Working Papers
[New!] The Market for Sharing Interest Rate Risk: Quantitites behind Prices (July 2023) – with Umang Khetan (University of Iowa), Jane Li (Columbia Business School) and Ishita Sen (Harvard Business School) – first version: BoE Staff WP 1031
We study the extent of interest rate risk sharing across the financial system. We use granular positions and transactions data in interest rate swaps, covering over 60% of overall swap activity in the world. We show that pension and insurance (PF&I) sector emerges as a natural counterparty to banks and corporations: overall, and in response to decline in rates, PF&I buy duration, whereas banks and corporations sell duration. This cross-sector netting reduces the aggregate net demand that is supplied by dealers. However, two factors impede cross-sector netting and add to dealer imbalances across maturities. (i) PF&I, bank and corporate demand is segmented across maturities. (ii) Large volumes are traded by hedge funds, who behave like banks in the short-end and like PF&I in the long-end. This worsens segmentation, exposing dealers to a steepening or flattening of the yield curve in addition to residual duration risk. Consistent with this, we find that demand pressure, in particular hedge funds’ trades, impact swap spreads across maturities. We also document that long-tenor pension fund trades are less likely to be centrally cleared, adding counterparty credit risk to demand imbalances.
(Upcoming) seminars and presentations: University of Iowa*, Commodity Futures Trading Comission, Bank of Canada, HEC Montréal, AFA Poster session*
The Ring-Fencing Bonus (Oct 2022)– with Irem Erten (Warwick Business School) and John Thanassoulis (Warwick Business School) – BoE Staff WP 999
We study the impact of ring-fencing on bank risk using short-term repo rates. Exploiting confidential data on the near-universe of sterling-denominated repo transactions, we find compelling evidence that banking groups subject to ring-fencing are perceived to be safer; repo investors lend to ring-fenced groups at lower rates, controlling for bank characteristics and collateral risk. Ring-fenced groups charge more to supply liquidity. We show that these effects are driven by the ring-fenced subsidiary; the other subsidiaries are not adversely impacted by ring-fencing to any meaningful extent. We further document that the banking groups reduce their risk-taking after the imposition of the fence. Our paper suggests that structural reforms can have a significant beneficial impact on risk in the banking system.
Presented at (selected): FMA Applied Finance, FMA Annual Conference*, OCC Symposium on Emerging Risks in the Banking System*, ERMAS, EFMA*, IFABS*, NFA Annual Meeting
Media coverage: Faculti.net , Financial Times: Martin Wolf , Financial Times: Helen Thomas, The Banker
In public policy: The Edinburgh Reforms, His Majesty’s Treasury Call for Evidence
Banks’ internal capital allocation, the leverage ratio requirement and risk-taking (2021) – with Quynh-Anh Vo (BoE) – BoE Staff WP 956
This paper examines how banks’ asset risk is affected by the level (ie group or business unit) at which regulatory requirements are applied. We develop a theoretical model and calibrate it to UK banks. Our main finding is that the impact differs depending on which regulatory constraint is binding at the group level. If that is the leverage ratio requirement, then the allocation of regulatory constraints to business units either maintains or decreases the riskiness of banks’ investment portfolios. However, if the risk-weighted requirement is the binding constraint at the group level, applying regulatory requirements at the business unit level can lead to banks selecting riskier asset portfolios as optimal. We also find that the impact on banks’ asset risk differs across bank business models.
Presented at (selected): EFMA 2021, IFABS 2021, 37th Symposium on Money Banking and Finance BdF 2021*, WinE EEA 2021, 20th Annual FDCI/JFSR Bank Research Conference 2021*, ASSA 2022, ERMAS Romania 2022
Work in progress
[Out soon!] Lash risk and interest rates (2023) – with Laura Alfaro (Harvard Business School), Saleem Bahaj (UCL & BoE), Robert Czech (Bank of England) and Jonathon Hazell (LSE)
In this paper, we first introduce a framework to understand and quantify a previously less-examined form of liquidity risk, Liquidity After Solvency Hedging risk or “Lash risk”. Unlike conventional forms of liquidity risk, Lash risk is not linked to maturity transformation and callable claims, rollover risk, or insolvency. Using regulatory data on the universe of sterling repo and swap exposures, we document a rapid increase of liquidity risks from leveraged exposures in non-bank financial institutions (NBFIs). We construct a measure of aggregated Lash risk across both instruments and link the evolution of this measure to recent monetary policy decisions. We show that low-interest rates have “sowed the seeds” of future liquidity crises, as NBFIs sought to hedge their long-dated liabilities and invest in higher-return assets at the same time. We then link the pre-crisis liquidity risk exposures of NBFIs to their UK government bond (gilt) trading activity during the UK mini-budget crisis and find that funds with larger exposures sold substantially higher quantities of gilts during the crisis, and thereby significantly contributed to the yield spike in the gilt market.
Details available upon request.
Dormant projects
Risk-Taking, Competition and Uncertainty. Do CoCo Bonds Increase the Risk Appetite of Banks? – BoE /CEPR WP Aug 2021 – with Mahmoud Fatouh (Bank of England) and Sweder van Wijnbergen (UvA)
Presented at (selected): Royal Economic Society Symposium 2021, IFABS 2021, Money Macro and Finance Society Annual Conference 2021, 19th Winter School of Mathematical Finance 2022
Multiple buffer CoCos and their impact on financial stability – Tinbergen Institute WP Jan 2020
Presented at: GSE Barcelona 2018, IFABS 2018, FMA European Conference 2018, INFER 2019, Queen Mary Economics and Finance PhD workshop 2019, ERMAS Romania 2019
(Updated: October 2023. * presentations by co-authors.)