When it Comes to Capital, Definitional Issues Loom Large
Click here for a printer-friendly version of this article
The financial regulatory reform bill and the new standards proposed by the Basel Committee on Banking Supervision would exclude obligations like trust preferred securities from being counted as capital for regulatory purposes. Redefining capital upward to exclude certain securities would lead to a much more conservatively financed and less crisis prone banking industry. However, the reform bill’s conference report provides special exemptions from these new standards, grandfathers existing securities, and provides discretion to regulators that neuters the provision and dramatically reduces its likely impact.
An appreciation for the bill’s proposed changes first requires a digression into what, precisely, is meant by “capital.” And there are different definitions for different businesses. In the world of nonfinancial businesses, like manufacturers, retailers, software companies, and the like, “capital” refers to both the obligations issued by the company and the “capital goods” those obligations finance. The obligations consist of debt (bonds and bank loans) and equity (common and preferred stock) securities, while the capital assets consist of the company’s property, plant, and equipment used in the production process. A firm’s “cost of capital” is the weighted average returns it must provide to the buyers of its obligations to attract the funds the firm then uses to acquire capital assets.
When the topic switches to financial firms, like banks, broker dealers, and investment management companies, the term “capital” is more specific, but, ironically, also more ambiguous. These firms have no capital assets, as their balance sheet consists of financial obligations like stock, loans, and bonds issued by other businesses. “Capital” is therefore only relevant in terms of financial firms’ liabilities and is used to reference their capacity to withstand losses. A “well capitalized” bank is one that could withstand a relatively large fall in the market value of its assets. Therefore, in the context of banking it is the bank’s net assets – assets minus liabilities – that best approximate what is meant by capital.
Arriving at the net asset figure first requires determining what types of liabilities should be subtracted from assets. There are two broad categories of liabilities: fixed and contingent. Common equity is an example of a contingent liability. Contingent liabilities are not really liabilities at all because they represent ownership stakes without any binding financial commitment on the part of the issuer. Equity holders receive the “residual” – whatever is left after all other claims are paid, if anything. If a bank was 100% equity financed, there would be no risk of default because contingent liabilities are default-free; as the value of the bank’s assets fell, so too would the value of the stock on a one-for-one basis. A mutual fund is a financial company whose assets are entirely financed with contingent liabilities. The value of the shares in the mutual fund is entirely dependent on the value of the assets the fund manager acquires with no recourse for shareholders if the investments tank.
Senior debt, conversely, is a fixed liability. The firm pledges to pay its creditors a specified amount each period and after a certain number of such periods the firm repays the creditors’ initial investment, or principal, in full. Unlike the equity holder, the senior debt holder has no ownership stake in the company, cannot exercise any rights with respect to the firm’s business decisions, and has no “upside” beyond getting repaid on the agreed terms. In exchange, the senior debt holder receives a priority payment that must be made if the firm wishes to avoid a bankruptcy filing (or bankruptcy-like proceeding for banks). If the firm cannot meet its debt obligations, the creditor has the ability to seize control of the firm and sell off its assets to recover some of the value of its principal. In the context of banks, senior debt also includes deposits, which must be repaid to the depositor on demand (or as specified in contract) for 100 cents on the dollar; the value of deposits remains static no matter how the bank’s investments perform.
If banks were entirely financed with deposits, debt, and common stock, there would be no trouble deciding what counts as “capital.” The capital ratio of the bank with the “Idyllic Balance Sheet” would be 20%, as its $100 portfolio of financial assets is financed by $80 of fixed liabilities that require full repayment and $20 in contingent liabilities for which the bank has no repayment obligation. The market value of the bank’s assets could fall by 19% without triggering insolvency; the book value of the deposits and bonds would remain unchanged, as obligations could be redeemed at par, while the value of the common equity would fall to $1.
Rarely do we see such a balance sheet in practice. Instead, banks and bank holding companies rely on a range of “hybrid” securities that have a blend of features that make them neither truly fixed nor contingent liabilities. Common equity is the most expensive form of capital. Because common shareholders have the most junior claims on the assets of the firm, they are all but assured to be wiped out in the event of default. This downside risk causes the variance of the returns on common equity to be much greater than the variance of the return on the assets of the firm or other obligations. Common equity issuance is also generally opposed by existing shareholders because of its dilutive effects. If a bank with $100 of expected future net income and 100 shareholders issues 20 new common shares to raise capital, the earnings per share of the existing shareholders falls by 17% (from $1 to $0.83), with a like decline in the market value of the stock. Making your existing shareholders poorer is generally not a winning strategy for corporate managers.
Rather than rely on common stock, banks use what has been called “innovative Tier 1 capital.” Tier 1 capital is the core capital for regulatory purposes. It serves as the primary measure established by bank regulators to measure the solvency of financial institutions. In addition to common equity, Tier 1 also includes certain forms of preferred stock, retained earnings (income owed to common shareholders but not paid out as dividends) and other securities that meet a multi-pronged test. As one might expect, the “innovation” associated with these securities largely consisted of pushing the envelope on what sorts of bells and whistles one would have to put on fixed liabilities to get them to count as capital.
The recent financial crisis was a case study in capital innovation. In 2007, Citigroup raised convertible debt to get its Tier 1 capital ratio above prudential minimums. Convertible securities were also issued by Lehman Brothers in 2008. In both cases, these hobbled firms were desperate to raise capital, but weren’t willing to incur the expense associated with issuing common equity, particularly because the dilutive impact on existing shareholders is inversely proportional to the share price (when the stock is $50, a firm would have to issue 200,000 shares to raise $1 billion compared to 500,000 shares if the stock price falls to $20).
The most consequential “innovation” in bank capital stemmed from an October 1996 Federal Reserve ruling that trust preferred stock could count as Tier 1 regulatory capital for bank holding companies. Trust preferred securities are complex instruments that transform straight debt into equity capital through a series of transactions. The bank creates a trust that issues redeemable notes to investors. The proceeds of those notes are used to purchase debt from the bank that has the exact same payment features as the notes issued by the trust. The bank makes tax-deductible interest payments to the trust that are passed through to investors. The securities are therefore a form of regulatory arbitrage, as they are simultaneously treated by the bank holding company as debt for tax purposes and Tier 1 capital for regulatory purposes.
The problem with trust preferreds and similar instruments is that a failure to make promised payment is treated as tantamount to default. Rating agencies generally include promised payments on trust preferred securities in the calculation of the cash flow coverage ratio, which is to say these payments are treated identically to coupon payments on debt. Rather the providing a cushion in times of stress, trust preferreds actually increase cash pressures by creating additional fixed obligations.
In 2007, Citigroup’s trust preferred securities increased by $14 billion (see page 75 of 208). This caused the firm’s Tier 1 capital to fall by only $1.6 billion even as losses on subprime-linked assets caused common equity to drop by more than $5 billion. This made a comedy of Tier 1 capital. As the Bank for International Settlements (BIS) explains, with Citi very much in mind, the Tier 1 definition:
allowed a number of banks to report high Tier 1 ratios but with low levels of common equity net of regulatory adjustments. As the crisis deepened, banks faced growing losses and write downs which directly reduced the retained earnings component of common equity, calling into question fundamental solvency. Many market participants therefore lost confidence in the Tier 1 measure of capital adequacy.
In the case of Citigroup, market participants no longer paid any attention to Tier 1 capital, focusing exclusively on tangible common equity.
When the Fed, the FDIC, and the Office of the Comptroller of the Currency conducted stress tests of the 19 largest U.S. financial firms in the spring of 2009, they used Tier 1 common equity as the capital standard. In effect, they acknowledged that their regulatory standard was an inadequate measure of these firms’ ability to withstand losses. As the Fed explained, common equity is “the first element of the capital structure to absorb losses, offering protection to more senior parts of the capital structure and lowering the risk of insolvency. All else equal, more Tier 1 Common capital gives a BHC (bank holding company) greater permanent loss absorption capacity and a greater ability to conserve resources under stress by changing the amount and timing of dividends and other distributions.”
At the end of 2009 an estimated $129 billion of trust preferred stock was counted as Tier 1 capital. If banking regulators were suddenly to acknowledge that this is not real capital, the financial system would have to sell assets or raise new common equity on a massive scale. To protect against this outcome, Section 171(b)(4)(B) of the financial regulatory reform bill provides a three-year phase in on new capital standards for securities issued before May 19, 2010. While this phase-in period is reasonable, Section 171(b)(5) of the bill entirely exempts banks with less than $500 million in assets from the new capital standard, while Section 171(b)(4)(C) grandfathers securities issued before May 19, 2010 by holding companies with less than $15 billion in assets as of December 31, 2009. This means that smaller banks (less than $500 million) could issue new trust preferreds in the future and have it count as Tier 1 capital and medium sized banks (less than $15 billion) could have all existing trust preferred securities (some with 30 years to maturity) count towards Tier 1 indefinitely.
These exemptions are not reasonable. If trust preferred securities are not properly classified as Tier 1 capital, that is just as true for small banks as it is for large ones. Although most agree the financial crisis was largely the product of overly aggressive investment and financing strategies at large financial institutions, it is important to remember that 227 banks have failed since the start of 2008. Most failed banks are tiny, with an average size of $2.7 billion. If $330 billion WaMu is excluded from the list, the average failed bank had only $1.3 billion in assets. The number of future bank failures would almost certainly be lower if the new capital standards were applied to all banks (after a phase-in) because it would lead to more conservative financing. While the financial regulatory reform bill was right to address the important definitional question of what counts as capital, varying this definition based on size only increases the likelihood of the financial system once again experiencing extreme volatility.