■ by Alexey Levkov, Senior Vice President, and Clark Peterson, Senior Vice President, The Clearing House
Bank equity capital is critical to maintaining a safe and sound financial position. It is the difference between a bank’s assets and its liabilities and is a funding source that acts as a financial cushion to absorb unexpected losses at times of distress. Beyond being a cushion, capital creates a strong incentive to manage a bank in a prudent manner because bank shareholders are at risk in the event of a failure.
The case for imposing capital requirements on banks rests on externalities associated with bank failure. In particular, a failure of a large bank has systemic costs that are not fully borne by the bank. These costs, among others, include disruptions in the financial markets, system-wide reduction in asset values, and losses among other market participants. The regulator’s task, through mandating higher capital requirements, is to make banks internalize the systemic costs of their potential failure and buffer the financial sector from the resulting systemic losses.
In this article we analyze the costs associated with changes in capital requirements. We distinguish between equity capital and debt capital and discuss both the private and social cost of requiring banks to hold additional equity capital. In the first part of the article we discuss the cost implications of changes in the mix of equity and debt. In the second part we consider how heightened capital requirements may affect allocation of credit in the economy with potential implications for financial stability and bank intermediation activities.
Our main conclusion is that when abstracting from the stylized theoretical framework, funding bank activities with equity is more expensive relative to debt. We should note that it’s certainly the case that bank equity levels can be too low. In this case, more equity would enhance market perceptions of bank safety, lower cost of capital and likely result in greater credit extension. But our read of academic literature suggests that substantial increases in equity capital requirements, both in the U.S. and elsewhere, result in reduced lending and economic growth and pushes some borrowers to seek credit from less regulated sources.
This leads us to two policy recommendations. First, we encourage policymakers to balance the costs and benefits of higher capital requirements (Posner, 2014), taking into account the close link between financial stability and economic growth (Rosengren, 2011; Hanson, Kashyap, and Stein, 2011). Second, capital calibrations should incorporate various banking activities that are vital for economic growth, not just the probability of bank failure, thus emphasizing the macroprudential aspects of financial regulation (Hirtle, Schuermann, and Stiroh, 2009).
Capital Structure and the Cost of Capital
At the firm level, the debate over bank equity capital requirements comes down to capital structure. Managers must make choices about how best to finance their assets. Broadly speaking, banks finance their assets with a mix of debt (and deposits, a form of debt) and equity, all of which are claims to the firm’s assets. Banks secure financing in the context of competitive and economic forces, as well as regulatory requirements. Capital structure choices have significant trade-offs for companies in general and for banks in particular.
Debt and equity are both sources of financing, or capital, but in the banking industry context “capital” is shorthand to refer to the firm’s equity capital. Equity is a (residual) ownership claim to a firm’s assets that adjusts in value depending on the value of cash flows that a company is able to generate from its assets. Leverage, or debt capital, involves the substitution of fixed cost debt for the residual claim of owner’s equity. Leverage increases expected return and it increases risk. Leverage allows a firm to create multiple risk-return combinations from any particular set of assets (Higgins 2011).
Banks create value by earning a rate of return on their assets for equity holders, and capital structure choices can impact that return. At an operating level, the relationship between leverage and return on equity (ROE) depends on a number of factors.
In addition to expecting a return, equity investors and creditors incur a cost when they finance a business’s assets. This cost relates to the risk of the assets they are financing. They could earn returns on other investments with similar risk profiles. This opportunity cost comprises the minimum rate of return the business needs to earn on its assets to meet the demands of its debt and equity investors (Higgins 2011). This is the firm’s cost of capital.
A company’s cost of capital is known as its Weighted Average Cost of Capital, or WACC. It states that a company’s overall cost of capital is the cost of each particular type of capital—both debt and equity—weighted by their proportion in the company’s capital structure:
This formula tells us that a company’s cost of capital is its cost of equity (
) times the proportion of equity in its capital structure (
) plus the cost of debt (
)times the proportion of debt in the capital structure (
). To understand how a decrease in leverage or an increase in equity capital requirements will affect loans rates, credit extension, and financial stability, we must understand how those changes impact a bank’s WACC.
Acquiring or holding assets requires an estimation of both their return and their risk. Banks create value for shareholders by investing in assets—making loans—that have a rate of return that exceeds their cost of capital. Thus, the cost of capital is used to appraise the attractiveness of new loans that will generate risky cash flows. The risk-return calculation on whether to make those loans is determined by “discounting” the cash flows from the loan by the cost of capital. Thus, if a bank’s cost of capital exceeds the rate of return derived from cash flows from a new loan, the loan will not be made or the bank will increase the rate charged for the loan so that it exceeds its cost of capital.
Since debt provides a fixed return and contractual certainty to the debt holder, the cost of debt is an observable data point and is simply the yield-to-maturity on the debt and other liabilities issued by the bank.
Estimating the cost of equity, an unsecured residual claim, is more challenging. The key is to examine the three types of risks generally related to holding a risky asset: (1) the time value of money,
How can we measure a firm-specific risk premium? The firm-specific risk premium is known as beta (ß).
Therefore, the cost of equity (
) equals the risk free interest rate on government bonds (
) plus the firm-specific risk premium: beta (ß) times the Market Risk Premium. The Market Risk Premium is defined as the market returns over and above the risk free rate
(
).
Using the formula above, we can estimate the cost of equity for an average risk company. The return on U.S. common stocks on average has exceeded the risk-free rate on government bonds by 6-7% (Higgins 2011). Let’s assume a market risk premium of 7%, a 3% rate on a long-term Treasury bond, and a beta of one. The cost of capital is 3% + 1 * (7%) = 10%. Under these assumptions, an average-risk U.S. company has a cost of equity of 10%.
Estimates vary, but banks have equity betas of approximately 1.2.
To answer that question, we need to first understand what theory predicts and then examine real-world factors that may deviate from the theoretical framework. Theory predicts that higher equity and less leverage in a bank’s capital structure will lower the bank’s beta, which will reduce the cost of equity (
). Leverage increases both risk and return. Bank betas capture the risk of the bank’s assets plus the incremental risk related to leverage. Betas can be “unlevered and relevered” according to target capital structure.
The idea that the cost of equity declines to precisely offset additional reliance on more expensive equity is conceptually sound though subject to numerous simplifying assumptions—assumptions that vary in the degree to which they reflect reality. Indeed, the foregoing analysis took into account only the narrow set of factors.
Discussions of capital structure and corporate finance often begin with the theoretical framework of Modigliani and Miller (MM), which shows that, under certain conditions and when cash flows are held constant, the amount of debt a company carries has no effect on its value. Capital structure is irrelevant. The MM framework illustrates an important point: most companies create value by managing their assets to maximize cash flows. Rearranging the proportion of paper claims to these real assets will have no effect on cost of capital, asset value, or firm value.
The MM irrelevance proposition is the starting point for analyzing capital structure; it is not the ending point. Its conditions are rigorous, including no taxes, no asymmetric information, and perfectly rational risk-based pricing—a frictionless world. MM does not reflect the real world, yet remains foundational because it forces us to be precise about those factors that make capital structure decisions important. In reality, there are a number of factors related to capital structure choice that do matter. But some use this theoretical, conceptual foundation to argue that there is “no cost…whatsoever” to higher bank equity capital requirements in terms of loan pricing, credit availability, economic growth, or financial stability (Admati and Hellwig 2013). The argument rests on a gross oversimplification of the complexities involved in bank capital structure.
Empirical support for beta’s responsiveness to changes in leverage is far less established in reality than it is in theory (Abuaf & Amit 1999, Baker & Wugler 2013). Corporate finance research that assesses the factors involved in estimating cost of capital and capital structure decisions often exclude banks altogether due to the unique nature of banks’ use of leverage and the role of regulation (King 2009). There are surprisingly few studies on the cost of bank equity and there is a clear need to examine further the trade-offs at stake with bank capital structure (Berlin 2011).
The Relevance of Bank Capital Structure
A number of real-world frictions deviate from the irrelevance condition of a Modigliani and Miller world. In general, debt can impact the value of cash flows for both banks and nonbanks in five primary ways: tax benefits, market signaling, flexibility, distress costs, and management incentives (Higgins 2011). In addition, a number of factors unique to banks impact their cost of capital, including socially valuable liquidity production; competitive and regulatory dynamics; and the effect of government programs like deposit insurance.
Taxes
The most common jumping off point from a MM perspective on capital structure is the effect of taxes. In the U.S., as in many developed countries, interest costs are a deductible expense for tax purposes. The effect of leverage in capital structure is to reduce taxes and create an “interest tax shield.” These are incremental cash flows that firms generate from their use of leverage. The interest tax shield is so prevalent in valuation that it’s integrated directly into the debt portion of the WACC formula: 
. This formula says that the real cost of debt is an after-tax cost (
). Thus, as a firm introduces more leverage into its capital structure, its WACC goes down, because debt is cheaper than equity even on a risk-adjusted basis. That is, an increased equity beta from an increase in leverage will no longer precisely offset in the cost of capital. A bank’s cost of capital, and hence the rate it charges on loans, will go up the more the bank is required to rely on equity. This effect on banks’ cost of capital has been quantified by a number of authors (Elliot 2009; Stein and Hanson 2010), each of which predict a moderate increase in lending rates.
Asymmetric Information, Management Incentives and Growth
Another important way that capital structure can impact bank lending is through the incentives that leverage provides to management in the context of asymmetric information regarding a bank’s assets and growth opportunities. A benefit of leverage is that in certain circumstances it can increase alignment between the interests of shareholders and management. The addition of fixed interest expenses to a business’s operating profile can help discipline management to deploy the firm’s cash flows to valuable projects (Higgins 2011).
The manner in which debt can align shareholder and management incentives varies by the investment opportunities a firm confronts. McConnell and Servaes (1995) empirically demonstrate that for “low growth” firms in mature industries, “value is positively correlated with leverage” (page 123). In their analysis, leverage helps to equilibrate between “overinvestment and underinvestment” problems. The set of investment opportunities combined with the interplay of shareholder-management incentives will create an optimal amount of leverage for firms. Given the well-documented asymmetric information problems unique to banks’ assets and the mature, highly developed nature of bank industry conditions, it’s likely that this effect is present and would impact banks’ lending decisions. McConnell and Servaes conclude that empirical, real world data show that “debt policy and equity ownership structure ‘matter.’”
Deposits and Liquidity Production
In the banking literature, banks are often described as “special,” and their specialness derives in part from the fact that they create liquid deposits that serve as money in our economy. But deposits are a funding source for banks, a form of bank debt, and banks are thus unique from most other companies in that they produce socially valuable services from the debt liabilities in their capital structure. Recall that nonbanks create value only from their assets. Because consumers and businesses demand a liquid, ready source of funds, there will be a pricing signal, or a premium, to provide liquid claims in the form of deposits (De Angelo and Stulz 2013).
In providing deposits, banks create value directly through their capital structure choices, and debt and equity are not precisely equivalent risk-adjusted sources of capital for banks. This is a clear departure from a theoretical MM world. DeAngelo and Stulz (2013) show that the provision of “(privately and socially beneficial) liquid financial claims” makes high leverage optimal for banks before taking into account taxes, asymmetric information, deposit insurance, or any other real-world frictions. They further argue that because MM does not contemplate the provision of deposit liabilities, MM “is an inappropriately equity-biased baseline for assessing” bank capital structure choices (page 3).
Returning to WACC, it follows that singling out banks for substantially heightened equity capital requirements will increase banks’ overall cost of capital and thus loan rates. This is due to a premium associated with bank deposits as a financing source and, compared with equity capital, bank deposits are a less expensive source of debt capital (even on a risk-adjusted basis). Moreover, bank deposits benefit from FDIC insurance, and while banks pay premiums for this insurance, the presence of official support will dampen the risk sensitivity of the providers of this funding source and reduce the cost of deposit funding compared to equity.
Equity Capital, Intermediation, and Financial Stability
The discussion thus far suggests that equity capital is more expensive relative to debt for banks. But what are the implications of requiring banks to have more capital in their funding mix? Would the heightened capital requirements have any implications for intermediation activity? If so, are there further implications for investment on the part of the borrowers? Would they seek credit elsewhere and what does it mean for financial stability?
Most discussions of bank capital regulation start from the premise that the goal is to keep the probability of bank failure below a certain fixed threshold, typically set in terms of a “confidence level.” A confidence level of 99.9%, for example, means that there is only a 0.1% chance that potential losses will exceed the sum of “expected” and “unexpected” losses, thus ensuring that probability of bank failure remains low. Once the confidence level is set, the necessary level of capital can be calibrated using information in each bank’s portfolio (Gordy, 2003).
Calibrating capital requirements only with respect to bank failure is problematic in that it does not account for bank lending functions and ignores the trade-off between a lower probability of bank failure and continuation of lending. In some states of the world, improvements in financial stability may outweigh the costs of lower output growth, thus justifying higher capital requirements. In other circumstances, however, improvements in financial stability may only be marginal relative to foregone lending opportunities, thus making capital requirements “inefficiently” high.
Changes in capital requirements may affect lending in two ways. First, as we have shown in the previous sections, raising bank capital requirements results in higher WACC (Baker and Wurgler, 2013). Thus, if a bank’s cost of capital exceeds the rate of return derived from cash flows from a new loan, the loan will not be made or the bank will increase the rate charged for the loan to cover its cost of capital. Too-stringent capital requirements may result in reduction in the supply of credit, underinvestment in the economy, and thus slower economic growth.
Alternatively, note that capital requirements are stated in terms of ratios. Faced with increasing capital requirements relative to total assets or risk-weighted assets, banks may respond by shrinking their size. To be specific, consider a bank with assets of $100 that is financed with deposits and $6 of equity capital, thus having 6% of equity capital relative to assets. Suppose that a new policy requires the capital ratio to be at least 10% relative to total assets. The bank can fulfill the new requirement by either raising $4 of fresh equity or by leaving its current level of equity unchanged and shrinking its assets to $60.
If asset shrinkage—perhaps by cutting back on lending—is limited to a single bank, then other banks can potentially pick up the slack without major implications for intermediation activity. However, there are serious consequences if multiple banks respond to higher capital requirements by reducing the flow of credit to the economy. Moreover, asset shrinkage on the part of the regulated banks may push borrowers to seek credit in a less regulated environment, with potentially negative consequences for financial stability.
Equity Capital and Intermediation
There is a large literature that explores how shocks to bank capital impact lending and economic activity. One of the cleanest studies in this area is by Joe Peek and Eric Rosengren (1997). Their key insight is that the sharp decline in Japanese stock prices between 1989 and 1992 was transmitted to the United States via U.S. branches of Japanese parent banks. Specifically, they find that a one percentage point decline in the Japanese parent’s risk-based capital ratio resulted in a 6% decline in loans originated by the U.S. branches. In a subsequent study, Peek and Rosengren (2000) analyze the implications of the decline in lending for real economic activity. They find that the decline in lending was followed by a sharp reduction in commercial real estate activity in the U.S., especially in states with more U.S. branches of Japanese parent banks.
Haubrich and Wachtel (1993) analyze the response of U.S. commercial banks to the 1988-89 announcement and implementation of higher capital requirements. They group the banks by their undercapitalization relative to the new minimum capital requirements and further segment the banks by their size. Haubrich and Wachtel find that relatively undercapitalized banks of all sizes shifted the composition of their portfolio in response to the new capital requirements, effectively shrinking their risk-weighted assets. The results are consistent with Bernanke and Lown (1991), who find a similar pattern of asset shrinkage in response to capital depletion among banks in New Jersey during the late 1980s.
In a more recent study, Francis and Osborne (2009) analyze the lending implications of bank-specific, time-varying capital requirements set by regulatory authorities in the U.K. They carefully construct a time series of bank-specific capital shortfalls and surpluses for the period 1996-2007 and estimate whether banks with capital shortfalls adjust their lending behavior. Their regression analyses suggest that a one percentage point increase in capital ratio results in a 2.4% reduction in risk-weighted assets, a 1.4% reduction in total assets, and a 1.2% reduction in lending over a period of four years.
Equity Capital and Financial Stability
Further, the implications of higher capital requirements for financial stability are not clear-cut. On the one hand, higher capital requirements reduce probability of default by forcing shareholders to absorb a larger fraction of losses in times of distress and by mitigating moral hazard concerns associated with excessive risk taking. On the other hand, faced with heightened capital requirements, banks can respond by selling illiquid assets causing asset prices to decline. Similar assets held by other market participants may decline in value as well, causing additional asset sell off. This self-reinforcing process increases system-wide financial fragility. Elements of this mechanism have been described in Shleifer and Vishny (1992).
Furthermore, increasing evidence shows that regulatory burden and capital requirements steer some of the traditional banking activities towards the shadow banking sector. With less supervisory oversight and more uncertainty about the quality of lending, the growth of shadow banks poses concerns about allocation of credit, output growth, and financial stability.
Aiyar, Calomiris, and Wieladek (2014) use a clever design to assess the impact of changes in capital requirements on bank lending and the allocation of credit. Specifically, the authors use differences in minimum capital requirements for U.K. banks between 1998 and 2007 and find that a 100 basis point increase in minimum capital requirement for banks regulated by the Financial Services Authority leads to 6-9% reduction in lending. Further, their analysis suggests that about one-third of the reduction in lending is picked up by branches of foreign-owned banks not regulated by the FSA. Although foreign-owned banks pick up some of the lending slack and offset the overall impact on lending, the “leakage” of lending to less-regulated areas of the banking system poses concerns for financial stability.
The present debate about the efficient level of bank equity focuses on the implications of additional capital requirements for financial stability and economic growth. At the core of the debate is the cost of equity relative to debt and how the equity-debt cost differential changes when banks are required to alter their funding mix.
On the one hand, theory suggests that capital structure has no impact on the cost of equity relative to debt (Modigliani and Miller, 1958) suggesting that there is no reason banks should not face significantly higher minimum capital requirements (Admati et al, 2013). However, when abstracting from the stylized and plainly unrealistic conditions in Modigliani and Miller with, for example, the introduction of taxes, asymmetric information, and deposit funding, then the equity-debt mix becomes important in determining the relative price of equity.
Regulators are probably aware that capital is more expensive than debt and too-stringent capital requirements may dampen intermediation activities. Otherwise, why not set requirements significantly higher? Indeed, existing empirical literature suggests that heightened capital requirements are often met with a reduction in lending, with implications for investment activity and migration of lending to less regulated environments. It is important that policymakers evaluate the substantial regulatory policies imposed on the banking industry thus far and carefully consider the macroeconomic implications of further increases in capital requirements.