What do you mean that banks use capital and debt to finance their asset portfolios?
When a bank purchases an asset—say a bunch of shipping loans—it must fund the purchase with something or other. Some of the purchase price is paid with bank capital—money raised by selling shares or earned by doing business—and some is paid with money the bank borrows. Increasing the level of debt financing relative to capital financing increases a bank's leverage.
Remember Bailey Building and Loan Association in "It's a Wonderful Life"? That bank made loans to people so they could buy houses. It financed a large part of those loans with savings deposits from customers, which is a form of borrowing. And that wound up getting the bank in trouble when depositors decided to withdraw their funds all at once. George Bailey wound up using his honeymoon savings to "recapitalize" his bank, thereby reducing the bank's leverage.
So what the regulators are doing here is placing a stricter limit on the relative level of debt that can be used to finance our biggest banks' assets.
So why is the increase in leverage ratios considered restrictive?
This is a bit confusing. When the leverage ratio increases to 5 percent, what that really means is that the ratio of debt to capital is decreased to 20:1, that is for every $20 of borrowed money a bank has to use $1 of capital to finance its assets. So the 6 percent ratio for FDIC-insured banks means that they have to use even more capital to finance their assets.
Why are regulators requiring this?
The goal is to make the biggest banks less risky, less prone to fail.
How does a regulatory leverage ratio accomplish this?
When a bank's balance sheet is financed with more capital, it is better able to absorb losses on its assets. So a bank is less vulnerable to a rise in defaults on its loans or a market downturn that depresses the prices of assets it holds.
Is bank capital like a rainy day fund?
No. Bank capital isn't a fund of money sitting in a vault somewhere. It's rather just a type of funding that the bank uses to finance the purchase of assets and the making of loans. It's generally money the bank has spent to acquire stuff, rather than borrowed to acquire stuff.
How does capital absorb losses?
Suppose we have a bank with $100 in assets, which have been financed with $3 of capital and $97 of debt. If the value of the bank's assets falls by $2, the bank can take the loss by writing down the equity portion. The bank will still be solvent because its assets are worth more than what it borrowed. But if the value of the assets dropped by $4 instead, the bank would now be insolvent. It would be unable to repay its debts.
The key to this is that capital is the amount a bank can afford to lose before its debts exceed its available assets. The capital portion can be written down because a bank's shareholders have agreed to absorb the losses by having the value of their contribution to a bank's balance sheet written down. The debt portion cannot be written down because debt is made up of promises to pay bondholders or depositors fixed amounts.
What do you mean "writing down the equity portion?"
That just refers to how the loss is accounted for on the bank's books. Basically, the reported value is reduced. And since equity holders are the residual claimants of a bank's assets—meaning, in a bankruptcy they get whatever is left over after all the debts are paid off—they take the first loss when a bank's assets decline.
For example, if a bank has $100 worth of assets and $97 of debt, the equity value of the assets is $3. That's what shareholders would get if the bank had to be unwound. If the asset value drops to $98, the debt holders still get paid $97 and the shareholders would get just $1. So we say the equity portion has been "written down."
So under the new rules, banks would have a 5 percent capital cushion, right?
Not quite. The leverage ratio is a regulatory minimum that the bank is always required to meet. This means that if a bank really is leveraged all the way to the limit, it cannot have any net losses whatsoever without being at risk of significant regulatory action, perhaps including being seized by regulators. Because of this reality, the banks are very likely to have capital in excess of the requirements—a voluntary cushion on top of the regulatory cushion.
When a bank's losses push its leverage toward the required ratio, it will have to raise more capital either by retaining earnings or issuing new equity. Alternatively, it can reduce its assets and liabilities by selling loans and paying off the debt it used to acquire them.
The required leverage ratio is meant to make sure that banks get swept into regulatory care before they actually go insolvent.
Why not leave it up to the market to decide what the leverage ratio of banks should be?
Theoretically, the market could decide the appropriate level of leverage. If the market decided a bank is overleveraged—meaning, it has borrowed so much that expected losses render it at risk of defaulting on its liabilities—the bank would find itself unable to borrow to finance its balance sheet. Since all banks depend on short-term borrowing at least to some extent, this would create a market check on leverage.
If there's a market check on leverage, why do we need a regulatory limit at all?
Basically, because government bailouts break the market process that would impose limits.
Take the most basic bailout available in our system: the FDIC's deposit insurance. The availability of deposit insurance means that depositors—who are an important source of credit to banks—don't worry about leverage. We don't get very many "It's a Wonderful Life" style runs on banks anymore. This is utterly intentional: We don't want bank runs so we make depositors immune to losses. The cost of this, however, is that we lose depositors as a check on bank leverage.
For very large banks—like the ones included in these new rules—there is also the problem known as "Too Big to Fail." Because the market doesn't believe the government will allow TBTF banks to go bankrupt, it won't monitor bank risk appropriately.
How can you say the market won't monitor bank risk? What was 2008 if not people freaking out about bank risk?
There were two types of fear we saw arise in 2008. One was the fear that the government would not stand behind some TBTF banks. This proved to be justified in the case of Lehman Brothers. After Lehman Brothers filed for bankruptcy, the fear that TBTF banks weren't as safe as they seemed triggered a widespread panic. Importantly, however, this kind of market check only seems to work during a panic—at which point it is too late. In normal times, markets don't seem to be good risk monitors of bank balance sheets.
The other fear, however, was that banks would have to cut dividends and raise new capital to avoid seizure by regulators. Meredith Whitney's famous call that Citigroup would cut its dividend was basically about the need to stay in compliance with regulations. And, again, this only really arises too late, when a crisis is at hand.
Won't banks make fewer loans if higher leverage limits are imposed? Is this bad for access to credit?
All things being equal, yes.
A bank that keeps its capital level exactly the same will have to shrink its balance sheet and make fewer loans in the future.
But all things are not equal. Making fewer loans means reducing income. This means shareholders will receive smaller dividends unless the bank dramatically cuts back on other costs. The banks hit by this rule, however, will likely choose to raise more capital—most likely by retaining earnings and reducing dividend payments—in order to be able to continue to grow their income under the new rules.
So there will be some loans banks won't make if the capital component is higher?
That's right. Some loans will have risk-adjusted returns too low to justify the capital required to make them. This is a regulatory choice we're making. We don't want banks reaching for these loans because it makes them too likely to fail.
Does the higher leverage limit apply to each and every loan a bank makes?
No. It applies to the total mix of bank assets. Some loans will be funded with less capital, some with more. But the aggregate mix of capital and debt will have to hit or exceed the target.
Mervyn King, the outgoing governor of the Bank of England, recently said: "Those who argue that requiring higher levels of capital will necessarily restrict lending are wrong. The reverse is true. It is insufficient capital that restricts lending." So will banks lend more?
This is a claim put forth by proponents of very high capital requirements but it is partially wishful thinking. Voluntarily high levels of capital can indeed encourage lending during a credit crunch. Regulatory capital requirements, however, have the opposite effect. They make it more likely that a bank will have trouble financing loans in a credit crunch because creditors will fear seizure by regulatory authorities. What matters, really, when it comes to the ability to make loans in a credit crisis is having enough capital to stay both solvent and to stay above regulatory requirements. High capital requirements makes the first part easier and the second part harder.
That said, regulators seizing a bank will seek to pay off creditors as best they can with the bank's assets. So a higher leverage ratio would reduce the ultimate loss from a bank defaulting on its obligations, which should mean that bank counterparties would be less fearful of seizure when a bank has a higher leverage ratio than a lower one.
Why are banks so resistant to this change?
Raising capital is expensive. It requires a bank to withhold dividends from shareholders or dilute existing shareholders by issuing new shares. Shareholders don't like either of these things and they tend to reduce the price of shares. So banks would prefer to avoid raising capital.
What's more, higher capital requirements mean that a bank can't expand just by expanding its debt financing when it spots opportunities. They must wait to expand their balance sheet while they accumulate capital. Banks worry that this will make them less nimble.
The economists Anat Admati and Martin Hellwig argue in their book "The Emperor's New Clothes" that capital is not expensive. Are you saying they are wrong?
That book is very instructive and I highly recommend it to everyone. But on this one point some of their readers have made the mistake of confusing regulatory capital requirements and voluntary capital levels. (I don't think Admati and Hellwig are guilty of this mistake, however.)
The basic argument is that higher capital levels make banks safer investments. So shareholders will be happy with lower returns because the risk is lower and rational shareholders want appropriate risk-adjusted returns. So additional capital shouldn't be all that expensive.
As far as this goes, it's fine. But it doesn't go far enough. Shareholder risk is not really diminished by higher requirements because the primary risk to shareholders is seizure by regulators—not actual insolvency. The higher capital requirements move the bar further away from involvency but don't better isolate shareholders from the seizure risk. Only voluntarily higher capital levels—above regulatory levels—do this.
That said, you won't necessarily get seized if you violate the leverage ratios. Instead, you'd have a bunch of meetings with regulators and they'd tell you to sell equity until your capital levels came up to standards. So you'd rather run seizure risk than solvency risk if you're a shareholder.
So this is bad for investors in banks?
It will likely put pressure on bank shares, especially on those with higher levels of debt to capital. In the long run, however, it should reduce the riskiness of banks, which would reduce share volatility. It could boost bank bonds because they will be less at risk as a result of the larger level of losses that would fall to the shareholders.
Which banks will be hurt the most?
Hurt is the wrong term. Some banks may find it easier to raise capital even if they are required to raise larger amounts because investors have more confidence in management.
The Wall Street Journal reports that analysts at Keefe, Bruyette & Woods predicted a shortfall under the new standards of $15.6 billion at JPMorgan Chase, $14.1 billion at Morgan Stanley, $13.2 billion at Citigroup and $4.9 billion at Goldman Sachs. Two of the eight bank-holding companies already exceed the required threshold, Bank of America and Wells Fargo.
Is it bad for consumers?
Consumers are unlikely to feel much of a negative impact from this. Perhaps some very risky credit lines will become unavailable—but those were mostly a trap for financially unsteady consumers anyhow. Getting rid of them is probably a benefit for consumers. Ordinary consumers seeking credit will continue to benefit from bank competition for their business, which will limit the ability of banks to pass on higher capital costs.
Is it bad for the economy?
Banks have been arguing that the requirements will slow economic growth. That's a strange argument to make considering the fact that we're living through the worst economic slump in generations thanks to the financial crisis of 2008. Whatever drag occurs on economic growth due to the new leverage requirements will likely be less over time than the drag that follows from a crisis brought about by undercapitalized banks. In the long run, it's likely to be an economic plus rather than a minus.
What's more, we are more than capable of adjusting fiscal and monetary policy to accommodate stricter bank leverage requirements. Better to have safer banks and let other policies adjust to counter any economic drag produced.
Will these requirements work? Are they strict enough?
No one really knows. Admati and Hellwig argue in their book for much higher and stricter requirements. At the very least, this looks like a step in the right direction.