More evidence that bail-ins are better than bailouts

Government responses to financial crises can be sorted according to who bears the losses. The state can concentrate the pain on the individuals and institutions who made bad decisions — they can be bailed in — or policymakers can bail out the people who caused the problem by spreading the damage across the rest of society.

Bail-ins are theoretically preferable because they preserve market discipline without causing undue harm to innocent people. Bailouts, by contrast, are unfair and inefficient. Governments tend to do them, however, out of misplaced concern about “preserving the system”. This stokes (justified) resentment that elites care about protecting their friends more than they care about helping regular people. Anyone concerned about institutional resilience should therefore be wary of copying too closely what American and European countries did in response to 2008.

As if that weren’t reason enough to prefer bail-ins to bailouts, new research from the Federal Reserve Bank of New York argues the bailout habit is also bad for growth. Bailouts encourage excessive risk-taking, which end up producing larger and longer-lasting busts. It would be better to accept pain earlier rather than postpone the inevitable when the total cost is far higher.

The NY Fed research focuses on the most opaque* form of bailout: “regulatory forbearance”.

The thinking is that banks can make more money when they are less encumbered by rules. This boost to earnings is supposed to help offset losses from past mistakes. The danger is that the rules were there for a good reason, which would mean the higher profits are temporary and the net result is larger losses in the future. Bear this in mind when Larry Summers says banks are “riskier” than they should be because they have fewer ways to rip people off than in the past.

Officials can also tweak the accounting regime to allow banks to recognise losses more slowly. This suppresses their funding costs and gives them more time to make good on their obligations. Done well, these tweaks can stop self-fulfilling crises from spiralling out of control — accounting rule changes were an important part of America’s policy response in early 2009. But in general the rules are there for good reason, and leaving banks free to obscure the quality of their assets encourages poor choices.

There are lots of ways to study the impact of regulatory forbearance. Comparing the Nordics and Japan in the 1990s is popular, since they took opposite approaches and more or less got opposite results.

Sweden took its lumps early, nationalising many insolvent banks and dividing up assets into a new “good bank” and a “bad bank” that would gradually liquidate over time. After a sharp and deep recession it ended up being one of the world’s most successful rich economies and largely avoided the 2008 crisis. Japan, by contrast, encouraged its banks to avoid recognising losses by lending new money at low rates to borrowers who would never be able to repay their original debts. The subsequent slowdown in growth led many to describe Japan has having entered a “lost decade”.

This new study looks at America in the 1980s, which has the advantage of comparing regions within the same country. The variation comes from differences in the type of supervision faced by lenders depending on whether they were under the purview of the Federal Deposit Insurance Corporation or the Federal Savings and Loan Insurance Corporation. Similarities in business models meant that the lenders covered by each regulator were broadly similar before the 1980s.

The FDIC leaned more towards bail-ins while the FSLIC preferred bailouts using regulatory forbearance. The places with a larger share of distressed lenders regulated by the FDIC tended to grow somewhat more slowly in the 1980s but did much better in the ensuing balance sheet recession, while places with more lenders regulated by the FSLIC had credit and housing booms in the 1980s followed by severe downturns in subsequent years:

Forbearance leads to a relative boom on some dimensions, followed by a broader, wide-spread bust that registers in aggregate output growth. Forbearance is estimated to initially lead to a greater supply of higher-risk loans, accompanied by greater construction activity, job creation and destruction, and new business starts.

But, after normal regulatory requirements are re-imposed nationwide, forbearance is associated with larger contractions in real estate and cumulative average declines of more than 3% in real GDP, coinciding with a recession in 1990-1991…Under more consistent regulatory policy, the estimates suggest GDP growth may not have been quite as high in the mid 1980s, but may have avoided turning negative in the early 1990s.

Back in the days of relatively stable prices, smaller banks could do quite well making mortgages funded by deposits and living off the interest rate spread. That model fell apart when short-term interest rates soared during Volcker’s disinflation campaign. Older mortgages made when rates were lower didn’t provide enough interest income to cover the cost of deposits, and this wasn’t likely to change.

In other words, many, if not most, of these lenders were insolvent. The longer they operated the longer they would waste resources that could be better-used elsewhere. Proper regulators would have shut them down, liquidated their assets to repay depositors up to the insured amount, and cover any differences with the resources in the insurance funds. This is what the FDIC tended to do.

However, the FSLIC only “had $6 billion in funds and 34 employees in 1980”. If it wanted “to deal with all of its insolvent institutions” it would have needed more than “$25 billion” — and presumably a lot more employees. The FSLIC ended up adopting more lenient standards than the FDIC. (National policy in the early 1980s also supported forbearance in part by justifying it as pro-competition deregulation.)

Instead of getting shut down, insolvent lenders were encouraged to gamble for resurrection, often by entering new markets they didn’t understand. Initially, this made them more money and appeared to help them with their problems. The strategy looked so successful that solvent banks decided to follow their neighbours and expand into new, riskier business lines. The extra competition meant it was easier for borrowers to get credit, which increased the danger of future losses while also inflating asset prices. It was also easier to commit outright fraud.

Banks are in the business of making bets with other people’s money, subsidised by the state. This is an inferior arrangement to a nationalised payment system with privatised capital allocation, but it can be made to work if lenders who make bad decisions are punished swiftly. As America’s experience in the 1980s shows, anything else is a transfer that is both unjust and inefficient.

Related links:
Anyone awake in the 1980s should have known about the dangers in the 2000s — FT Alphaville
Banks have a dubious business model and markets have noticed — FT Alphaville
Illiquid, insolvent, what’s the difference? — FT Alphaville
“People want money” — FT Alphaville

*The most straightforward form of bail-out is to transfer money from regular people to investors who would otherwise have lost money using the tax and spending powers of the fiscal authorities. Borrowing, rather than taxing, is a slightly subtler form of transfer. For one thing, the extra debt could have been used to fund more worthwhile programmes. And the debt will eventually be serviced through some combination of taxes, financial repression, and surprise inflation/default, all of which come at the expense of regular people.

Slightly less straightforward is to bailout the culprits using the central bank. Monetary stimulus makes it easier for borrowers to repay their debts at the expense of savers. It also tends to increase the difference between short-term and longer-term interest rates, which boosts bank profits and helps them “earn their way out” of their previous poor choices.

Central banks can directly support risky asset prices by accepting them as collateral for loans or by buying them outright. Offering below-market loans to stressed financial firms (emergency lending) is an especially opaque way to discreetly transfer resources from the many to the few. During a crisis, it’s impossible to distinguish between a fundamentally healthy business facing a temporary bit of trouble and a bad business that should succumb to its wounds — these are contingent conditions.

Copyright The Financial Times Limited 2019. All rights reserved. You may share using our article tools. Please don't cut articles from FT.com and redistribute by email or post to the web.

Read next:

Read next:

Lookout, there’s a dollar crunch!

FT Alpha Tweets