Since the financial crisis, bank capital positions have improved considerably. However, calls for heightened capital requirements have not abated. Federal Reserve Chair Ben Bernanke, Vice Chair Janet Yellen, and governors Daniel Tarullo and Jeremy Stein have all suggested that Basel III may not go far enough, particularly for large and complex banks. The legislation proposed by Senators Sherrod Brown and David Vitter lays out a plan to raise requirements further, and there are other recommendations both by policymakers and academics.
It is hard to measure the long-term benefits and costs of capital requirements with precision. The issue is multidimensional, involving agency problems in banks, asymmetric information, international coordination, bank governance, taxes, government subsidies, systemic risks, shadow banking, and so on. Recent experience suggests that the benefits of a safer and more stable banking system, while hard to precisely quantify, are considerable.
A central concern, however, is whether substantially heightened capital requirements will increase the overall cost of capital, raise lending rates, and limit the growth of economic activity in the process. Bankers have appealed to a simplistic logic about the cost of debt and equity. If equity is more costly than debt, then more equity means a higher overall cost of capital.i
Many economists view this as a fallacy.ii
In theory, reducing leverage with increased capital requirements makes banks safer and lowers the returns that shareholders require. Even though equity is costlier than debt, banks that decide—or are forced—to hold more equity will not change their overall average cost of capital. As long as capital markets are efficient and free of other distortions like taxes, the reduced cost of equity exactly offsets its increased weight in the capital structure, leaving the overall cost of capital unchanged. This simple and powerful theoretical insight won Nobel Prizes for Franco Modigliani and Merton Miller.
Of course, capital markets are not perfect. For example, if debt is tax advantaged, heightened capital requirements will raise the cost of capital somewhat.iii
Our new research (Baker and Wurgler, 2013) focuses on an underappreciated and even more fundamental market imperfection at the core of the Modigliani-Miller logic: the notion that there is a reliable tradeoff between risk and return in the equity markets.
Certain risk-return tradeoffs across asset classes are sensible and apparent: Treasuries tend to return less than corporate bonds, which in turn tend to return less than stocks. But within the stock market, risk and return have not been well connected. If anything, the relationship is backward. Shareholders have long earned lower returns, on average, for bearing more risk. For instance, in a study published with Brendan Bradley of Acadian Asset Management (Baker, Bradley, and Wurgler, 2011), we found that a dollar invested in low risk U.S. stocks in 1968 grew to more than 59 dollars over the next forty-one years. In contrast, the same dollar in high-risk stocks actually shrunk to 58 cents. After inflation, the high-risk investor is essentially wiped out, despite bearing four decades of extreme risk.
This low risk anomaly is a paradox hidden in the data and runs counter to traditional finance assumptions and principles. The pattern is obscured by episodic crises and easy to see only in a long-term analysis. But it is as real as it is striking. A similarly odd risk-return tradeoff appears in many developed stock markets.
Banks and banking regulators need to pay attention because this fact also applies to banks. We collected eighty years of stock returns on several thousand U.S. banks. As theory led us to expect, we found that highly capitalized banks had lower risk, measured by stock beta or volatility. But, as prior research on the low risk anomaly suggests, low risk banks have historically paid more for their equity—in the sense of delivering higher average stock returns—than relatively riskier banks.
If the cost of equity actually rises instead of falls, as banks are required to hold more of it, the overall cost of funds rises, too. And, because lending rates must be higher than funding rates—banks like to make a profit—this will be passed on to borrowers. The effect is quantitatively big. Using the inverted relationship between risk and return in the historical data suggests that a ten percentage-point increase in Tier 1 capital to risk-weighted assets—a stricter increase than Basel III mandates, at least for most banks, but less than others have called for—would have increased the overall cost of capital for U.S. banks by between 60 and 90 basis points per year.
By our estimates, an increase of this magnitude would more than double the spread over Treasury yields that banks usually pay for capital. It would, in competitive lending markets, also increase lending rates by a similar amount. Higher rates would deter investment or direct borrowers toward the less-regulated shadow banking system for better terms.
What causes the backward risk-return relationship among bank stocks? The simplest explanation is that shareholders like their stock risky. They inadvertently accept generally lower returns on risky stocks, even the occasional collapse, for the shot at a big upside, the proverbial “next Microsoft.” This risk-loving attitude may derive from a mix of well-documented biases in our psychology including optimism, overconfidence, and preferences for lottery-like payoffs. Another explanation is that the mere categorization of a security as equity leads investors to demand a return that resembles the return on other stocks, regardless of its particular risk properties. This sort of behavior arguably influenced the pricing of collateralized debt obligations that were labeled as AAA. A final explanation is that the smart money is often on the sidelines. Institutional equity managers are typically benchmarked against the overall stock market. A portfolio of low risk stocks appears risky to them, in the sense that it tends to lag in bull markets.
But frankly, for bank regulators and banks, the exact explanation does not matter. All that matters is the market reality. While forcing banks to reduce leverage will reduce the risk of equity just as intended, it will also—if historical experience is a guide and investor attitudes do not change—increase the cost of equity and in turn lending rates. There are subtler implications here, too. Banks may respond by increasing their risk in other ways to offset the increase in capital requirements, by pairing lending with riskier activities like investment banking, market making, asset management, and brokerage. Regulators might lean toward rules that allow a greater reliance on unsecured or convertible debt, which leaves equity risk largely unchanged. Perhaps the link between risk and return can be corrected, if new and very low risk bank equity is marketed to investors as different, more like near-investment-grade debt than prototypical equity.
Let us also be very clear. We are not arguing that undercapitalized banks are a good idea. We are not sure what level of capital requirements best balances the costs and benefits. Clearly, fewer financial crises would be welcome. Our point is that regulation that makes bank equity safer will raise the overall cost of capital, and thereby lending rates. Borrowers in particular will then pay a price for safer banks. This is a heretofore-unrecognized cost of capital requirements and it needs to be added to the regulatory conversations in Basel, Brussels, London, and Washington.
Malcolm Baker is Robert G. Kirby Professor of Business Administration at the Harvard Business School. Jeffrey Wurgler is Nomura Professor of Finance at the NYU Stern School of Business. Both serve as consultants to Acadian Asset Management.
For example, Elliott (2013) quotes a former managing director of JP Morgan and policy analyst as writing, “the first-order effect of increasing the ratio of common equity to total assets for banks from 5% to 30% would clearly be very high. Assume that the annual cost of bank equity is 5 percentage points higher than the after-tax cost of bank deposits and debt…”ii
For example, see Admati, DeMarzo, Hellwig, and Pfleiderer (2011).iii
Other authors, including Hanson, Kashyap, and Stein (2010) have attempted to calibrate the effect of increased capital requirements in light of corporate taxes.
1) Admati, A. R., DeMarzo, P. M., Hellwig, M. F., and Pfleiderer, P. C. 2011. “Fallacies, Irrelevant Facts, and Myths in the Discussion of Capital Regulation: Why Bank Equity is Not Expensive.” Working Paper, Stanford University.
2 )Baker, M., Bradley, B., and Wurgler, J. 2011. “Benchmarks as Limits to Arbitrage: Understanding the Low-Volatility Anomaly.” Financial Analysts Journal 67, 40-54.
3) Baker, M., and Wurgler, J. 2013. “Do Strict Capital Requirements Raise the Cost of Capital? Banking Regulation and the Low Risk Anomaly.” Working paper, Harvard University.
4) Elliott, D. J. 2103. “Higher Bank Capital Requirements Would Come at a Price.” Internet Posting, Brookings Institute.
5) Hanson, S., Kashyap, A. K., Stein, J. C. 2010. “An Analysis of the Impact of ‘Substantially Heightened’ Capital Requirements on Large Financial Institutions.” Working Paper, Harvard University.