Bank Capital Redux: Solvency, Liquidity, and Crisis

Abstract

What is the relationship between bank capital, the risk of a financial crisis, and its severity? This paper introduces the first comprehensive analysis of the long-run evolution of the capital structure of modern banking using newly constructed data for banks’ balance sheets in 17 countries since 1870. In addition to establishing stylized facts on the changing funding mix of banks, we study the nexus between capital structure and financial instability. We find no association between higher capital and lower risk of banking crisis. However, economies with better capitalized banking systems recover faster from financial crises as credit begins to flow back more readily.

Bank capital: Excess credit and crisis incidence

“…We should note that capital and liquidity are significant in our full period experiments, even those with the more restricted crisis definition in Table 2. This is contrary to the post 1945 results in Jordà et al (2017) and would lead us to very different policy conclusions from theirs for the current, post Bretton Woods, period. If we added the 35 years between the end of the Second World War to the start of our data, we would add no crises until after 1972, and then only crises in the UK and Spain.…”

“…Hence it would not surprise us if capital became insignificant if we added those observations to our data, and the lack of growth in credit up until 1972 meant that it seemed to explain (the lack of) banking crises. However, we think the liberalised post-Bretton Woods era should be explained by different factors than the repressed 1940s to early 1970s, and it does not surprise us that our results differ from those of Jordà et al (2017) and hence so do our policy conclusions.…”

“…1980, at least as far as estimates in Jordà et al (2017) suggest. Hence it would not surprise us if capital became insignificant if we added those observations to our data, and the lack of growth in credit up until 1972 meant that it seemed to explain (the lack of) banking crises.…”

“…It has been common to link this crisis, and others to the twin problems of excessive credit growth and the subsequent unsustainable growth of asset prices, and, particularly, property prices. The link between credit growth and bank based financial crises has been emphasised in a series of papers covering a period of over 130 years of history in 17 developed economies by Schularick and Taylor (2012), Jordà, Schularick and Taylor (2011) and Jordà, Richter, Schularick and Taylor (2017) and has been supported by the views and the publications of the Bank for International Settlements (BIS). 2 The evidence to link crisis incidence to credit growth over the last forty years is, however, weak, and evidence of the link from earlier periods may not be relevant for the analysis of policy problems in a set of financially liberalised advanced economies.…”

What Do the Two Global Crises Tell us: History as a Mirror

“…Higher capital ratios are not equivalent to lower systemic financial crisis risks [15] . Liquidity indicators are more reliable than capital ratios for measuring financial fragility on balance sheets.…”