Anat Admati is professor of finance and economics at Stanford Graduate School of Business. Martin Hellwig is former director of the Max Planck Institute for Research on Collective Goods. In 2013, only a few years after the global financial crisis, they came out with the much-heralded The Bankers’ New Clothes, and now they have released a “new and expanded edition.” “Our purpose in writing this book,” they proclaim, “is to demystify banking and explain the issues to widen the circle of participants in the debate.”

They define “the bankers’ new clothes” as “flawed and misleading claims that are made in discussions about banking regulation.” What’s wrong with banking, they say in short, is that bankers use too much debt to cover their investing. Accordingly, Admati and Hellwig want to require bankers to use more equity. This second edition contains new chapters on central banking, bailouts, and the rule of law.

Equity and debt / If you think bank capital means deposits, you are thinking the way bankers want you to think, but that is wrong. That money, in essence, is lent by depositors to banks, and the banks are obligated to pay it back. Meanwhile, bank capital, specifically, is money the banks receive from their investors. “In the language of banking regulation,” Admati and Hellwig clarify, capital just “refers to the money the bank has received from its shareholders or owners.” Capital is equity: a nonborrowed source of funds.

Bankers argue that requiring them to hold more equity will reduce their lending and slow economic growth. “In fact,” the authors inform us, “capital regulation does not tell banks what to do with their funds or what they should hold.” Every dollar of shareholder equity may be lent in the same way that every dollar borrowed may be lent.

But bankers resist financing their asset purchases with more equity because equity is a more costly source of funds. Admati and Hellwig give the bankers’ perspective:

The view that it is more expensive to use equity funding than to fund by borrowing is sometimes justified by the observation that for each dollar they invest in a bank’s shares, shareholders “require” a higher return than debt holders require.

The authors deem that perspective as “yet another article of the bankers’ new clothes.” Required rates of return, they explain, depend on the combination of equity and debt that corporate managers use. Using more debt increases the risk of default and causes lenders to ask higher interest rates. Likewise, shareholders raise the return on equity they require when a corporation uses more debt. Put the other way around, using more equity and less debt reduces risk. Shareholders will reduce the return on equity they require.

Following the authors, suppose Kate has $30,000 for a down payment on a house, borrows $270,000, and buys a house for $300,000. Her equity-to-asset ratio is 10 percent. Paul puts $30,000 down, borrows $120,000, and buys a house for $150,000. His equity-to-asset ratio is 20 percent. Paul is using more equity relative to his investment. The return on equity (ROE) equals the return on assets (ROA) times the reciprocal of the equity-to-asset ratio. If the price of Kate’s house rises 5 percent, her ROA = $15,000 ÷ $300,000 = 5%. Her ROE = 5% ÷ 10% = 50%. If the price of Paul’s house also rises 5 percent, his ROA = 5% and his ROE = 5% ÷ 20% = 25%. Paul’s return on equity is lower than Kate’s because his $30,000 funds a smaller investment. But by using more equity relative to his investment, Paul is further from insolvency in case of a decrease in the price of his house. A decrease in the price of Kate’s house that is over 10 percent will render her insolvent. To render Paul insolvent, it would take a decrease in the price of his house greater than 20 percent. That is the sense in which using more equity is less risky.

Funding costs and returns / Whether a corporation that uses more equity will experience higher “funding costs” depends on the corporation’s profits. For instance, “using a different mix [of equity and debt] might affect such things as the amount of taxes the corporation pays, the subsidies it receives, or the investment decisions it makes.” Given that a corporation may subtract interest expense from income before paying taxes, using more equity may increase funding costs.

Admati and Hellwig add that banks offload the consequences of bad investments onto the Federal Deposit Insurance Corporation (FDIC) and taxpayers, too, in the case of bailouts. “This can make borrowing cheaper and more attractive for banks,” they continue, “but such cost savings are paid for by others and therefore should not affect policy.” The implication is that if a bank uses more equity, it may experience an increase in funding costs. However, that is not a valid reason not to require banks to use more equity because the cost to taxpayers will decrease.

Some bankers say dumb things. “Because we have this substantial self-funding with retail deposits,” said one banker, “we don’t have a lot of debt.” That banker does not know that deposits are a bank’s liabilities, or he is trying to mislead. Recall that bankers claim that using more equity is more costly because shareholders require a higher return on equity than lenders require on their loans. Admati and Hellwig report that “bankers and others routinely claim that having more capital would ‘lower returns on equity.’” That is not necessarily the case.

Consider the authors’ numerical illustration. Suppose a more leveraged bank has a return on assets of 1 percent and an equity-to-asset ratio of 10 percent. Its ROE will be 1% ÷ 10% = 10%. A less leveraged bank with an equity-to-asset ratio of 20 percent and the same return on assets of 1 percent will have an ROE = 1% ÷ 20% = 5%. That is a banker’s view: using more equity lowers shareholders’ ROE. Admati and Hellwig open the reader’s eyes to the alternative scenario when a bank incurs a loss. Suppose the return on assets is –1% for both banks. The more leveraged bank’s ROE will be –1% ÷ 10% = –10%. The less leveraged bank’s ROE will be –1% ÷ 20% = –5%. The less leveraged bank’s ROE is higher.

When bankers claim that using more equity will lower the return on equity, they are not telling the whole story. Banks that use more equity will have higher returns on equity (“less negative”) during bad times such as financial crises. Perhaps bankers do not know this, or their self-interest clashes with better policy.

Bankers’ incentives / Bankers fixate on ROE because their pay depends on it. The trouble is that bankers borrow more to increase ROE and their pay, and that increases risk. Bank managers and bank shareholders are not the only bearers of that risk; lenders and taxpayers share it.

Admati and Hellwig portray bank shareholders as naive. They claim that “because derivatives trade over the counter and not on organized exchanges…, shareholders might not even be aware of the risks that are taken.” Even if, as seems likely, bankers own shares, their compensation is sufficient over time to outweigh losses that eventually appear.

Lenders are not naive, according to the authors. Depositors accept low interest rates because their deposits are insured by the FDIC. Banks pay insurance premiums to the FDIC; however, the premiums are not based on the risks a bank takes. Banks benefit from “implicit government guarantees” such as government bailouts, Federal Reserve purchases of “lower quality” securities, and borrowing directly from the Fed at low interest rates. Buyers of bonds sold by banks expect the government to support banks in times of crisis, and therefore they accept lower interest rates.

Policy proposals / Admati and Hellwig characterize the banking industry as “fragile,” by which they mean near insolvency. They urge regulators to prohibit banks from making dividend payments so that the banks will increase their equity. With more equity, banks can make more loans or pay down their liabilities. Also, the authors urge regulators to mandate that banks sell new shares of stock to increase equity. The incoming funds can likewise be lent out or used to pay down liabilities. If a bank has no earnings to retain or cannot find investors willing to buy shares, regulators should consider closing it.

The concept of “risk weighting” assets seems sensible: Banks should use equity to finance riskier assets, but may use debt to finance the purchase of riskless assets. But the application of this concept is problematic. Admati and Hellwig put it this way:

In theory, risk weights are meant to adapt equity requirements to the risks of the banks’ investments; in practice, the weights are determined by a mixture of politics, tradition, genuine and make-believe science, and the banks’ self-interest. In this mixture, some important but real risks are overlooked.

Their critique is supported by evidence that during the tumultuous year of 2008, banks with more equity relative to total assets were less likely to fail than banks with more equity relative to risk-weighted assets.

The authors want banks to increase their equity until their equity-to-asset ratios are in the range of 20–30 percent. Bankers will resist. But as Nobel Economics Prize winner Merton Miller pointed out, bankers expect similar commitments from their borrowers. The chief benefit of requiring banks to use more equity is a reduced likelihood of financial crises.

The authors demonstrate this with a numerical example: If a bank’s equity is 3% of its assets and the value of its assets decreases by 1%, 33% of its equity disappears. But if a bank’s equity is 25% of its assets and the value of its assets decreases by 1%, just 4% of its equity disappears. We may expect fewer bank failures and fewer bailouts paid by taxpayers. That would be great—except politics makes such policy unlikely.

Readers will learn what happened at Silicon Valley Bank (SVB). The primary source of SVB’s funds was “uninsured short-term deposits,” and its primary use of those funds was to buy bonds. When the Fed raised interest rates in March 2022, SVB veered toward insolvency. Some of its depositors moved their money to money market funds paying higher interest rates. Simultaneously, the value of its bonds decreased. It tried to meet deposit outflows by selling bonds and selling new shares of stock, but the bond sales at depressed prices were insufficient and there was no appetite for its stock. California’s Department of Financial Protection and Innovation declared SVB insolvent and took over the bank. The FDIC paid off SVB’s depositors, including uninsured deposits over the coverage limit of $250,000.

The case of SVB serves as another lesson for regulatory reform. Based on current accounting rules, a bank “usually” puts bonds on its balance sheet at face value. If bond prices fall, the bonds remain on the balance sheet at face value on the condition that the bank “intends to hold debt securities until they are repaid.” When faced with deposit outflows, a bank may need to sell bonds before they mature. If the proceeds from the bond sales are insufficient to meet the deposit outflows, current accounting rules enable a bank to “hide” the insolvency. Thus, Admati and Hellwig propose “market-value accounting” whereby a bank lists bonds on its balance sheet at market prices.

Conclusion / Readers who blame markets for bad outcomes will appreciate Admati and Hellwig’s condemnation of bad bank behavior. For instance, they describe how, during the hot real estate market of the 2000s, Washington Mutual, Bear Stearns, and probably JP Morgan Chase “marketed mortgage securities that [their] employees knew were of poor quality while misrepresenting the quality to investors or providing inaccurate information when asked.” Readers who blame government officials will appreciate insights such as this: “Politicians want the bankers’ cooperation to make the investments the politicians favor—or campaign contributions.” Admati and Hellwig state, “There are many instances where flawed government activities are a problem, but without laws and without law enforcement by government, markets would not work and we would all be much worse off.”

Requiring bankers to use more capital is a reasonable way of assigning the costs of risky behavior to those engaging in the behavior. If the authors are correct that there are no valid objections to this proposal, perhaps the political obstacles to it will be overcome and the future will show whether they are in fact correct.