Notes from the Vault
Larry D. Wall
December 2017

The end of 2017 seems an appropriate time to look back at the many regulatory changes adopted in the decade since the financial crisis. With that in mind, the Atlanta Fed recently hosted the workshop Financial Regulation: Fit for the Future?, which was cosponsored by the Center for the Economic Analysis of Risk at Georgia State University. This post reviews many of the papers and presentations from the workshop dealing with a variety of regulatory policy issues. Two upcoming macroblog posts will review the papers and presentations that focus on discretionary macroprudential policies and the regulation of fintech (new technologies applied to finance).

Market discipline
During the financial crisis, many governments provided support to their domestic banks, thereby protecting bank creditors and owners from losses they would otherwise have taken. Immediately after the crisis, the G-20 agreed changes were needed to end so-called too-big-to-fail policies, both to discourage excessive risk taking by banks and to protect taxpayers. Two papers presented at the workshop examine the effectiveness of these measures.

The European Union adopted the Bank Recovery and Resolution Act in 2014, which requires (among other things) that at least 8 percent of liabilities and equity be written down before any state aid can be provided. The requirement exposes many bank bondholders to loss given the low-level equity-to-asset ratios at many European banks.1 Office of Financial Research economist Benjamin Kay and his coauthors analyze the credibility of some of these changes in the paper "The Market-Implied Probability of European Government Intervention in Distressed Banks." The paper exploits some unique features of the market for credit default swaps on European bank bonds to analyze market perceptions of the probability of a bailout. Kay et al. find evidence of reduced expectations of a government bailout from the fourth quarter of 2014 through the second quarter of 2016. However, expectations of a bailout were reversed starting in the third quarter of 2016 and have returned to pre-reform levels. The authors contend the timing of this reversal suggests it is likely due to the treatment of some distressed Italian banks.

The second paper examined two hypotheses about the relationship between capital requirements on bank equity values. One hypothesis is that bank equity holders would prefer low capital levels so that if the bank fails, most of the losses fall on creditors and deposit insurers. The other hypothesis holds that a bank charter can provide a valuable stream of profits that would be lost if the bank fails, implying shareholders benefit from more capital. Professors Joseph P. Hughes of Rutgers University and Choon‐Geol Moon of Hanyang University, along with Cleveland Fed President Loretta J. Mester, study the effects of bank capital levels on shareholder value in "Market Discipline Working for and against Financial Stability: the Two Faces of Equity Capital in U.S. Commercial Banking." The authors find that banks with greater investment opportunities benefit from higher capital and these banks tend to be smaller. Conversely, they find banks with lower-valued opportunities could increase their stock price by lowering their capital ratio; these banks tended to be the largest, systemically important banks. The paper concludes that the empirical findings support the need for regulatory capital requirements.

Regulatory capital requirements
Two of the conference papers provided theoretical analysis of the impact of regulatory capital requirements. "Equilibrium Theory of Banks' Capital Structure," by Professors Douglas Gale of New York University and Piero Gottardi of the European University Institute, develops a model of socially optimal capital levels in a setting where banks are lending to commercial firms. The model has two important frictions: (a) deposits are valued for both their financial return and their liquidity services (the ease of spending on goods and services); and (b) if the bank or firm defaults on its debt, then it incurs deadweight losses in bankruptcy (costly default). In this model, corporate borrowing must come from the banks and the banks finance the loans with a combination of debt and equity.

The analysis in the paper provides a somewhat surprising result: if the credit risk of loans rises or falls for all loans at the same time, then the optimal solution is for firms to hold all of the capital and banks to be solely financed by deposits. The reason is that capital held at the firms does double duty: it protects both the firms and the banks from failing. However, in the more general case where firms have idiosyncratic risk (some firms become riskier, while others become safer), the bank optimally holds some capital to guard against the possibility that its borrowers turn out to be more risky than the average borrower. Importantly, in both cases, the firms and banks will maintain the socially optimal amount of capital without regulatory action. However, the last part of the paper is careful to explain that its model does not include many of the reasons typically given for capital regulation, including moral hazard and the possibility of government bailouts. Thus, the paper should be viewed more as providing solid micro-foundations to start the analysis of capital regulation than as being the final word on the subject.

The other theoretical exploration of regulatory capital is "How Is the Likelihood of Fire Sales in a Crisis Affected by the Interaction of Various Bank Regulations?" by Divya Kirti, an economist at the International Monetary Fund, and Vijay Narasiman. The paper focuses on which assets a distressed bank will shed to raise its capital ratio, given a binding risk-based capital ratio. The paper concludes that a distressed bank will prefer to reduce its holdings of low-risk assets if the weights on low-risk/low-return assets are the same as those on high-risk/high-return assets. The bank will do this to earn higher expected returns and benefit from the greater option value of risky assets. However, if the weights on the high-risk assets are sufficiently larger than those on the low-risk ones, the bank will reduce its holdings of high-risk assets, as that is a more efficient way of reducing the amount of capital the bank must hold.

Credit risk transfer
We ordinarily think of suppliers of funds (lenders, investors, or depositors) as bearing the risk of credit-related losses. However, there are a variety of ways in which the risk can be transferred to third parties, including the provision of collateral to creditors and the securitization of assets where the seller retains part of the risk. Two papers explored some of the subtle ways these techniques can influence the stability of individual banks.

Banks provide collateral to some creditors as a way of reducing the banks' cost of funding by lowering the amount of credit risk borne by the funds' suppliers. Examples of such collateralized borrowing include repurchase agreements, covered bonds (in Europe), and Federal Home Loan Bank advances (in the United States). "Asset Encumbrance, Bank Funding, and Fragility" by Toni Ahnert of the Bank of Canada and his coauthors shows that such transactions come with both a benefit and a cost. The benefit is that the access to additional, cheaper funding allows a bank to earn higher profits in good states of the world. However, such asset encumbrance also shift risks onto the bank's unsecured depositors and creditors if the bank fails. As a result, unsecured creditors have an incentive to run at higher levels of capital than would be the case if none of the bank's assets had remained unencumbered.

A bank may also transfer some of its risk of loss by securitizing loans or pooling a group of loans together, creating securities backed by the loans' cash flows, and selling these securities to investors. Securitization was a popular way of financing riskier mortgages prior to the crisis. However, after the crisis, some observers allege mortgage lenders engaged in reckless behavior, knowing that if the loans went bad the losses would be borne by the investors in mortgage-backed securities.2 One policy response to this concern after the crisis has been to require lenders and/or securitizers to have some "skin in the game" by retaining some of the risk associated with securitized loans. The Bank of Canada's Maarten R.C. van Oordt observes in "Credit Risk Transfer and Bank Insolvency Risk" that the direct effect of such risk retention requirements may be to increase the insolvency risk to the securitizing bank. The problem arises because securitizations transfer risk in those states in which the underlying loans default, but do so at the cost of reducing earnings in those states where the loan is good. If the states where the loans default are those in which the bank would fail in any case, the securitization has not reduced the bank's risk. Conversely, by reducing profits in the states where the bank suffers losses but might still be viable, securitization reduces the bank's earnings and its chances of survival.

Swaps markets
Prior to the crisis, the over-the-counter (OTC) swap markets faced criticism from various quarters. First, the transactions created bilateral credit risk, implying that the failure of one big market maker could adversely affect the condition of others. Second, no one party could fully understand the overall risk exposures in this market because the parties did not report their individual transactions to any outside party. A third criticism was the OTC market was not sufficiently competitive. The financial crisis revealed these concerns were not merely hypothetical, as government policymakers struggled to understand the extent to which OTC swap markets were creating a risk of contagion among systemically important banks. In response to these concerns, the major developed countries agreed to require OTC derivatives transactions be reported to a trade repository, and the more actively traded contracts be traded on an exchange and be centrally cleared.3 Two papers presented at the conference analyze some of the changes to the swaps markets using data now available to the trade repositories.

The paper "OTC Premia" by economists Gino Cenedese and Michalis Vasios of the Bank of England and Professor Angelo Ranaldo of the University of St. Gallen analyzes banks' use of central clearing. During the December 2014 to February 2016 sample period, European counterparties to swap contracts had the choice whether to use central clearing or continue to manage the credit risk of swaps bilaterally. If they chose central clearing, a central counterparty (CCP) would take a position between the two swap counterparties, promising each would be paid what it was due even if the original counterparty defaults. In return, the CCP required that both counterparties in the swap provide collateral to the CCP to reduce the CCP's risk of loss if one counterparty did default. The paper analyzes the choice of central clearing for plain-vanilla U.S. dollar interest rate swap contracts in the United Kingdom. It finds these contracts are more likely to be centrally cleared if they are for larger contracts for longer maturities, and if one of the counterparties is a dealer, bank, or hedge fund. However, the authors also find swap contracts are less likely to be cleared if one of the counterparties is less sophisticated and lower rated, and the contract is traded during a period of market stress.

The other paper related to swap markets analyzes changes mandated by the Dodd-Frank Act to increase competition in the U.S. swap markets by requiring some standardized contracts to be traded on swap execution facilities. "Mechanism Selection and Trade Formation on Swap Execution Facilities: Evidence from Index CDS" by Lynn Riggs, an economist at the Commodity Futures Trading Commission, and coauthors analyzes the trading options provided on swap execution facilities. These facilities offer customers three options for trading swaps.

One option is similar to that of trading equity: to place a market or limit order directly on the central limit order book. However, the paper finds customers typically prefer to trade through one of two other mechanisms. First, the customer may request streaming indicative (suggestive) quotes from dealers throughout the day. These quotes are not firm, however, so a dealer may reject a customer's proposal to trade at the dealer's quote. Second, a customer may send a request for a quote to three or more dealers. In this case, a deal is final if the dealer provides such a quote and the customer accepts the proposal.

One interesting aspect of the Riggs et al. paper is its focus on how many dealers a customer contacts when the customer uses a request for a quote. The authors find that customers seem to be trading off the competitive advantages of contacting more dealers against dealers' concern about the so-called winner's curse, that is, the risk the dealer who submits the winning bid will turn out to have quoted a price too favorable to the customer. Consistent with such concerns the authors find that: (a) customers contact fewer dealers if they have a large or nonstandard order, (b) dealers are more likely to respond to a request for a quote if fewer dealers are contacted, and (c) the quoted spread and transactions costs are increasing in the number of dealers contacted.

Bond and stock markets
The effect of postcrisis regulatory changes of the bond and stock markets was the subject of two conference papers. Prior to the crisis, an investor who thought a stock was overpriced could either short the stock or buy a put option. If the investor bought a put option from an options market maker (OMM), the OMM would seek to offset the risk by shorting the stock. Unlike the investor, however, the OMM could naked short-sell the stock, or short-sell without having found someone who would lend the OMM the stock. That meant the OMM could effectively provide an unlimited supply of shares for those betting on the decline of a particular stock. However, the OMM's ability to do naked short sales also sometimes means the OMM would not be able to find someone who would lend the stock when it came time to settle the short sale. In these cases, the OMM would "fail to deliver" the stock. This failure came at a cost to the OMM, which was required to pay a fee to the trader. Moreover, this failure to deliver created the potential for increased risk to equity market participants if the OMM became insolvent before delivering the stock. As a result, the U.S. Securities and Exchange Commission (SEC) sought to stop OMM naked short sales with a 2008 rule that banned this activity and a 2013 rule that prohibited a way in which OMMs had gotten around the 2008 rule (called "reverse conversions").

"Converting Short Sale Constraints" by Professors Jesse A. Blocher of Vanderbilt University and Matthew C. Ringgenberg of the University of Utah examines the effect of the SEC's 2008 and 2013 rules. The authors find the rules accomplished their intended purpose, as failures to deliver were substantially lower after the implementation of the rules. However, the rules also had the effect of raising the cost of shorting some stocks, which reduces the potential returns to investors who think a stock is overpriced. As a result, the paper finds that price efficiency is lower and difficult-to-borrow stocks are more mispriced after the SEC's rules.

The corporate bond market, like the OTC swaps market, has long been a dealer market. Prior to the crisis, dealers held relatively large inventories of bonds to facilitate their trading in the market. However, some postcrisis changes in bank regulation discouraged banks from holding such large inventories. One prominent change was the Dodd-Frank Act's requirement that banks not engage in securities "speculation," called the Volcker rule. A second change is the increased leverage ratio requirement, which had the effect of forcing banks to hold more capital to support their corporate bond inventories. Bond market investors complained these regulatory changes had the effect of reducing the supply of liquidity in the bond market and pointed to decreases in bond dealers' inventories. However, others observed that some measures, such as bid-ask spreads, suggested that liquidity had not been reduced.

University of Illinois at Urbana-Champaign Professor Jaewon Choi and Federal Reserve Board economist Yesol Huh analyze changes in corporate bond liquidity in "Customer Liquidity Provision: Implications for Corporate Bond Transaction Costs." The paper posits that the lack of change in bond spreads could be misleading. Prior to the regulatory changes, bond dealers provided most of the liquidity by being willing to buy or sell upon request from their customer. However, since the regulatory changes, far more of the liquidity is being supplied by other bond investors. For example, a dealer may respond to a customer request to buy a particular corporate bond by seeking out another customer willing to sell the bond. If the dealer succeeds, the reported transaction will typically have a lower spread than if the dealer had supplied the liquidity. The flip side is the customer will often have to wait longer to complete the transaction. Once the paper adjusts for cases where the customers are supplying liquidity to each other, it finds that spreads have increased by 35 to 50 percent in those cases where the dealer is supplying the liquidity.

Conclusion
In the years since the financial crisis, regulatory agencies around the world have taken a number of steps to reduce the probability of another crisis. The Atlanta Fed workshop Financial Regulation: Fit for the Future? sheds new insight into the many ways these regulations are affecting the financial sector.

Larry D. Wall is executive director of the Center for Financial Innovation and Stability at the Atlanta Fed. The author thanks Scott Frame for helpful comments. The view expressed here are the author's and not necessarily those of the Federal Reserve Bank of Atlanta or the Federal Reserve System. If you wish to comment on this post, please email atl.nftv.mailbox@atl.frb.org.

_______________________________________

1 Equity to asset (or leverage) ratios for European banks are reported in the June 2017 Global Capital Index.

2 See a post by Atlanta Fed economist W. Scott Frame for a discussion of the evidence that prior to the crisis, the securitized mortgage loans were riskier than those held in portfolio.

3 The Financial Conduct Authority provides a discussion of swap reporting obligations in the United Kingdom. Similarly, the Securities and Exchange Commission and the Commodity Futures Trading Commission provide posts on their respective regulations on swap reporting. See an Economic Perspectives article by Chicago Fed economists Ivana Ruffini and Robert S. Steigerwald for a discussion of the market infrastructure, including some discussion of swap execution facilities.