Organizational Memory and Bank Accounting Conservatism

ABSTRACT

This paper is the first to investigate the impact of banks’ organizational memory of past history on the conservatism of accounting policy. Specifically, we investigate two types of bad time history: banks’ undercapitalization and the failures of other banks during financial crises. Using a large sample of U.S. banks over the period 1997–2013, we find that both types of bad times are positively related to timelier recognition of earnings decreases versus earnings increases in accounting income. We also find that following bad times, banks increase their allowance for loan losses. The results of path analysis and survey research indicate that bad time memory of banks impacts bank accounting conservatism through CEO tenure and board of directors’ tenure. Collectively, our results suggest that banks’ organizational memory of bad times and macro-level banking crises lead to greater accounting conservatism in banks.

Keywords:

• Organizational memory
• Bank accounting conservatism
• Bank history
• Allowance for loan losses

Notes

1 The institutional/individual memory literature argues that institutions/individuals put more weight on realizations experienced during their lifetimes than on other available historical data. Unique insights can emerge from this argument (Berger & Udell, 2004; Malmendier & Nagel, 2011, 2016): (1) young institutions/individuals, react more strongly to recent experiences than do older institutions/individuals, who already have a longer data series accumulated in their lifetime histories; (2) the memory of past experiences vanishes over time, but effects of extreme events can last for a long time. In this paper, we assume that young and older institutions/individuals react to recent experiences in a similar way due to the lack of data on young and older institutions/individuals.

2 ‘Routines not only include the forms, rules, procedures, conventions, strategies, and technologies around which organizations are constructed and through which they operate, but also include the structure of beliefs, frameworks, paradigms, codes, cultures, and knowledge within the organization that buttress, elaborate, and contradict the formal routine’ (Levitt & March, 1988).

3 We thank the anonymous reviewer and the associate editor for suggesting us to do this survey study and confirm our regression results with U.S. bank CEOs and CFOs.

4 Using the internet websites of the banks listed in Q4 2012 call report, we found that 601 banks out of 6,752 in the U.S. posted the emails of senior executives (i.e., CEO, CFO, president, and chairman) on their websites. We collected 969 emails of senior bank executives (424 CEOs, 301 CFOs, 174 presidents, and 70 chairmen). We sent our survey questionnaire emails to 969 senior bank executives; 163 emails were returned to us because 1) some banks’ firewall systems blocked our emails; and 2) some bank executives were on vacation and sent automatic reply emails to us.

5 The locations of these bank executives who responded to our survey questions are randomly distributed throughout the U.S.

6 Alternatively, we remove the duplicates in each executive’s choices and add up the choices of the 13 executives. We get 10 choices supporting ‘by bank managers’ incentives’, 10 choices supporting ‘by board of directors’ inside monitoring’, and 8 choices supporting ‘by auditors’ external monitoring.’ Some bank executives did not differentiate between federal auditors and CPA firm auditors. They stated that federal auditors are important forces driving bank accounting conservatism by exerting pressure on managers, boards, and auditors and guiding them to increase reserves for loan and lease losses. Thus, part of the 8 choices supporting ‘by auditors’ external monitoring’ actually support federal auditors’ external monitoring.

7 DSCR is debt service coverage ratio; LTV is loan-to-value ratio; and ALLL is allowance for loan and lease losses.

8 CECL is current expected credit losses.

9 We broaden the tests to include the bank’s cumulative history in the past three or five years, as in Bouwman and Malmendier (2015). Our main results remain robust to these additional tests.

10 Tier-1 risk based capital ratio is the ratio of a bank’s ‘core capital’ to its risk-weighted assets. Risk-weighted assets are constructed by assigning different weights to assets with different levels of risk and summing the totals. The tier-1 risk-based-capital ratio measures how much buffer a bank has as a percentage of its riskiness. We focus on this particular ratio because it excludes more ‘exotic’ elements from the calculation of capital and so serves as a better approximation of an adequate capital ratio.

11 Beatty and Liao (2011) scale the loan loss allowances by non-performing loans. Our results are robust to this minor difference.

12 The incidence of bad times is low, based on Table 1, Panel A. Undercapitalization occurs in 0.2% of bank-years, and the average percentage of banks to fail in a state in any given year is 0.3%. This makes it difficult to assess the economic significance of the results, as well as whether the results might generalize to bad times more broadly defined. The results may be driven by a small set of observations. This compounds the concerns about generalizability inherent in focusing exclusively on financial institutions.

13 In path analysis where there is no variation among our sample for the dummy variable $UNCAP$ due to the significantly reduced number of observations after filtering the sample by CEO tenure, board tenure, or auditor tenure, we define $UNCAP$ as the negative of the Tier 1 risk-based capital ratio, so that a high value of $UNCAP$ represents bank-specific bad times low capitalization.

14 The Office of the Comptroller of the Currency (OCC) is an independent bureau within the United States Department of the Treasury and serves to charter, regulate, and supervise all national banks and thrift institutions in the United States.