Academic journal article Journal of Economics and Finance

A Longitudinal Study of Net Interest Margin by Bank Asset Size: 1992-2005

Academic journal article Journal of Economics and Finance

A Longitudinal Study of Net Interest Margin by Bank Asset Size: 1992-2005

Article excerpt

Abstract

With the consolidation in banking over the past 20 years, interest in the comparative performance of big and small banks intensified. This study expands this research and examines the profitability of intermediation (measured by net interest margin or NIM) through a longitudinal model that uses panel data. Banks are assigned to one of five asset classes for each year of the 1992-2005 period, and the classes serve as the panels. Results show that interest rate effects on NIM vary by asset class, but the presence of economic effects and fixed effects on NIM depends on the model's specifications.

(JEL G21)

Introduction

The Interstate Banking and Branch Efficiency Act (1994) permitted interstate branching by July 1997 and accelerated the consolidation in banking that has been underway since 1982. Mammoth banks appeared while the number of small banks shrank. This heightened interest in the comparative performance of big and small banks. To expand this research, this study examines the comparative profitability of intermediation at all banks, controlling for asset size over the 19922005 period. This study examines the effect of size on the profitability of intermediation over the 1992-2005 period using net interest margin (net interest income to average assets), or NIM, as the relevant profit measure for intermediation.

The study first presents a discussion on net interest margins that lays out the basic assumptions behind the forces affecting NIM. Recent research on NIM's behavior is summarized and contrasted with the current study. Data and data methodologies are then addressed. To focus on the effect of asset size on profitability, all banks are assigned to one of five asset classes for each year of the 1992-2005 sample period, average NIM is calculated for each asset class for each year, and differences in NIM by asset class over the sample period are identified.

Next, the longitudinal model is developed in which the asset classes serve as panels and as the cross-section identifiers. Empirical results are then reported, and the discussion focuses on the effects of various interest rates and economic variables on NIM. Final results suggest that (1) interest rate effects differ by asset class, (2) economic variables have no effect, and (3) fixed effects, which are often important in panel studies, were unimportant in the final model, suggesting that time-constant factors unique to each asset class have probably been identified.

Background on Net Interest Margins

Customarily, net interest income is presumed to be inversely related to the level of interest rates. The variation in net interest income stemming from changes in interest rates has four aspects (Basel Committee on Banking Supervision, 2003): (1) repricing risk, (2) yield curve risk, (3) basis risk, (4) optionality risk. The first, repricing risk, stems from a mismatch in the maturity structure of assets and liabilities. The presumption is that banks are liability sensitive, meaning that liabilities reprice faster than assets as interest rates change. Mirrored in a positive duration gap typical for hanks, liability sensitivity is amplified by the rotation in the yield curve as interest rates change, forming the second source of interest rate risk, i.e.. yield-curve risk. Yield-curve risk results from the flattening and possible inversion of the yield curve when rates rise. The opposite is experienced when interest rates fall.

The link between the other two aspects of interest rate risk and the inverse relationship between interest rate movements and NIM is less clear-cut. Basis risk stems from imperfect correlation in the timing of interest rate changes for assets and liabilities. Banks may well use this as a means of reducing the effects of the first two sources on NIM. The last, optionality risk, stems from options embedded in various mortgage-related instruments, though options may also be embedded in liabilities. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.