This guest post was submitted by Peter Fox-Penner, a leading expert on regulation at The Brattle Group. The views expressed herein are those of the author alone.
At present, the debate among economists over whether our financial regulations should protect institutions on the basis that they are “too big to fail” (TBTF) still rages. Like many other economists, I distrust the reasoning behind the TBTF justification and rue the fact that the measures taken to prop up the U.S. financial system have made the largest banks even larger, while small banks are failing at record levels. In my first guest post I argued patchwork attempts to strengthen financial regulation without a “clean sheet” review were likely to be inadequate.
In this second post I look past short term bailouts and address the broader issue of establishing regulation of TBTF firms. Policymakers are faced with challenge of establishing a large regulator that retains the specialized expertise needed to manage complex markets – specialization more often found in a network of smaller agencies. To do so they will need to address the size and complexity of the financial sector itself. As before, I turn to examples from the utility industry, specifically the establishment and repeal of the Public Utility Holding Company Act of 1935 (PUHCA), that provide lessons for crafting regulation of complex industries.
There should be no question that any firm considered too big to fail must be regulated thoroughly and effectively. This is unquestionably the $64 trillion question. We cannot give an explicit or implicit government guarantee to rescue a private firm without whatever control is necessary to prevent moral hazard. But what if the need to impose appropriate regulation itself becomes a limit on the allowable size or complexity of firms? In short, can a firm be too big to regulate?
In any such discussion it is important first to distinguish between several dimensions of big. The first dimension is the multiplicity of products and markets in which a company is involved. In context, this means companies that operate in commercial and investment banking, trading, insurance, and dozens of other product lines that are all “financial,” but which have different attributes, externalities, and regulatory requirements. Call this “firm complexity”.
“Organizational complexity”, the second dimension, tends to follow firm complexity, but it isn’t the same. This dimension has to do with the number and diversity of business structures owned by a single parent company. The larger and more varied the number of corporate entities owned or controlled by a single parent, the more layers of vertical ownership; and the more complex the cross-ownership claims, the more complicated the structure.
The third and final dimension is “structural bigness”. This is about having a large market share in a well-defined product and geographic market. This is the traditional meaning of big in the industrial organization field of economics. It is bigness within a market, or market dominance.
For most financial (as well as non-financial) products, structural bigness is policed by competition (antitrust) laws. (The insurance industry has enjoyed an antitrust exemption that Congress is now reconsidering.) Because I assume these laws will continue to be enforced, I assume structural bigness is not a factor in making a firm too big to regulate.
For the other two dimensions of bigness, product and organizational complexity, it’s a different story. In a nutshell, when the range of one firm’s market and organizational activities grows too large, it often becomes politically or administratively impossible to do a good job regulating it, whatever the particular tools and processes regulators use.
Firm complexity challenges regulation because it requires regulatory agencies with enormous resources to understand the linkages between extremely different financial product markets. To regulate a firm engaged in insurance, banking, investment banking, trading, and other products, a single regulatory agency will have to possess an unbelievably broad range of skills, tools, and resources. This is not to deny that there is a careful balance to be struck between too much and too little regulatory overlap, acknowledging sometimes conflicting problems such as regulatory capture and forum shopping.
It is politically unimaginable that the U.S. would ever establish a single financial sector super-regulator (an outcome far beyond the Fed getting the job of policing systematic risk). However, if we will rely on multiple specialized agencies to police one market at a time, then we have the current patchwork quilt that never allows narrowly specialized regulators to see the linkages between markets. It also encourages firms to create products that fall in the cracks between agency jurisdictions and that no agency understands well enough to monitor. Placing limits on the number of markets a firm can operate in may reduce synergies and scale economies, but there should be no argument that it makes regulation far more likely to succeed.
For purely practical reasons, organizational complexity also makes regulation ineffective. As businesses get successively more complex and varied business structures, the ability of regulatory agencies to understand the company’s financial position simply fades away. It is well-documented, for example, that Enron built a financial structure so complex that regulators could never understand what it was up to, even following its downfall. To cite one example, when the investigative staff of the U.S. Federal Energy Regulatory Commission (FERC) was directed to look into Enron’s electricity trading practices, here’s what they had to contend with:
The complexity of the issues confronting Staff and the agencies cooperating with the Commission is such that more time would be required to fully understand Enron’s and other market participants’ activities in the energy markets. For example, we spent a considerable amount of time analyzing Enron’s massive information technology (IT) systems that were used to harness information and use such information for Enron’s advantage. In short, the IT systems were functionally equivalent to the IT systems of a national trading exchange, e.g., a stock exchange, coupled with the credit and risk systems of a large national bank, and linked to a large telecom company. The IT systems were designed to keep transactional data, such as a telecom IT system much do with telephone service customers such as customer service, billing, scheduling, and provisioning, but also link it to a sophisticated, on-line trading platform, and calculate the credit and risk exposure of each transaction. Because Enron traded 1700 different products on-line around the world, the trading had to be linked together in a secure manner.
Although Staff has focused its energies on relevant data, the size of the task is enormous. For example, as described herein, Staff is now reviewing approximately 1.8 terabytes (TB) of data, which is equivalent to the amount of data produced by a large telecom company. In addition, because the data had to be easily accessible to Enron employees, we are also reviewing nearly 1,000 spreadsheets that were populated with data from the IT systems. The spreadsheets were approximately 40 megabytes (MB) each and dozens were created daily.[i]
Fortunately for the FERC, its objectives in the investigations were confined to Enron’s role in the Western power crisis of 2000-2001. No full accounting of Enron’s actions was sought, and none has yet been produced.
The question of the size and institutional division of the analytic resources needed to properly regulate financial markets was the subject of my prior guest post. In this area, it is interesting that a National Institute of Finance has been proposed to create an independent body of expertise usable by regulators. Other regulated industries have this, such as the National Regulatory Research Institute, the Institute of Public Utilities at Michigan State University (MSU), and the Regulatory Assistance Project for energy regulators.
Structural Limits in the Utility Industry
The energy utility industry has struggled to find the right balance between enabling effective regulation and allowing market and organizational complexity. Starting in the late 1880s, municipalities and small industrial firms began installing electric systems that were seldom larger than a handful of plants and lines. In the first part of the 20th century, industrial titans like J.P. Morgan and Samuel Insull did what we would now call a roll-up: they purchased dozens of small systems and assembled them into massive holding companies.
Although most of the local subsidiaries of these utilities had rates set by state or local regulators, these agencies could not understand or control enormous multistate holding companies. With massive, complicated holding companies, regulators could not determine the true cost of serving customers and therefore could not determine the true cost of serving customers and therefore could not set good cost-based rates. Complexity gave holding companies the tools to overleverage their companies, mislead investors and regulators, overcharge regulated, captive customers and subsidize unregulated lines of business.
Following the ’29 crash, Congress asked the U.S. Federal Trade Commission (FTC) to investigate the financial practices of utility holding companies. The 101-volume FTC report found 19 categories of financial misdeeds, including:
. . . the issuance of securities to the public that were based on unsound asset values or on paper profits from intercompany transactions; the extension of holding company ownership to disparate, nonintegrated operating utilities throughout the country without regard to economic efficiency or coordination of management; the mismanagement and exploitation of operating subsidiaries of holding companies through excessive service charges, excessive common stock dividends, upstream loads and an excessive proportion of senior securities; and the use of the holding company to evade state regulation.[ii]
In what one historian called “the most bitter legislative battle of Roosevelt’s first term,” Congress passed the Public Utility Holding Company Act of 1935, known as PUHCA. PUHCA essentially banned complex financial holding company structures for utilities, gave the new Securities and Exchange Commission (SEC) authority to approve utility mergers and security issuances, and made it extremely hard for utilities to buy or engage in non-utility lines of business. Drawing on the notion that utilities had large scale efficiencies, but only if systems were physically connected, it also required that all mergers between utilities created integrated systems.
These strong restrictions were based on a consensus that it was simply not realistic to try and regulate the rates and securities of extremely complicated holding company structures. It was “the very heart of the title,” said the Senate Report accompanying PUHCA, to “simply…provide a mechanism to create conditions under which effective Federal and State regulation will be possible.”[iii] Over the next 11 years, under federal supervision, the large holding companies were slowly divided and sold.
Although it is likely that the utility industry lost some efficiencies from the bright-line prohibitions introduced by PUHCA, there is little question that increases in efficiency continued. Over the fifty years following the passage of the Act, the industry grew by a factor of 100, the average power plant grew more efficient by a factor of five, and pollutant emissions (other than CO2) declined by a factor of 10,000. Real electricity rates dropped nearly continuously throughout the period. And if the industry was missing scale economies, this was surely more the case in the third of the industry that was publicly owned or a cooperative, where are still nearly 3,000 separate systems, as opposed to only about 140 investor-owned firms.
Many a utility CEO loathed the shackles PUHCA put in place. Utilities with interstate operations could rarely get permission to diversify into non-utility businesses, thus hampering shareholder growth. Utilities that did not cross state lines were exempt from PUHCA, and these intrastate firms often dabbled in insurance, real estate development, fuel production, and other non-utility lines of business.
From this experience, we have something of a laboratory comparing utilities allowed to diversify and utilities that were not. I think it is a fair to say that the utility economics literature has not found large gains in new product synergies where PUHCA was not binding. Utilities inside a single state went through a wave of diversification in the 1970s, were largely unsuccessful, and have since primarily abandoned non-utility businesses. It could not have been costless, but PUHCA appeared to have its intended effect, making regulation relatively simple, practical and effective until the late 1970s.
As deregulation of utilities (along with financial markets) gained currency in the 1980s, the industry pushed to repeal what it felt was an aging and unnecessary statute. The arguments for repeal were good. Financial regulation was far more comprehensive and seasoned by the 1990s than it was in 1935, with almost no one seeing the decay in its effectiveness now so apparent. Federal and state utility regulation was also now far more accomplished. Most of all, the requirement for physical integration made no sense in an industry that was intentionally de-integrating in order to introduce competition in parts of the sector. The economic structure of the industry was at odds with the Act.
PUHCA was repealed in the 1992 energy bill, but many of its provisions, meant to protect against excessively complex utilities, were essentially transferred to the FERC and state regulators from the SEC. Federal and state regulators were guaranteed access to utility holding companies’ books and records, and the FERC cannot allow a merger that impairs the effectiveness of state regulation. Many state regulators discourage their regulated utilities from engaging in non-utility lines of business, and nearly all of them require a separation between non-utility and utility assets so that losses in unregulated lines of business don’t affect the financial viability or rates of the regulated part of the firm. For a recent example, see this press release regarding the recent EDF-Constellation deal.
Technically, the removal of the physical integration requirement and other industry changes have enabled mergers and acquisitions. Between 1993 and 2007[iv] there was a wave of utility mergers, with 84 U.S. to U.S. combinations completed and a handful of acquisition of U.S. utilities by foreign companies. Warren Buffet’s holding company has purchased one utility, the Texas Pacific private equity firm recently purchased the former Texas Utilities Company, and several foreign utility holding companies now own state-regulated U.S. firms. Ownership of deregulated power generators is even more diverse.
Nonetheless, there are still strong limits on trades between utilities and non-regulated subsidiaries in a single holding company. State and federal regulators require that the regulated entity. And lately, the pace of utility mergers and acquisitions has slowed to a crawl. Many in the industry believe that this is because state regulators are starting to once again feel that utilities were becoming too big to regulate.
Relevance to Financial Regulation
Regulation of the financial sector is a vastly more complex problem than regulating electric or gas utilities. There are dozens of products deeply interconnected in several different ways in a geographically global system. It is clear this complex, interconnected system needs far better regulation than it has today.
Many of the financial regulatory reforms under discussion involve limits on “bigness” as I have defined it. The Baseline Scenario has long been a strong voice for limiting bank size based roughly on overall systemic risk; more recently, the idea seems to have been embraced by Mssrs. Volker, Greenspan, and Soros; (Two of my Brattle Group colleagues, George Oldfield and Michael Cragg, have also voiced this view. See “Life Boats for the Banks–Let the Holding Companies Swim,” in The Economists’ Voice.). Still others call for no absolute size limits, but rather higher capital requirements the larger and riskier the institution. Chairman Bernanke, for example, calls for an approach that preserves “the economic benefit of multi-function, international [financial] firms.”
Among these bigness policies, Mervyn King’s (and others’) proposal to separate commercial and investment banking (so-called utility and non-utility banks), comes closest to the policies that guide energy utility regulation today. Under King’s proposal, commercial banks would be limited in the amount of risk they could take on and would receive protection against failure, while investment banks would have fewer constraints on risk but would receive no survival guarantee
While this is analogous to energy utility regulation, and may be a good idea for financial regulatory reform, it cannot be a complete solution. Even with good separation of “boring” and “non-boring” banks, (somewhat different) regulation will be necessary for both types of firms, and close coordination will be necessary for all financial regulators. In addition to limits on “boring” banks, it may be necessary to limit, or at least oversee, the riskiness or size of “non-boring” financial entities, however they are defined. Vexing questions regarding the need for transparency, common clearing platforms, and position limits, in all financial product markets, remain. (See these comments from David Brooks, Sophia Grene, Brooke Masters, and Gillian Tett.
The interconnections between disparate financial products and markets and the need for overall prudential regulation creates an extremely complex tradeoff between multiple regulators who have the resources to specialize — and can serve as checks on each other — and the danger of missing the big picture or letting new markets and risks develop in the cracks between their jurisdictions. In this context, limits on firm size and complexity help by lowering the difficulty of assessing risks and policing activities within a single sprawling firm, in addition to reducing systematic risk. One has to wonder at passages like this in the Financial Times:
Then there is the idea of obliging the biggest, most complex banks to draw up “living wills” – certificates that would lay out a blueprint for how a bank should be wound up in the event it should fail, perhaps forcing it to ring-fence certain operations, such as its retail and investment banking, in separate subsidiaries. Regulators are determined the chaos that followed the collapse of Lehman Brothers a year ago should not be repeated. It has yet to be decided what form a living will should take, but the very notion has many banks up in arms that they will have to spend months, even years, untangling the complex corporate structures they have evolved both to comply with local regulations and to maximize the tax efficiencies.[v]
If the banks that have complete self-knowledge and control will take years to simplify their own structures, how will regulators do it during a crisis?
In short, limits on the complexity and product offerings of commercial banks may well be an essential part of the solution, but it is nowhere near the entire solution.
By Peter Fox-Penner
[i] See “Initial Report on Company-Specific Separate Proceedings and General Reevaluations; Published Natural Gas Price Data; and Enron Trading Strategies.” Docket No. PA02-2-000, Federal Energy Regulatory Commission, August 2002, p. 9.
[ii] “The Regulation of Public-Utilities Holding Companies.” Division of Investment Management, U.S. Securities and Exchange Commission, June 1995, p. 3.
[iii] Ibid, p. 9.
[iv] “EEI 2008 Financial Review,” The Edison Electric Foundation, 2008, p. 45.
[v] See Patrick Jenkins, “Banking on the Future,” The Financial Times, Future of Finance Section, October 19, 2009.