[su_dropcap style=”flat”]I[/su_dropcap]N CHAPTER 2, we discussed a general business’s risk of failure. The terms “credit underwriting,” “risk assessment,” “risk analysis,” and “credit analysis” all describe the qualitative and quantitative reviews that experts perform to judge the likelihood of failure of a business entity. Failure, as we also discussed, is the inability to make contractual payments (most often for debt) as the payments become due.
This chapter focuses on credit analysis for banks. We address directly some features that distinguish banks from other corporate entities: government support; high leverage; and mismatched funding. Our presentation here is broad, generic, and qualitative in that we do not perform a complete review on specific banks such as Citigroup or (pre-bankruptcy) Lehman. Neither do we estimate a numerical failure likelihood (e.g., such as “10% failure probability over the following year”).
Rather, our goal here is to motivate and argue the summary view that banks are risky entities in the absence of government support. It is only the presence of real and implied government support that enables banks to operate at their typical leverage levels and with their typical funding practices. Breaking the banks’ dependence on government support and bailouts requires new choices for leverage and funding.
Let us first discuss the roles of leverage, funding, and asset quality for banking credit analysis assuming an absence of government support. This is the standard treatment of credit rating agencies’ approach to bank reviews. Next we overlay the implied government support. This bifurcation is sensible since it permits the analyst and her colleagues to debate and dissect the evolving views of government support.
Risk of High Leverage
Figure 5-1 shows the evident balance sheet leverage of Lehman less than four months prior to its bankruptcy filing in 2008. The debt-to-equity leverage exceeds 20 (i.e., $613 billion of debt divided by $26 billion of equity with reference to Lehman’s SEC 10-Q filing of May 2008 for the debt and equity values).
The picture of figure 5-1 shows the failure risk well even though the drawing is not truly to scale! A lender to Lehman in May 2008 might have worried that a drop in the $639 billion asset value of just 5% (about $32 billion) would render Lehman insolvent. Whether for Lehman in 2008 or any other bank today, the potential lender to the bank must consider the risk that a plunging asset value could create insolvency and the risk that the stated asset value may be inaccurate or uncertain even before a fall in value.
A prudent lender might choose to lend to a bank with leverage of, say, 10 or higher for a tenor of a year or longer if her analysis finds that (i) the bank assets have highly certain value and low volatility (i.e., low uncertainty of future value), (ii) balance sheet solvency is largely not relevant to the bank’s ability to repay debt, or (iii) the government or CB will somehow ensure debt repayment. Let’s ignore this last condition for now since the current exercise is to determine the bank’s failure likelihood in the absence of government support.
Regarding the relevance of balance sheet solvency, it is certainly conceivable that a business might have negative equity and yet have ample cash flow to pay obligations as they come due. Even aside from paying such obligations from its available resources, such an insolvent business can reliably persuade lenders to roll its debt if historical cash flow is consistently strong. Kellogg, Campbell Soup, and Clorox are examples from chapter 4 of this (unusual) type of business with strong and consistent cash flow overcoming near insolvency. But banks typically cannot muster this necessary consistency and also have the complication of excessive short-term debt concentration.
Regarding asset value and volatility, banks have the blessing and curse that their assets are overwhelmingly financial assets. The assets produce income – which is both good and readily understandable to the investor. But values of financial assets have widely varying certainty and volatility. Financial disclosure requirements of the U.S. SEC stipulate that banks must show the portion of assets that are “Level 3” (an accounting judgment that the certainty of asset value is low). In the Lehman example of figure 5-1, for example, the bank’s equity is $26 billion. But Lehman’s “Level 3” assets of May 2008 are $41 billion (from Lehman’s SEC 10-Q filing of May 2008). Without any further scrutiny of this bank’s assets, it’s fair to conclude that the asset value is uncertain relative to the distance from insolvency.
Risk of Funding Mismatch
Figure 4-2 of chapter 4 shows the preponderance of short-term funding for banks. A typical bank’s ratio of current liabilities (maturity of one year or less) to total liabilities is roughly 80%. For Lehman in the year prior to its failure in 2008, $200 billion of debt – almost one-third of total liabilities – matured every day. Hence, on every business day, Lehman employees would arrange the new financing (essentially refinancing) of as much as $200 billion.
The risk to all lenders is simply that the day may come when other lenders will not roll the short-term debt. Whether lending overnight (i.e., for 1-day maturity) or for thirty years (by buying a 30-year bond), the lender will suffer loss if short-term lenders do not roll the loans. How does the credit analyst judge or otherwise estimate the likelihood of the borrowing bank’s ability to roll its debt continuously? The first and perhaps most obvious reply is to reason that “if the bank can’t roll its funding and therefore defaults on its debt repayment, then liquidation of assets will repay in full all creditors.” But such logic applies only if the bank’s asset value comfortably exceeds the value of the bank’s liabilities. The high leverage of typical banks means that asset value is nearly the same as liability value which implies high likelihood that liquidation of assets cannot repay the liabilities.
The second attempted reply to gain comfort that a bank will always be able to roll its short-term funding is to review the bank’s history. It may well be that the bank has been operating with a similar funding profile for the past five years with no incident. Stated differently, all other lenders seem comfortable and the bank has been rolling its debt for many years. But this is earthquake risk. If one buys a house on an earthquake fault line, years and decades may pass without even a worrying tremor. Insolvency born of short-term funding is not just part of banking history, it is banking history. There’s a crass saying on Wall Street to the effect that “industrial companies die from cancer, banks die from heart attacks.”
One obvious encouragement for short-term lenders is that their exposure is short-term. An overnight lender really needs to be “right” that the bank will not fail for just one day of risk. In addition, overnight repo lenders have collateral in the form of a pledged debt or equity security. In truth, it is highly unlikely that these lenders would ever take a large loss. But the lending is unstable. The short-term lenders will protect themselves by choosing not to roll their loans when the bank reports poor results or is the subject of negative rumors. This refusal to roll becomes an immediate crisis for unsecured creditors, the shareholders, and the bank.
Returning to the pertinent question, how does the credit analyst judge the likelihood of such a funding crisis? In the absence of government support and lacking – due to high leverage and assets having much longer maturity than the funding – a clear indication that the bank’s resources can pay all liabilities, there just is no capability to make such a judgment. Short-term lenders wondering whether to roll their loans to a dubious bank look sideways at the other short-term lenders. If one lender conspicuously drops out, others may follow rapidly. It is not possible to quantify reliably this sudden and abrupt flight of “confidence.”
Our hypothetical exercise in which we ignore government support becomes questionable on the topic of bank deposits. Depositors are unsecured lenders who have the right to remove any portion of their funds without notice or with very short notice (such as one week) depending on deposit type and jurisdiction. Since most deposits in OECD (Organization for Economic Cooperation and Development) commercial banks enjoy insurance from the respective government, it is difficult to specify how depositors behave in the absence of this insurance. By historical inference, though, we conclude that depositors without government insurance would indeed remove their funds upon negative rumors or doubts of solvency.
Figure 5-2 Run on the Bank
Figure 5-2 shows “run on the bank” depositors waiting for their bank to open in the U.S. of the 1930’s. Banking “panics” recurred every ten to twenty years in the U.S. prior to the creation of the FDIC and establishment of deposit insurance in 1934. Contemporary records imply, however, that many observers did not consider deposit insurance to be an appropriate solution to banking problems leading into the 1930’s. The Canadian Department of Finance noted in the 1920’s that deposit insurance had failed within individual U.S. states and that the concept is unsound:
“[Deposit insurance] has proved unworkable in the United States and is basically unsound as it means that the conservative and properly operated bank would be called upon to bear the losses through mismanagement, fraud, and otherwise incurred by competitors over whom it has no control. The final outcome would be a disaster as the public would not be called upon to discriminate between sound institutions, with whom their funds would be safe, and the others.”
Writing in 1932, the soon-to-be U.S. president, Franklin Delano Roosevelt, also did not favor deposit insurance:
“[Deposit insurance] would lead to laxity in bank management and carelessness on the part of both banker and depositor. I believe that it would be an impossible drain on the Federal Treasury.”
Even with deposit insurance, though, depositors withdraw their funds as a bank spirals into failure. In the weeks before the collapses of the U.S. banks Washington Mutual (Washington Mutual, Inc.) and IndyMac (IndyMac Bank, F.S.B.) in 2008, for example, bank deposits fell sharply. The same flight of (insured) deposits hit Northern Rock (Northern Rock plc) before its failure in the U.K. in 2007 – though one explanation is that the insurance protected just 90% of funds above a threshold of £3,000.
It is not irrational for depositors to withdraw funds when they know deposit insurance protects them. They merely prefer the certainty of having cash out now rather than navigating a government process of reimbursement that may have burdensome validation and filing requirements and delays to say nothing of possible political wrangling (such as whether some depositors should receive preference over others). If there is no cost to removing cash now, it’s not surprising that deposit insurance fails to stem the bleeding of deposits when a bank is stressed.
The next excerpts for Chapter 5 will discuss “Critical Role of the Quality and Transparency of Assets,” “Government Support” and “Who Understands Banking Risk?”
 See the excellent and authoritative Lehman Brothers Holdings Inc. Chapter 11 Proceedings Examiner Report, published in 2010.
 See, for example, J. Grant, Money of the Mind: Borrowing and Lending in America from the Civil War to Michael Milken, Farrar Straus & Giroux, 1992.
 J. Carr, F. Mathewson, and N. Quigley, “Stability in the Absence of Deposit Insurance: The Canadian Banking System, 1890-1966,” Journal of Money, Credit and Banking 27, 1137-58, 1995.
 See page 139 of N. Prins, It Takes a Pillage: Behind the Bailouts, Bonuses, and Backroom Deals from Washington to Wall Street, John Wiley & Sons, 2009.
 See K. Grind, The Lost Bank, Simon & Schuster, 2012.