed‑rate bond portfolios
    Source sentence
    The dataset on which our empirical study is based includes end‑of‑month portfolio allocations from December 2020 to June 2023

Costs

Not applicable The paper reports no trading or execution result.

Finds 3

  • A 100‑bp upward shift of the yield curve raises the average value of swaps by 3.65% of CET1, partially offsetting bond losses of 2.64% (fair value) and 5.98% (amortised cost) of CET1. performance
    Source sentence
    a 100‑basis‑point upward shift of the yield curve increases on average the value of swaps by 3.65% of Common Equity Tier 1 (CET1), compensating in part for the losses of 2.64% and 5.98% of CET1 recorded on debt securities valued at fair value and amortised cost
  • Aggregate gains on swaps fully offset the losses on fair‑value bonds, indicating effective hedging, but do not fully offset losses on amortised‑cost bonds. statistical
    Source sentence
    Another interesting finding from Figure 8 is that the aggregate gains on swaps fully offset the losses on bonds carried at fair value, indicating that banks may use swaps to hedge the value of these assets against changes in rates. However, this does not seem to hold true for losses on AC bonds
  • Errors in fitting swap value changes over time are significantly lower than errors in fitting the absolute values themselves. pricing-calibration
    Source sentence
    we show that the errors in fitting value changes over time are significantly lower compared to those in fitting the values themselves

Checked by others

No other paper in the archive has been found to support or contradict these findings (compared with 21 related papers). Coverage is partial, so this is neither replication nor refutation.

Code

No repository by the authors found in the paper.

Extracted 2026-11-14 · same card as JSON · A language model chose these items: check the source sentence before relying on one.