Banking Sector Accounting Practices and Financial System Stability

Abstract

This paper investigates the underexplored nexus between heterogeneous accounting practices within the banking sector and the emergent stability properties of the financial system
as a whole. Departing from traditional analyses that treat accounting as a neutral, transparent reporting mechanism, we conceptualize accounting frameworks as active, constitutive
elements of the financial ecosystem that shape bank behavior, risk perception, and interbank
dynamics. Our novel contribution lies in modeling the financial system as a complex adaptive
network where nodes (banks) employ one of three distinct accounting paradigms: Historical
Cost Accounting (HCA), Fair Value Accounting (FVA), and a proposed hybrid, Dynamic
Provisioning Accounting (DPA). We develop an agent-based computational model to simulate the propagation of liquidity and solvency shocks under varying compositions of these
accounting practices. The model incorporates feedback loops where reported accounting figures directly influence market confidence, collateral values, and interbank lending decisions,
thereby endogenizing systemic risk. Our results reveal several non-linear and counterintuitive findings. First, a system predominantly using FVA exhibits higher volatility and faster
shock transmission, confirming some post-crisis critiques, but also demonstrates a greater
capacity for early loss recognition and rapid system cleansing. Second, a homogeneous HCA
regime fosters apparent short-term stability but can lead to the accumulation of hidden
losses and larger, delayed systemic collapses—a ’stability illusion.’ Third, and most originally, we find that a deliberately heterogeneous mix of accounting practices, particularly one
that strategically embeds DPA banks as ’circuit breakers,’ can enhance systemic resilience by
dampening pro-cyclical feedback and creating asynchronous response mechanisms to shocks.
The optimal mix is non-trivial and depends on network topology and shock origin. We conclude that financial system stability is not merely a function of individual bank capital but is
profoundly mediated by the diversity and design of the accounting rulebook itself. This argues for a macroprudential approach to accounting standard-setting that considers systemic
network effects, moving beyond the micro-level ’representational faithfulness’ paradigm that
has dominated standard-setting discourse.