Source: Ideas and Institutions Issue #27
Analysis
The Science and Art of Bank Failures
The failures of certain prominent banks in the United States and Europe have brought the issues of banking regulation and resolution back into public discussion. Whether this is an early stage of a systemwide crisis or just a series of one-off failures would depend not just on the quality of the bank assets but also on how the political economies and public administration systems respond. There is speculation that the U.S. Federal Reserve may, in the interest of financial stability, halt the monetary tightening it has pursued since the middle of 2022. There are also efforts underway to resolve the failed banks in a manner that panic about other banks can be avoided.
While the proximate causes for a bank’s failure are easy to see, the deeper causes are usually difficult to identify. We are still litigating the causes for the 2008 global financial crisis. So, each crisis provides an opportunity to learn, but it is not obvious that the right lessons are learned. The regulatory systems in the U.S. and Europe did change a lot after the 2008 crisis, and it remains to be seen whether they are better placed to avoid a major crisis. As of now, it is too early to say.
Banking has always been crisis-prone. As discussed in a previous issue of this newsletter, since banks run maturity mismatches—they take deposits that are withdrawable on demand while giving out long-term loans—if a large number of depositors go to a bank to withdraw their deposits at the same time, most banks would be unable to meet all the demands because they cannot immediately convert their assets into cash. If they try to do so, they will incur losses that may lead to their failure. Depositors rush to get their deposits from banks when they perceive that a bank is failing. Such a bank run becomes a self-fulfilling prophecy, making even a previously healthy bank fail. So, maintaining trust in the banking system is crucial particularly to avoid failure of healthy banks.
There are two public policy concerns relating to the safety and soundness of banks—the protection of depositors and the maintenance of stability and resilience of the financial system. Since banks take deposits from public for safekeeping, the protection of depositors makes the case for state intervention for maintaining safety and soundness of banks and resolving failed banks. Further, when a bank becomes systemically important, its failure could create system-wide repercussions.
Given its crisis-prone nature and the presence of the aforementioned public policy concerns, banking is a sector in which some state intervention has long been considered potentially beneficial, even though the actual benefits cannot be taken for granted as they depend on how the systems of state intervention work. There are three main forms of state intervention for maintaining safety and soundness of banks.
- Prudential regulation and supervision of banks: This is a preventive form of intervention that aims to preserve the safety and soundness of banks, so that the probability of bank failure remains at an acceptable level. This takes the form of requirements around capital, lending and investment standards, liquidity coverage, risk management, internal controls, and so on. Supervision includes monitoring, stress testing, corrective action, and so on.
- Systemic risk regulation of banks takes the form of additional prudential regulation for systemically important banks (additional capital requirements, stress tests, and so on), measures to protect the health of the banking system during a cyclical downturn (for example, countercyclical buffers), and additional supervisory attention.
- As the lender of last resort, the central bank provides liquidity (usually against some security) for the banks facing liquidity crisis. This facility is particularly useful for banks that are solvent but are facing a liquidity crisis.
Since bank failure is sometimes triggered by a perception of poor financial health, other banks can help send a positive signal about the bank’s health. In the U.S., a group of large banks placed $30 billion of deposits with the First Republic Bank to signal their trust in the bank. This effort is still underway and may or may not work out. Such efforts may be facilitated by a public authority, but they are usually based on rational choice by the fellow banks, because the latter are worried about the effects of the bank’s failure on the depositors of the banking system.
Despite all this, banks do fail sometimes. In fact, some bank failures are necessary because creative destruction is essential for the efficiency of the banking system. However, a failed bank needs to be resolved in an orderly manner to minimize the harm to the depositors and to systemic stability and resilience. Many countries have created specialized resolution mechanisms for detecting bank failure at an early stage and resolving the failed bank quickly (typically by merging it with another bank) so that a bank failure does not lead to panic in the banking system. In recent years, banks have also been asked to issue certain instruments whose function it is to absorb the losses when a bank is failing. In Credit Suisse’s case, the resolution involved write-off of these instruments and sale of the bank to UBS.
The resolution mechanism is usually accompanied by deposit insurance with a cap on the insured amount—when the loss of asset value is so large that the deposits cannot be repaid, the deposit insurance mechanism pays to the depositors to the extent of the coverage. To avoid the problem of a moral hazard—banks taking on too much risk in the hope of being bailed out—the deposit insurance framework guarantees protection of only deposits and that too up to a limit for which it collects insurance premium from banks. Further, resolution imposes losses on liabilities in an order of priority.
While the regulatory framework for banking in the United States is often criticized, there is one element that works quite well. This is the resolution and deposit insurance framework administered by the FDIC. The main virtue of this framework is its efficiency. Between 2009 and 2011, about 400 banks failed in US, and the FDIC was able to resolve most of them without any run on the banks and without a need for bailout. In the case of the Silicon Valley Bank, the entire receivership process was done over a weekend. The process started on a Friday evening, and the depositors could get access their deposits from Monday.
A key fact to consider regarding resolution of failed banks is that while there are extensive laws and regulations governing the decisions, they are not complete, and exceptions are sometimes made, usually when a systemwide crisis is feared. In fact, the laws themselves provide room for exceptions by giving broad, discretionary powers to regulators, central banks, and the government. The long history of state intervention in banking crisis is also a history of highly discretionary (and creative) decision-making in the face of uncertainty.
The most controversial of discretionary interventions are bailouts of banks using public funds. Such bailouts are often seen as an unfair use of public money, but whether they are an inefficient use of public money or not depends on a counterfactual—what would have happened in the larger banking system had the bailout not been done? Such decisions therefore remain controversial. The key question is whether the systemic implications of the bank’s failure outweigh the opportunity costs of public funds and the costs of moral hazard created by the bailout. However, short of a direct bailout backed by fiscal resources, other exceptions are also possible.
For instance, in the case of the SVB, after the bank failed and went into receivership by the Federal Deposit Insurance Corporation (FDIC), while losses were imposed on shareholders and bondholders, the depositors were allowed to access all their deposits immediately. Since the deposits are insured only up to $2,50,000, this is an undue benefit given to the depositors with large value deposits. This was considered necessary to avoid panic among large depositors of other banks. In SVB, many large value deposits were withdrawn just before the bank failed. A run by large depositors could trigger failure in other banks as well. Since the losses have been imposed on shareholders and bondholders, this is not a full-fledged bailout.
The experience of SVB shows yet again that the systemic importance of a bank cannot be fully determined in advance based on some static rule (say, the size of the balance sheet). Even a mid-sized bank like SVB can appear systemically important if its failure could trigger a systemwide crisis. The key point to understand is that it is impossible to do away with discretion, as the laws cannot be complete. Some discretion in the determination of systemic consequences cannot be done away with, and this is a challenge for public administration.
Since the global financial crisis of 2008, many countries have established resolution capabilities. India continues to lag on this front. Its effort to reform the resolution framework for the financial system failed in 2018 when the bill that was introduced for this reform was withdrawn. The Financial Stability Board’s “2021 Resolution Report” finds that India’s bank resolution regime is lacking in almost all the expected key attributes. As of now, India only has a deposit insurance mechanism that operates as a paybox, but not a specialized resolution mechanism. When a bank fails, the Reserve Bank of India and the union government come up with some resolution plan, which usually involves denying access to deposits for some time.
One probable reason for the complacency around this reform is that the main source of safety in India’s banking system is the fact that two-thirds of bank deposits rest with government-owned banks (including regional rural banks). Since it is widely assumed that the government would fully back these deposits, the chances of a run on these banks are very small. However, the cost for this is incurred in many ways, including the periodic recapitalization of these banks. Further, as the experience of Yes Bank showed, even large private banks can fail in India. In any case, several small banks fail every year in the country. Developing a specialized resolution capability is essential for the growth of banking in India. The ease of exit is essential for the ease of entry and increases the options available for policymakers when faced with a bank failure or a banking crisis. A well-developed resolution capability can reduce the costs of a banking crisis by ensuring early recognition of failure, quick and orderly resolution of the failed banks, and an almost immediate access to the insured deposits. All this can help maintain trust in the banking system.
When a crisis happens in a developed financial system, people in other jurisdictions may feel tempted to congratulate themselves. However, the overall efficiency of a system cannot be determined only by looking at its costs, which reveal themselves most clearly during such crises—a system’s benefits should also be considered before drawing any broad conclusions. After all, U.S. and Switzerland have two of the most developed financial systems in the world.
—By Suyash Rai
Review
Studying the Link Between Migration and Agricultural Production in India
India’s process of structural transformation is bound to affect the patterns of food production. As household members in rural areas migrate to urban areas in search of jobs and higher wages, what happens to the production of food? Do the remaining members of the household in rural areas replace this lost labor with hired labor or substitute it with technology? Or do they instead give up on agriculture? If it is the latter, what implications does this have for food production?
A recent working paper by Madhok, Noack, Mobarak, and Deschenes tries to answer these and other questions. Their paper titled “Rural-Urban Migration and the Re-organization of Agriculture” (2022) tries to understand the impact of rural-urban migration on agriculture and food production in India. According to the authors, while previous literature suggests that labor-losing areas modernize quickly, this is not true of India, where there is a “positive correlation between labour exit and crop production within districts. Districts experiencing labour loss also experience output contraction.”
Like many other transitioning economies, India has seen a gradual shift away from agriculture. Even though the agriculture sector employs the largest proportion of the country’s workforce, this proportion has reduced significantly over time. Madhok et al. state that aggregate employment in agriculture dropped by almost 20 percent between 1991 and 2011. However, there is little research to highlight the impact this shift has had on agricultural production, labor-substitution, and the spatial organization of agriculture.
In their paper, the authors first examine night-light data to show that crop production declines in labor-losing districts but increases in more remote areas that witness lower migration. They then use regression analysis to confirm this fact. They find that migration leads to a spatial reorganization of agricultural production—from labor-losing areas to more remote areas that witness lower production.
The authors first use a partial equilibrium model of household production to generate predictions about “technology responses to labor reallocation” in a fixed-price regime where migrant wages are completely remitted. They find that in response to higher urban wage potentials, households reallocate labor from agriculture to employment in urban areas.
In response to labor loss, households reduce technology use and contract farm sizes and output. This finding, as the authors note, is contrary to existing literature that posits agricultural modernization in labor-losing areas. But this decision to reduce technology use, according to them, is rational if technology use complements labor instead of substituting it.
They also find that declining output due to the loss of labor is offset partially through an indirect effect—declining land prices, which prompt farm expansion. They are also offset because of another indirect effect of migration and reduced production—increased crop prices, which incentivize other households whose members have not migrated to urban areas to increase production. As an aggregate consequence, agricultural production declines in areas with high migration and increases in areas with little or no migration.
These findings, as the authors note, show that structural transformations in the economy do not necessarily lead to a shift away from agriculture altogether. In India, they are resulting in a spatial shift towards more remote areas where migration is lower. The paper states that as per their analysis, “. . . the spatial reorganization of agriculture mitigated 61% of the aggregate food decline” and “. . . 48% of total crop value.”
One issue the paper does not address properly is the degree to which better provision of agricultural infrastructure affects agricultural production in more remote areas. In other words, production in remote areas may be increasing due to the extension of agricultural infrastructure to some extent and the indirect effects of migration from peri-urban areas to urban areas. The degree to which these other factors cause a spatial shift in agricultural production is not clear from the paper.
The paper also provides food for thought for subjects it does not directly focus on. While the research in this paper is based on Indian Human Development Survey’s data covering all of India, it would be interesting to know the redistributive impact of out-migration within specific states and districts. For example, states like Bihar are witnessing both outmigration as well as increased agricultural production. How do the authors then map the spatial redistribution of food production caused by outmigration in a state like Bihar? In which areas of Bihar does agricultural production shift due to migration from Bihar to other states?
In addition, the paper ends with a question worth consideration—what are the income consequences of agriculture shifting to more remote areas? For migrant households, it would be logical to assume that the act of migration provides higher incomes in comparison to the households whose only means of subsistence is agriculture. If agricultural production increases drastically in remote areas, does this lead to a reduction of poverty in these areas? In other words, is migration to peri-urban areas a mechanism for poverty reduction in more remote areas?
Finally, the paper draws attention to the fact that in India, migration leads to a reduction, rather than an increase, in the use of technology, because the technology used is labor complementary. It points to the fact that the consequence of reducing the pressure of labor in agriculture in peri-urban areas is that, rather than enabling households to invest more in labor substitution and modernization, it leads to a shift of agricultural production towards non-migrant households.
This highlights, first, the precarity of household incomes and the lack of available surplus for investing in technology that substitutes labor (as opposed to being labor complementary). It also highlights the possible economic utility of migration as an escape from the agricultural sector rather than a source of additional income to help expand agricultural production.
Overall, the paper provides an interesting analysis of the relationship between migration and agriculture in India and highlights the need for a more nuanced understanding of this relationship to support both migration and food security.
—By Anirudh Burman