6 ideas
Bank capital is not money set aside and left unused; it is the share of a bank's funding that comes from shareholders, not borrowers. The industry's line that capital requirements force banks to 'hold back' money that could be lent confuses the funding side of the balance sheet with the asset side. A bank with more equity can lend exactly as much as before.
A bank's equity works like a homeowner's down payment. The smaller the down payment, the more a small fall in house prices wipes out the owner and leaves the loss with the lender. Treating banks as borrowers makes plain that high leverage raises the chance of insolvency and the spread of losses to others, just as a thin down payment does for a household.
More debt makes equity riskier, so shareholders demand a higher return. Measured against that risk, equity is not truly 'expensive', and bankers' claims that it is ignore this relationship from Modigliani-Miller. Most of the private cost banks attribute to equity is really a loss of subsidies, not a cost to society.
Creditors expect governments to rescue large banks, so they lend to them cheaply regardless of risk. This, together with tax deductibility of interest, makes debt artificially cheap and rewards banks for piling on leverage. Taxpayers bear the downside while shareholders and managers keep the upside.
Bank executives are often paid on return on equity, which rises with leverage and hides the risk that comes with it. Judging performance by ROE gives managers a reason to resist equity requirements and take on hidden tail risk. The public debate therefore reflects managerial incentives, not what is good for the economy.
The authors propose that banks fund at least 20–30% of total assets, not risk-weighted assets, with equity. Until then, banks should retain earnings and stop paying dividends. Measuring against total assets avoids the manipulation that risk weights allow, and this level of loss absorption would make crises far less likely at little social cost.