The focus of this book is on how bank lending and liquidity flows interact with the wider economy. So long as credit losses remain at acceptable levels, they do not interrupt those flows. Given that the focus is elsewhere, this section will just offer a high-level perspective on how banks manage credit risk, without attempting to discuss what strategies individual banks use to analyse credit risk. Although this section will mainly refer to lending decisions, liquidity provision to capital markets participants also requires credit risk analysis.
Banks aim to limit their credit losses with a multi-layered strategy. The layers are listed next, which will then be described below.
Internal controls for fraud and incompetence by loan officers.
Make good loans based on income or collateral.
Get guarantees for loans.
Get loans off the balance sheet.
Have good loss mitigation practices on troubled loans.
Limit concentration risk.
Internal Controls
The first line of defence for a bank is to have internal processes and examiners that contain the risks created by incompetent or fraudulent lending officers. Once a bank is large enough to have multiple branches, bank management needs to delegate lending authority to loan officers. The usual structure is to make the system hierarchical, with the size of loans that can be made determined by the level within the hierarchy. This hierarchical structure means that low-level loan officers largely work on a quota basis, and do not interact with the teams managing the overall strategy of the bank.
The concern of top management is that junior loan officers are either unable to judge risk properly, may wish to understate risks to create a larger loan volume, or might even conspire with outsiders to defraud the bank.
An even deadlier risk is when top management of the bank themselves defraud the bank. This is described in Bill Black’s book The Best Way to Rob a Bank is to Own One: How Corporate Executives and Politicians Looted the S&L Industry. “S&L” refers to Savings and Loan banks, which were the small American banks caught up in a banking crisis that culminated in the early 1990s. The root of the crisis was that these banks were hit hard by the high interest rates brought in by Fed Chairman Volcker, as they historically operated in a regime where interest rates were regulated. The push to deregulation (often described as neoliberalism) led to attempts to cover up the interest rate losses via letting the S&L’s take more credit risk. Although this stratagem might have worked, it also coincided with a change that allowed individuals to own S&L banks. Free market theorists refused to believe that the owners of firm might deliberately run it into the ground while lining their own pockets, and that is exactly what happened.
Make Good Loans
The simple way to eliminate credit risk is to not make any bad loans. Although some bond investors can operate in this fashion (lend to governments, quasi-governments, entities guaranteed by governments, and highly capitalised banks that are too big to fail), most banks need to deal with riskier clients. Banks need to take risks in lending. How a bank does that is a core part of its business strategy.
One traditional way of looking at the riskiness of loan is to look at the “4 C’s of Credit”: capacity, collateral, covenants, and character. Capacity refers to the borrower having the income flows to support the loan. Collateral provides a second line of defence against credit losses in a liquidation. Covenants refer to the legal encumbrances within a lending contract placed upon the borrower that are meant to protect the lender from adverse transactions by the borrower. (Although tighter covenants might seem superficially better, they also raise the odds of a bankruptcy if they constrain the borrower’s financial freedom of action too greatly.) Character is the willingness and desire of the borrower to pay back the loan, which can be hard to measure based on quantitative scores. (Credit scores look at past payment behaviour to measure “character,” but it is entirely possible for someone to diligently pay their credit card bills then decide to flee to an extradition-proof country after running a major fraudulent transaction.)
Lending against capacity and collateral are the strategies that are amenable to quantitative analysis.
Lending against capacity looks at measures like the ratio of the debt outstanding or interest payments versus the income of the borrower. Although the borrower’s assets and any collateral for the loan are looked at, the hope is that the income stream will allow repayment of the debt without any need to seize collateral. The obvious risk to such loans is a drop in income.
Lending against collateral allows lenders to take more risks against the income stream of the borrower on the basis that the value of the collateral will keep credit losses limited. Liquidity provision in capital markets relies heavily on collateralisation. The most famous example of collateral-based lending was subprime lending in the United States. (American residential mortgages are divided between higher quality prime loans that qualify for securitisation by the Government Sponsored Agencies (GSE’s), and sub-prime loans that did not.) The theory was that since home prices had not fallen at the national level since World War II, a geographically diversified securitisation pool of subprime mortgages would be relatively safe. (And even safer once the financial engineers used the magic of the Gaussian copula to create Collateralised Debt Obligations (CDOs).) The passage of time revealed that these pools were not in fact safe.
Loan Guarantees
If the original borrower looks somewhat shaky, the risk can be greatly reduced by getting another entity to guarantee the loan. This provides two layers of protection to the lender: if the original borrower is unable to service the loan, the lender can go after the entity that provided the guarantee.
Loan guarantors can come from all sectors of the economy. Courtesy of the Canadian housing bubble, young borrowers relying on a guarantee from The Bank of Mom and Dad on their mortgage has become common. At the corporate level, firms might guarantee other firm’s debts as part of a partnership. Within the financial markets, credit default swaps (CDS) allowed a variety of financial market entities to offer credit protection for target borrowers. Governments guarantee a wide variety of loans as a way of expressing policy objectives without the government spending money. In Canada, the housing market is mainly backstopped by the mortgage insurance provided by the Canada Mortgage and Housing Corporation (CMHC), which is a full faith and credit obligation of The Government of Canada.
The CDS market offered a way to deal with credit risk, but the problems started once the finance academics went to town. Prior to 2008, it was the era of hedge funds consisting of two people with a financial pricing services terminal trading large amounts of CDS based on hare-brained theories using spreadsheets with formulae using gaussian copulas. Once the financial crisis hit, people discovered that credit guarantees by entities with non-robust balance sheets are not that valuable. (Admittedly, some large firms with robust balance sheets also got into trouble.) It is possible for a financial system to sell credit insurance against a small number of weak links, but the aggregate system cannot self-insure, which is what effectively happened in 2008.
Get Loans of the Balance Sheet
Once again, getting loans off the bank balance sheet is the secret sauce for risk reduction. If you do not own it, you cannot take credit risk losses on it.
Loss Mitigation
Banks must have the capacity to work with troubled borrowers and find the way to best reduce the losses when restructuring a loan.
Having a dedicated credit loss mitigation team sets banks apart from most fixed income investors. The bulk of invested bond funds are in bonds that have a high enough credit rating to qualify for “investment grade” indices. Such bonds are not supposed to immediately default. (To the embarrassment of credit rating agencies, it does happen.) Since their bonds are not supposed to default, there is no need to keep an expensive legal team on staff to deal with debt restructurings. Since bond investing is a bulk business with razor thin expense margins, firms are very happy to avoid that expense.
The way such firms deal with weakening credits is to sell the bonds either when they are downgraded or they think the bonds are about to be downgraded. The bond prices will adjust much lower and will generally be picked up by “high yield” (investors who own bonds with ratings below investment grade) or even distressed debt investors. Distressed debt investors specialise in participating in debt restructurings, and the bond yields are high enough to allow management fees that cover the expenses of a dedicated legal team.
Concentration Risk
The last line of defence for a bank’s capital is attempting to ensure that it has diversified its credit exposures. There are three components to such diversification. Firstly, it needs to avoid having large exposures to a single borrower. (In this context, we determine whether a loan is large enough by comparing it to the size of the bank’s equity or capital.) Secondly, the bank should attempt to diversify the types of loans it makes. Finally, it ideally can diversify its loans across geographical regions. (This is difficult for small banks or banks in postage stamp-sized countries.)
Concentration risk can be easily quantified and is thus a natural target for regulators. For example, see the discussion of concentration risk by The Office of the Comptroller of the Currency (an American regulator) in the references. The problem with quantifying concentration risk is determining which credit risks are highly correlated (if one borrower defaults due to an event in the economy, the other is also exposed to the same event). For small loans, there are enough default events that it is likely that banks can develop somewhat reliable rules of thumb. Larger loans and novel forms of lending will not have a large data set to work with.
Recessions and Asset Market Busts
Recessions (and real estate busts) are worrisome for banks as there is coordinated drop in income flows across the economy. Instead of bankruptcies being isolated events that can be dealt with statistically, failures are widespread. Central bankers are typically forced to apply bandages to the banking system and hope that a recovery will take the pressure off the banking system. Since it takes time for credit losses to be realised, banks can stumble along in a half-dead state for some time.
A major problem that banks (and bank regulators) face is that they necessarily deal with the system from a micro perspective: they are looking at individual borrowers, and their own credit book. Although they can obviously develop an intuition about the trends of the business cycle, they have little choice but to ride the wave of capitalist expansion. A bank that has too conservative lending practices will lose market share to more aggressive competitors. This is not a sustainable commercial stance in an environment where financial crises are separated by decades.
The economist Hyman Minsky described what he called the Financial Instability Hypothesis: that a stable financial system encourages greater and greater risk taking, which eventually undermines the stability of the financial system. The system is then hit by a major financial crisis (e.g., The Great Depression of the 1930s, the 2008 Financial Crisis) which forces more conservative behaviour on regulators and financial market participants. This stabilises the system – until the “animal spirits” return. I already described this in one of my earlier books (Recessions: Volume I), and I point interested readers either to that book, or the book by Minsky in the references.
Too Big To Fail
A "sound" banker, alas! is not one who foresees danger and avoids it, but one who, when he is ruined, is ruined in a conventional and orthodox way along with his fellows, so that no one can really blame him. (John Maynard Keynes, The Consequences to the Banks of the Collapse of Money Values Aug. 1931)
The final important macroeconomic point to note is that banks herd together and tend to all blow up at the same time (as noted in the Keynes quote above). The advantage of this behaviour is obvious – regulators and central bankers do not have the freedom of action to completely shut down the economy as a result of shuttering several major banks. The banking system in aggregate is “too big to fail,” and so bankers can hope for a bailout in a system-wide crisis.
The way to mitigate this tendency is for regulators to toughen standards and be vigilant for excessive exuberance by lenders and borrowers. At the time of writing, the Financial Crisis of 2008 is still fresh in memories, and practices have been tightened. The question remains when will the siren call of deregulation to “unchain capitalism” strike again?
Concluding Remarks
Credit risk is the main killer of banks – it takes considerable incompetence to kill a bank solely based on interest rate risk. (Admittedly, the best and brightest in Silicon Valley did demonstrate that it was possible to do so in 2023.) The ultimate failure is most likely going to feature a liquidity event, but that liquidity event is the result of funding sources drying up in response to the prospect of the bank being insolvent because of credit losses. Insolvency causes illiquidity.
References and Further Reading
Black, William K. The best way to rob a bank is to own one: How corporate executives and politicians looted the S&L industry. University of Texas Press, 2013.
Comptrollers Handbook: Concentrations of Credit. Office of the Comptroller of the Currency (United States regulator). URL: https://www.occ.gov/publications-and-resources/publications/comptrollers-handbook/files/concentrations-of-credit/pub-ch-concentrations.pdf
Minsky, Hyman. Can it happen again?: Essays on instability and finance. Routledge, 2016.
No comments:
Post a Comment
Note: Posts are manually moderated, with a varying delay. Some disappear.
The comment section here is largely dead. My Substack or Twitter are better places to have a conversation.
Given that this is largely a backup way to reach me, I am going to reject posts that annoy me. Please post lengthy essays elsewhere.