Disproportionate costs of uncertainty: Small bank hedging and Dodd‐Frank

Abstract

Uncertainty in banking regulation may impose widespread economic costs by increasing financial constraints on credit availability. Four years of Dodd‐Frank uncertainty over undecided risk weightings increased regulatory uncertainty for smaller banks, restricting “vanilla” interest rate hedging activities. This paper uses newly reported mortgage banking data as an identification strategy and finds that when costs of uncertainty are removed, small banks hedge 97%–120% more interest rate risk while mortgage securitization income increases by 65.2% compared to large banks. These findings support the need for tailored regulations that consider the higher costs of regulatory uncertainty for smaller banks.

Hedging securities and Silicon Valley Bank idiosyncrasies

“…Using this approach, I obtain similar overall findings as Table 4, confirming this paper's findings are not driven by sample selection bias. The first stage in Equation (2) is a Probit model where the independent variables are motivated by Bliss et al (2018), Kim (2021), Purnanandam (2007), and Sinkey and Carter (2000) specified as $$$\begin{array}{ccc}Pr({{Hedger}}_{i,t}=1) & = & \phi \left[\alpha +{\gamma }_{1}log(Assets)+{\gamma }_{2}\frac{Loan{s}_{i,t}}{Asset{s}_{i,t}}+{\gamma }_{3}\frac{NP{A}_{i,t}}{Asset{s}_{i,t}}+{\gamma }_{4}\frac{Deposit{s}_{i,t}}{Asset{s}_{i,t}}\right.\\ & & \left.+{\gamma }_{4}\frac{{[{HTM}{\unicode{x02007}}{Losses},\unicode{x02007}{AFS}{\unicode{x02007}}{Losses},\unicode{x02007}{Maturity}{\unicode{x02007}}{Gap}]}_{i,t}}{Asset{s}_{i,t}}+{\varepsilon }_{i,t}\right].\end{array}$$$…”

“…This raises concerns that the panel data in Table 4 is subject to selection bias when hedging IRD is the dependent variable. Similar to Bliss et al (2018), Kim (2021), and Purnanandam (2007), I deploy a Heckman (1979) two‐stage model that uses maximum likelihood estimation to correct for any sample selection bias. Using this approach, I obtain similar overall findings as Table 4, confirming this paper's findings are not driven by sample selection bias.…”

“…Hedging should be evaluated separately from trading to understand the bank's actual risk exposure and management proficiency. Similar to Kim (2021), Table 3 explores the impact of 10-year swap rates on interest rate derivatives. This table also explores the impact of 10-year swap rates on trading versus hedging interest rate derivatives.…”

“…Jiang et al (2023a) argue banks were under‐hedged since 6% only of bank assets were hedged while McPhail et al (2023) find negligible net swap exposure used to hedge balance‐sheet risks. However, McPhail et al (2023) net both hedging and trading interest rate derivatives which is problematic since trading primarily speculates on macroeconomic factors while hedging mitigates bank balance‐sheet risks (Begenau et al, 2015; Chernenko & Faulkender, 2011; Kim, 2021; Ruprecht et al, 2013). This paper uses Call Reports Schedule RC‐L (Derivative and Off‐Balance Sheet Items) data which requires banks to report interest rate derivatives used for hedging (RCON8725) and those used for trading (RCONA126) separately.…”

“…InKim (2021), hedging IRDs are used to address mortgage rate risk during the origination process.9 Of 6539 unique banks in the sample, 1884 have used interest rate derivatives for hedging at least once during the sample period of 2015Q1-2022Q4.…”