Overview of U.S. Capital Ratios

The data shown in Exhibit 19.1 shows the ratio of equity capital to assets of U.S. banks during the 1896–2007 period. Equity capital, which is the book value of assets less the book value of liabilities, is different than regulatory capital that can include subordinated debt and some adjustments for off-balance sheet items. It also differs from economic capital, which is a statistical estimate of risk and capital that will be discussed shortly.

As shown in Exhibit 19.1, banks in the United States had equity/asset ratios of 23.5 percent in 1896. The ratios in 1896 and 1900 reflect a time when many banks were operating under the “real-bills doctrine”—borrowing short term and lending short term.

The equity/asset ratio gradually declined to less 5.8 percent in 1980. Over the years, bankers expanded their lending horizons and made longer-term loans, including real estate loans. They were still borrowing short term, but the longer-term loans increased their risk. During the 1985–1992 period, 1,373 banks that were insured by the Federal Deposit Insurance Corporation (FDIC) failed.³ In addition 1,073 savings and loan associations and 1,707 credit unions failed. All were federally insured. Thus, there was pressure in the United States for increased regulations dealing with bank capital. The end result was the passage of the Federal Deposit Insurance Corporation Improvement Act of 1991 (FIDICIA), which increased bank capital requirements. FIDICA included Prompt Correct Action (PCA) rules of how to deal with undercapitalized banks that have risk-based capital ratios of 6 percent or less. Well-capitalized banks have risk-based capital ratios of 10 percent or more.

Two additional factors contributed to increased growth opportunities and risks for banks. The first factor is the laws that allowed simple commercial banks to form multibank holding companies, and then expand into underwriting securities, insurance, merchant banking, insurance, and other complementary activities.⁴ This opened the door for banks to become Large Complex Banking Organizations (LCBOs). Second was the growth of securitization in mortgage lending. Securitization is the packaging and selling pools of mortgage loans to investors.⁵ Securitization can involve complex structures such as Mortgage Backed Securities (MBS), Collateralized Debt Obligations (CDOs), and Structured Investment Vehicles (SIVs) backed by pools of MBS and CDO bonds.

Exhibit 19.1 also shows that nonfinancial corporations had equity/asset ratios that ranged from 69 percent to 35 percent. This is substantially greater than the banks’ equity/asset ratios for several reasons. One reason why banks have lower equity/asset ratios is that they are regulated by federal and state agencies and subject to various laws such as the Federal Deposit Insurance Corporation Insurance Act or 1991 FIDICIA that requires bank regulators to take prompt corrective actions if a bank’s risk-based capital falls below predetermined levels. Risk-based capital ratio refers to a percentage of a bank’s risk-weighted assets (e.g., loans) to its capital accounts. Well-capitalized banks have risk-based capital ratios of 10 percent or more. Undercapitalized banks have ratios of 6 percent or less. Other reasons include access to the Federal Reserve’s discount window, government intervention, and the Too-Big-To-Fail doctrine.⁶

Basel I

On the international scene, the Basel Committee on Banking Supervision was established in 1974. It focused on facilitating and enhancing information sharing and cooperation among bank regulators, and developing principles for the supervision of internationally active large banks. Following the large losses from the less-developed countries (LDC) in the late 1970s, the Basel Committee became increasingly concerned about the failure of large banks and cross-border contagion. In particular, they were concerned that large banks did not have adequate capital in relation to the risks they were assuming. In the 1980s, their concerns were directed at Japanese banks that were expanding globally. The end result was a uniform one-size-fits-all 8 percent capital requirement that became known as the 1988 Capital Accord, or Basel I.

Under Basel I, bank capital consisted of two tiers. Tier 1 includes shareholder equity and retained earnings, and it is 4 percent. Tier 2 includes additional internal and external funds available to the bank and also is 4 percent.⁷ Thus, Basel I required 8 percent risk adjusted capital.

Basel I focused primarily on credit risk, and risk-weighted assets ranged from 0 percent weight for claims on Organization for Economic Cooperation and Development (OECD) central banks and governments to 100 percent weights for commercial and consumer loans and loans to non-OECD governments.

Along this line, banks were required to hold more capital against ordinary mortgages than against pools of mortgages that were securitized. Therefore, banks began to change the way they did business from holding the mortgage loans to securitizing them and selling them to other investors. While banks continued to make mortgages and other loans, the securitization process allowed them to shift the risk to the investors who bought the securitized loans.⁸

The 8 percent risk-based capital ratio is an arbitrary ratio that is used to monitor risk. The 8 percent “Minimum capital is a guidepost.… It was not and is not intended as a level toward which the firms should aim nor as a standard for internal risk management.”⁹ It does not measure risk. Equally important, a large number of failed banks had capital ratios in excess of 8 percent shortly before failure. According to a FDIC study, 26 percent of the 1,600 U.S. banks that failed between 1980 and 1994 had CAMEL ratings (capital, asset quality, management, earnings, and liquidity) of 1 or 2, one year before failure.¹⁰ CAMEL ratings are used by bank regulators to evaluate banks. The ratings range from a high of 1 to a low of 5.¹¹ The study went on to say that “… bank capital positions are poor predictors of failure several years before the fact.”¹²

In 2007, all FDIC insured commercial banks in the United States held an average of 12.23 percent risk-based capital, far in excess of the 8 percent regulatory capital required by Basel I.¹³ The smallest banks (assets of less than $100 million) held 19.84 percent risk-based capital, while the largest banks (greater than $10 billion) held 11.86 percent. The holding of capital in excess of regulatory requirements is due in part to FIDICA, higher earnings, goodwill due to mergers, and to take advantage of growth opportunities.¹⁴

From 1980 to 1996, 133 of the International Monetary Fund’s (IMF) 181 member countries experienced significant banking sector problems.¹⁵ As a general rule, bank failures tend to be in large numbers, they are frequently associated with financial shocks (e.g., foreign exchange), and real estate defaults is the most common cause of bank failures.

In the early 1980s, Chile also experienced systemic banking problems. Falling copper prices, a severe recession, rising interest rates in the United States, and the 90 percent decline of the peso adversely affected foreign exchange–linked loans to domestic borrowers. The Central Bank in Chile took over 14 of the 26 commercial banks and 8 of the 17 domestic finance companies. Eight of the banks and all of the finance companies were liquidated.¹⁶

The bottom line about Basel I is that bank capital matters. However, a number of changes occurred that undermined Basel I. These changes include developments in derivatives, globalization, and the consolidation of LCBOs. Equally important, the basic business model of commercial banks has shifted from the real bills doctrine (borrowing short term and lending short term) to borrowing short term and lending long term (i.e., buy and hold), and more recently to borrowing short term and selling assets (i.e., originate and distribute to other investors through syndications, securitization, and credit derivatives).¹⁷ Along this line, Federal Reserve Governor Susan Schmidt Bies (2007) said,

U.S. supervisors support the 2005 Basel/International Organization of Securities Commissions’ (IOSCO) revisions to the 1996 Market Risk Amendment (MRA). Since adoption of the MRA, banks’ trading activities have become more sophisticated and have given rise to a wider range of risks that are not easily captured in the existing value-at-risk (VaR) models used in many banks. For example, banks are now including more products related to credit risk, such as credit-default swaps and tranches of collateralized debt obligations, in their trading books. These products can create default risks that are not captured well by the methodologies required under the current MRA rule—which specifies a ten-day holding period and a 99 percent confidence interval—thereby creating potential arbitrage opportunities between the banking book and the trading book.

In a nutshell, former Federal Reserve Vice Chairman Ferguson (2003) said that “Basel I is too simplistic to address the activities of our most complex banking institutions.” It is not sufficiently risk sensitive. Thus, Ferguson supported Basel II that was proposed by the Basel Committee on Banking Supervision in 2001.

Basel II

Basel II¹⁸ is a work in progress. It is an attempt to align regulatory capital with the risks that banks face. There are two distinctly different but related aspects of Basel II. One involves the three pillars, and the other involves enterprise risk management. Each is discussed here.

• Pillar 1: Minimum Capital Requirements—The regulatory capital requirements are based on credit risk (defaults by counterparty), market risk (price changes—on- and off-balance sheet), and operational risk (failed processes, people, systems, events). The definition of total capital in Basel II is the same as in Basel I. Total capital divided by credit risk, market risk, and operational risk must be equal to or greater than 8 percent.
  (19.1)
• Pillar 2: Supervisory review process—Foster supervisor-bank dialogue on risk management.
• Pillar 3: Market discipline—Based on disclosure of information.
• Without going into details, banks quickly discovered that it is relatively easy to get around the Basel II capital requirements.¹⁹ There are three ways to compute the capital requirements—(1) the Standardized Approach, (2) the Foundation internal rating based (IRB) Approach, and (3) the Advanced IRB Approach.

The data shown in Exhibit 19.2 shows that the minimum capital for a $100 commercial loan can vary from $1.81 to $41.65 depending on the credit risk and the approach used to calculate the required capital.

In addition, Exhibit 19.3 illustrates regulatory arbitrage for two $1 million loans where the Expected Loss (EL) is equal to the Probability of Default (PD) times the Loss Given Default (LD). Although the PD and LD of the two loans vary widely, they both have the same capital charge.

Banks can avoid charge-offs by restructuring loans that may become nonperforming loans. Alternatively, they can make a second loan to the obligor that would cover the payments of the first loan and keep it from defaulting. Finally, they can securitize loans—get them off the balance sheet.

Several Quantitative Impact Studies (QIS) have been run in order to test various aspects of Basel II. The FDIC’s View of the Fourth Quantitative Impact Study-QIS-4 said that “The results of QIS-4 show Basel II would most likely lead to an unacceptably large decline in capital for the largest banks.… Competing head to head with large banks, holding in some cases a fraction of the capital that non–Basel II banks hold on the same loan portfolio, would be a daunting challenge for the nation’s community banks.”²⁰ There was a 31 percent (median) reduction in Tier 1 Capital.²¹ This is not conducive to a level playing field for all banks in general, and community banks in particular. However, as previously noted, small community banks in the United States held capital far in excess of minimum required levels of regulatory capital, and it does not seem to have hurt their competitive positions.

Nevertheless, in July 2007, U.S. Federal Banking Agencies reached an agreement to implement Basel II.²² It will be tested over a three-year transitional period, and it will allow a cumulative capital reduction of no more than 15 percent. The primary impact will be on the LCBOs and other large banks that choose to use “opt in.” Smaller banks will be using the “standardized” approach. In June 2008, the FDIC approved the standardized approach for all banks except the largest, most complex banks that are subject to the advanced approaches.²³

Federal Reserve Governor Kroszner (2007) observed that while Basel I is based on “rules,” Basel II is based on “principles.” He said that “Taking a more principles-based approach means that we must allow bankers some flexibility in meeting the requirements and permit a reasonable amount of diversity of practices across banking organizations.” In other words, the capital requirements for the same loan may vary from bank to bank.

Another complicating factor is adoption of Fair Value Accounting for valuing bank assets, liabilities, and certain financial instruments.²⁴ Barth (2004), and Gup and Lutton (2008) point out that there is added volatility of assets and liabilities associated with fair value accounting. The volatility could have a positive or negative effect on bank capital adequacy.

FDIC Chairman Sheila Bair (2008) in a speech about Basel II and risk management said that there was a major lack of transparency in structured finance (i.e., Collateralized Debt Obligations, CDOs). She said that “The advanced approaches in general represent a heavy bet on the accuracy of models and quantitative risk metrics.” The unintended consequence is that the Basel II framework results in lower capital requirements for most credit classes with a favorable loss history. “And this can encourage banks to lever up … to boost their return on equity.” Therefore, she says that the “advanced approaches can be far off the mark. Now (there is) widespread recognition that there is more to sound risk management than mathematical formulas….” She favors a simple “leverage ratio.” In testimony before the U.S. Senate, she argued that “The leverage ratio complements the risk based capital requirements by ensuring a base level of capital exists to absorb losses … even in situations where risk-based metrics erroneously indicate risk is minimal and little capital is needed. These safeguards, along with the Prompt Corrective Action framework … will preserve capital and promote a safe and sound banking system …”²⁵

In July 2008, an interagency statement was issued concerning the “U.S. Implementation of Basel II Advanced Approaches” for selected banks and bank holding companies.²⁶ The statement said that banks and bank holding companies planning to operate under the advanced approaches must follow certain procedures that will lead to their implementation no later than April 1, 2011.

Enterprise Risk Management (ERM) and Economic Capital

Basel II must be considered in the context of enterprise risk management (ERM). The Committee on Sponsoring Organizations (COSO) defines ERM as “a process affected by an enterprise’s board of directors, management, and other personnel that is applied across an enterprise that is used to identify, assess, and manage risks within its risk appetite, to provide reasonable assurance of achieving its objectives.”²⁷

Federal Reserve Governor Susan Schmidt Bies (2006) commented on the COSO definition of ERM, and she said that it can mean different things to different people, but “all banking organizations need good risk management. An enterprise-wide approach is appropriate for setting objectives across the organization, instilling an enterprise-wide culture, and ensuring that key activities and risks are being monitored regularly.”

The key point here is that ERM is forward looking. It takes into account economic conditions and a wide range of risks and other factors affecting banks in the future. Regulatory capital is history—not the future.

ERM employs the concept of economic capital—a statistical concept that measures risk, and it reflects the bank’s estimate of the amount of capital needed to support its risk-taking activities. It is not the amount of regulatory capital held.²⁸

A study of Risk Based Capital by the Government Accountability Office (GAO) found that “although the advanced approaches of Basel II aim to more closely align regulatory and economic capital, the two differ in significant ways, including in their fundamental purpose, scope, and consideration of certain assumptions. Given these differences, regulatory and economic capital are not intended to be equivalent.… Economic capital models may explicitly measure a broader range of risks, while regulatory capital as proposed in Basel II will explicitly measure only credit, operational, and where relevant, market risks.”²⁹ Thus, economic capital reflects the bank’s estimate of the amount of capital (not book value capital or regulatory capital) needed to support its risk-taking activities. In statistical terms, it is a conditional random variable. In practical terms, some large banks use a Return on Risk Capital (RORC) that is related to Economic Capital in making lending and investment decisions throughout the organization.

In the context of ERM, risks for global banks go far beyond credit risk, market risk, and operational risk. They include, but are not limited to, breakdown of critical infrastructure, changing laws and regulations, changes in technology, defaults of sovereign debts, hurricanes, oil prices, terrorism, political instability, and other factors. Although some of these risks may seem remote, they have happened in the past and may happen again. For example, there have been sovereign defaults in Latin America/Caribbean dating back to the early 1800s. Chile, for example, defaulted on government debts in 1826, 1880, 1931, and 1983. The most recent sovereign default in that region was in Dominica in 2003.³⁰

Risk is measured in terms of probability, expected impact, and standard errors. Thus, Economic Capital is the difference between a given percentile of a loss distribution and the expected loss. It is sometimes referred to as the unexpected loss at the 99.97 percent confidence level. That means a 3 in 10,000 probability of the bank becoming insolvent during the next 12 months. It is important to recognize that measures of economic capital will vary from bank to bank and over time as conditions change. Exhibit 19.4 illustrates the concept of economic capital.

In March 2007, the Basel Committee on Banking Supervision gave its Risk Management and Modeling Group the mandate to assess the range of practices for measuring economic capital.³¹ The areas of potential emphasis include:

• New measurement approaches for credit risk.
• Diversification effects.
• Complex counterparty credit risks.
• Interest rate risk.
• Firms’ approaches to validation of internal capital assessments.³²

In theory, economic capital sounds good. In practice,

the Federal Reserve conducted a review across a number of large banking organizations to assess these firms’ use of so-called “economic capital” practices, which are a means for firms to calculate, for internal purposes, their capital needs given their risk profile. Consistent with other findings, we found that some banks relied too extensively on the output of internal models, not viewing model output with appropriate skepticism. Models are dependent on the data used to construct them. When data histories are short or are drawn mostly from periods of benign economic conditions, model results may not be fully applicable to an institution’s risk profile. We concluded that banks would generally benefit from better evaluation of inputs used in their internal capital models, stronger validation of their models, and broader use of stress testing and scenario analysis to supplement the inherent limitations of their models.³³