Banking On Failure: Junk Banks (Continued)

WORLDWORKS by Joe Pimbley

Editor’s Note: This is the next in a Continuing Series of Articles by Joe, as excerpted from his Book; “Banking on Failure” co-authored by Laurel McDevitt.

The previous excerpt for Chapter 11 explained why “Banks are Junk”

Government Support of Banks is Expensive

In the worlds of capital markets and investing, one frequently observes a bank or other institution guarantee the debt or payment obligations of another firm. Banks, for example, provide “letters of credit” on behalf of borrowers that obligate the banks to make the borrowers’ debt payments when the borrowers cannot do so themselves. There also exist Banking on Failure“bond insurers” that provide guarantees for municipal bonds.[1] Banks and bond insurers (“guarantors”) receive premium payments for providing the guarantees. The clear, common sense understanding of guarantors is that the risk of the guarantee is simply the risk of the “underlying” borrower or bond that receives the guarantee. When the borrower or bond defaults, the guarantor takes a direct loss equal to the amount it must pay. The guarantee of a high-risk borrower, therefore, is itself high-risk and the guarantor should receive a commensurately large premium payment.

We just concluded in the preceding section that “underlying risk” of banks is junk (i.e., non-investment-grade). Thus, governments’ support and guarantees for banks are expensive, even during periods of no direct loss payments, due to the high risk that the governments assume. When direct loss payments are necessary, they can be substantial. Federal taxpayers paid more than $130 billion to resolve failed U.S. savings and loan (“S&L”) banks in the 1980’s.[2] The U.S. government will ultimately lose more than $150 billion for the ongoing bailouts of Fannie Mae and Freddie Mac.[3] Ireland’s government paid €64 billion to bail out its financial system[4] – a huge amount relative to its recent annual GDP of €40 billion.[5]

Beyond the financial risk of the government guarantee and direct payments for losses, there are political and indirect economic costs. Voters and taxpayers despise government funding of banks. Regardless of any supposed merit to the argument that it is “practical” to violate the principle opposing government subsidy of private-sector firms, citizens of all political persuasions are outraged. Indirect financial costs include the so-called “financial repression” that governments impose through the central banks to keep interest rates low.[6] This repression benefits debtors at the expense of creditors and results in distorted and diminished economic activity.[7]

Ironically, a “successful bailout” does not truly solve core problems. When a government injects funds to restore depleted bank capital, the action serves only to restore the bank to its prior “junk” condition. The rescued bank will need the expensive government guarantee going forward. If instead of cash injection the regulator encourages a merger of the failing bank with a stronger bank, then the government merely concentrates its guarantee risk into a larger, also “junk,” institution.

Within this paradigm of “governments support and guarantee banks,” there’s a simple observation. As with any financial guarantee, the cost and risk decline when the “underlying borrower” has lower risk of failure. A paramount goal of bank regulation and oversight should be restructuring to create banks and a banking system that can survive and operate without government support. It is both ineffective and expensive to guarantee “junk banks.”

Why Not Let Banks Fail?

As we noted early in chapter 2, failure is a positive force in capitalist economies. Better businesses and better business practices emerge from a landscape in which good ideas win and bad ideas die. Even the dominant global bank regulator, the Basel Committee on Banking Supervision (BCBS), endorses “market discipline.”[8] So why not just let all banks fail if and when they cannot make contractual payments?

This question brings us back to our earlier “banks are different” dilemma. Though not supportive, we grant the face-value plausibility of what appears to be the majority view of regulators and bankers that government bailouts are preferable to failures for large banks. Let’s go through the best arguments for the “bailouts are better” opinion.

First, a government’s choice not to bail out a large, failing bank sends a screaming message to depositors, other investors, bankers, and short sellers that the game has changed. Previous assumptions no longer hold. Logic implies that guarantees are similarly unreliable for all other large banks. Thus, a government’s refusal to save one bank may put many banks “in play.” Recall our statement of chapter 5 from a rating agency bank analyst that “nobody really understands banks.” With this absence of analytical confidence, there is certainly a risk that investors will shun banks en masse with a government’s abrupt choice to end bailouts. We don’t recall anybody making this argument pre-Lehman, but it certainly resonates post-Lehman.

The second pro-bailout argument is that it would be catastrophic to lose the “payment system” that banking provides to the entire economy. Whether for individuals or businesses, the unquestioned, unhindered ability to make payments electronically or by card or check or currency is certainly crucial. Loss of the payment system is akin to loss of all electricity service and running water. The weakness of this point is simply that it assumes a failed bank would halt its payment functions voluntarily or that counterparty banks would halt their payments to the failing bank. Such interruption need not necessarily occur – especially if the failure event consists of just a single bank.

A third argument favoring bailouts for large banks is that the government potentially faces larger losses from deposit insurance claims of several simultaneously failing banks than it would from “propping up” the weak banks via bailouts. Conveniently, the “propping up” of banks may appear cheap in some cases. A typical bank may have just 10% equity relative to total assets. If the bank loses this entire equity amount, the government can pay the 10% of assets as a bailout and thereby “prove” to the market that it will not let banks fail. The government will then compel the rescued bank to repay the bailout over, say, a 5-year period and then tell the citizens “the bailout didn’t cost anything.”

By letting the bank fail and honoring deposit insurance commitments, however, the government could easily lose 20% of bank assets with no chance of recouping the loss.[9] This comparison is certainly “unfair” because the no-loss conclusion of the bailout case makes the rosy assumptions that the bank will survive, pay back the bailout funds, and never make a claim on the deposit insurance fund. Unfortunately, such optimistic assumptions are typical for “policy analysis” supporting political decision making.

Systemic Risk

These arguments favoring bailouts of large banks have plausible aspects, as we say. Yet the end result that banks will seemingly not fail regardless of malfeasance, incompetence, lassitude, or for any number of deserving reasons due to taxpayer bailout is unacceptable. It is not just “fear of failure,” but also actual failure, that prods organizations to minimize destructive practices such as political in-fighting, paying unreasonable compensation, building bloated bureaucracies, scrounging and lobbying for government subsidies and protections, et cetera.

Ironic and instructive is the recollection that bank regulators have been wringing their hands for years over “systemic risk” to the financial world with vague statements that derivative transactions and other “interconnectedness” are sources of instability.[10] We find it most likely that systemic risk – the possibility of bank failures clustering together rather than being independent so that the entire “banking system” is at risk – arises from actions of governments and regulators. One example is the drama of Lehman’s failure in 2008. This failure demonstrated that government bailout support was uncertain and thus rattled and weakened all banks. Another example is a lesson of Calomiris and Haber:[11]

“[G]overnment safety nets tend to destabilize banking systems, and such safety nets tend to arise not for reasons of economic efficiency but because they are outcomes of political bargains.”

Lehman aside, the bank regulatory regime imposes standard rules on all banks and thereby forces banks to measure risks similarly and to take essentially the same risk positions. This conformity-inducing regulation creates high correlation (“systemic risk!”) among banks.[12] When one particular asset class for which regulators assign low “risk weight,” such as investment-grade sovereign debt or triple-A tranches of subprime mortgage securitizations, fails, every bank holding large positions in the troubled asset class will take losses simultaneously.

Dowd supports this common regulatory regime argument forcefully and lucidly:[13]

“Any system of regulatory capital modeling inevitably means that the regulators have a preferred model of their own and then pressure regulated institutions to adopt similar models. Regulated banks’ risk models will then converge to something similar to the Fed’s own models. The end result is that banks will have much the same models and much the same risk management strategies. They will therefore take much the same risks and make much the same mistakes – dramatically magnifying systemic risk. This is exactly what happened with subprime.”

[1] J. M. Pimbley, “Bond Insurers,” J. Appl. Fin. 22(1), 36-43, April 2012.

[2] See “The Savings and Loan Crisis and its Relationship to Banking,” chapter 4 in An Examination of the Banking Crises of the 1980s and Early 1990s, 167-188.

[3] N. E. Weiss, “Fannie Mae’s and Freddie Mac’s Financial Status: Frequently Asked Questions,” Congressional Research Service, R42760, August 2013. Though not banks, Fannie and Freddie are good examples for this discussion since they are financial institutions with (prior to 2008) implied support of the U.S. government.

[4] See, for example, “Government was told in 2008 bank bailout would cost €16.4 bn,” Irish Times, March 20, 2014.

[5] See . As reference, the U.S. GDP is roughly $15 trillion, so a Fannie/Freddie loss of $150 billion, though painful enough, is just 1% of GDP.

[6] See, for example, .

[7] Over an entire economy, debtors and creditors balance. Since government is a large net debtor, the private sector is a large net creditor. The private sector employs capital far more effectively than government. Hence, financial repression has negative impact.

[8] See, for example, several points within “Basel III: A global regulatory framework for more resilient banks and banking systems,” Bank for International Settlements, June 2011.

[9] To reach this 20% of assets rough estimate, we note that a large bank’s deposit liabilities are in the range of 50% of assets. See, for example, the balance sheet of the fiscal year 2013 SEC 10-K filing of Bank of America Corporation in which assets for end-of-year 2013 are $2.1 trillion, shareholder equity is $233 billion, and deposits are $1.1 trillion. We estimate the FDIC loss fraction on its deposit disbursements to be 35% based on Table A-1 of “A Brief History of Deposit Insurance in the United States,” International Conference on Deposit Insurance, Washington, D.C., 1998. Multiplying 50% by 35% and rounding up a bit gives 20%.

[10] See, for example, J. M. Pimbley, “My Perspective on Bank Regulation,” Financial Engineering News, January-February 2007.

[11] See chapter 14 of C. W. Calomiris and S.H. Haber, Fragile by Design – The Political Origins of Banking Crises and Scarce Credit, Princeton University Press, 2014.

[12] See J. M. Pimbley, “Bond Insurers,” J. Appl. Fin. 22(1), 36-43, April 2012. A widespread misunderstanding is that derivative trading among banks is a dominant source of systemic risk. Based on deep understanding of derivative trading among banks as well as the history of such trading, we disagree. See J. M. Pimbley, “My Perspective on Bank Regulation,” Financial Engineering News, January-February 2007; and J.M. Pimbley, “Collateral damage,” Credit, November 2008. This latter article also explains why the AIG experience is not a valid argument that “derivatives cause systemic risk.”

[13] See page 15 of K. Dowd, “Math Gone Mad,” CATO Institute 754, 1-64, September 3, 2014.


Joe Pimbley
Joe Pimbley
Joe is Principal of Maxwell Consulting, a firm he founded in 2010. Joe is expert in complex financial instruments, financial risk management (certified as FRM by the Global Association of Risk Professionals), valuation, structured products, derivatives, and quantitative algorithms. His recent and current engagements include financial risk management advisory, valuation and credit underwriting for structured and other financial instruments, and litigation testimony and consultation. In a prominent engagement from 2009 to 2010, Joe served as a lead investigator for the Examiner appointed by the Lehman bankruptcy court to resolve numerous issues pertaining to history’s largest bankruptcy. Joe and his colleagues discovered Repo 105 and also reported the critical importance of pledged collateral mishaps and mischaracterizations to the Lehman failure. Joe holds a Ph.D. in Theoretical Physics and is a co-author of Banking on Failure (2014), Simple Money (2013), and Advanced CMOS Process Technology (1989). He serves on several corporate and academic Boards, has written more than thirty finance articles, presented more than sixty finance seminars, and holds numerous patents for engineering inventions.

SOLD OUT! JOIN OUR WAITING LIST! It's not a virtual event. It's not a conference. It's not a seminar, a meeting, or a symposium. It's not about attracting a big crowd. It's not about making a profit, but rather about making a real difference. LEARN MORE HERE