The following remarks were delivered by Commissioner Daniel M. Gallagher of the U.S. Securities and Exchange Commission in Washington D.C. on January 15, 2014. 

Thank you, Sarah [Kelsey, Exchequer Club Secretary], for that introduction.  I’m very pleased to be here this afternoon.

Today, I’d like to talk about regulatory capital.  Given the usual reaction I get when I raise this subject, just to be safe, I’ve barred the exits!

In all seriousness, though, there’s been a great deal of attention paid to regulatory capital recently, including new Dodd-Frank requirements, Basel III implementation (or non-implementation) issues, and even bipartisan Congressional efforts to raise capital requirements for large banks.[1]  Almost all of that attention has naturally centered on the question of how much capital a financial institution should be required to hold.  What’s missing from the conversation, however – and what I’d like to focus on today – is a proper understanding of the theories behind capital requirements, both for banks and for non-bank financial institutions.

You may have noticed my reference to the theories behind capital requirements, rather than a single theory.  If so, you’re one step ahead of many policymakers both here and abroad, who often implicitly or explicitly advance a single, one-size-fits-all approach to capital.  As I’ll explain, this is a mistaken view that has the potential to impact the U.S. economy.

Understanding capital requirements begins with addressing the fundamental question of why financial institutions need minimum capital levels.  In the banking sector, where the regulated entities operate in a principal capacity and are leveraged institutions, capital requirements are rightly designed with the paramount goal of enhancing safety and soundness, both for individual banks and for the banking system as a whole.  Bank capital requirements serve as an important cushion against unexpected losses.  They incentivize banks to operate in a prudent manner by placing the bank owners’ equity at risk in the event of a failure.  They serve, in short, to reduce risk and protect against failure, and they reduce the potential that taxpayers would be required to backstop the bank in a time of stress.

Capital requirements for broker-dealers, however, serve a different purpose.  In the capital markets, we want investors and institutions to take risks – informed risks that they freely choose in pursuit of a return on their investments.  Eliminate the risk of an investment, and you eliminate the opportunity for a return as well.  Capital markets, in short, are predicated on risk.

Whereas bank capital requirements are based on the avoidance of failure, broker-dealer capital requirements are designed to manage failure by providing enough of a cushion to ensure that a failed broker-dealer can liquidate in an orderly manner, allowing for the orderly transfer of customer assets to another broker-dealer.

These two models of capital requirements, in other words, differ in fundamental ways – it’s certainly not a matter of comparing apples to apples.  Applying bank-based capital requirements to non-bank financial entities, in fact, is rather like trying to manage an orange grove using apple orchard techniques – it’s the equivalent of trying to determine how best to grow oranges to be used in orange pie, orangesauce, and, as a special treat, delicious caramel oranges on a stick.  If you think that metaphor is a bit strained, well, it is – but nowhere near as strained as imposing a bank capital regime on broker-dealers.

In order to fully understand the danger of imposing bank capital requirements on non-bank institutions, it’s helpful to take a bit of a detour to review the actions of the Federal Reserve during the height of the financial crisis, which leads us to that dreaded word: bailouts.

The word “bailout,” of course, has come to stand for everything wrong with the federal government’s response to the financial crisis.  As with, I imagine, everyone except for bailout recipients themselves, I find the idea of using taxpayer money to prop up insolvent financial institutions repugnant.  There’s a basic ant-and-grasshopper dynamic to bailouts: we naturally recoil from the idea of using the resources of prudent taxpayers to rescue institutions felled by their lack of prudence.  So let’s be absolutely clear: I hate bailouts.  We should all hate bailouts.  Case closed.

But…what, exactly, is a bailout?  Notwithstanding the risk of being misunderstood on an incredibly sensitive topic, I believe it is critically important to understand what is, and what is not, a bailout.  And here we come to the concept of the Federal Reserve as the lender of last resort and the crucial difference between insolvency andilliquidity for financial institutions.

The Federal Reserve Act of 1913 established the Fed, through its use of the discount window, as the nation’s lender of last resort.  The best starting point for understanding the concept of a lender of last resort remains Walter Bagehot’s seminal work Lombard Street.  Writing in 1873, Bagehot, who may be familiar to you from his work as editor-in-chief of The Economist or his treatise on the English constitution, set forth what is sometimes referred to as “Bagehot’s Dictum.”  My friend Paul Tucker, former Deputy Governor of the Bank of England, succinctly summarized Bagehot’s Dictum as follows: “to avert panic, central banks should lend early and freely…to solvent firms, against good collateral, and at ‘high rates.’”[2]

As you may recall me noting, I’m starkly against bailouts.  But offering access to the discount window to illiquid, but not insolvent, banks against good collateral comports with the traditional role of a central bank as the lender of last resort and falls outside even an expansive definition of the dreaded concept of a bailout.  Indeed, it falls squarely within the traditional understanding of a central bank’s paramount purpose.

In 2008, however, the Fed went well beyond offering access to the discount window to depository institutions in its capacity as the lender of last resort.  Instead, what happened in 2008 was that the Fed became the investor of last resort, a tremendously different concept which does indeed lend itself to the terrible title of “bailout.”  The acquisition of almost 80 percent of AIG in exchange for an $85 billion loan, for example, as well as the ownership of $29 billion in former Bear Stearns assets, marked a fundamental departure from the Fed’s traditional role.  As explained by Professor Allan Meltzer, author of a history of the Federal Reserve, “This is unique, and the Fed has never done something like this before.  If you go all the way back to 1921, when farms were failing and Congress was leaning on the Fed to bail them out, the Fed always said, ‘It’s not our business.’ It never regarded itself as an all-purpose agency.” [3] One reporter aptly deemed the Fed’s actions in the financial crisis as a transformation into “The Fed Inc.”[4]

As the Fed explains on its web page detailing its “Mission,”[5] in an amusingly understated manner, “Over the years, its role in banking and the economy has expanded.”[6]  The Fed describes its current duties as conducting the nation’s monetary policy, supervising and regulating banking institutions, maintaining the stability of the financial system and providing financial services to depository institutions, the U.S. government, and foreign official institutions.[7]  Notwithstanding the breadth of this mandate and the full plate of work you’d expect it to engender, the Fed has also taken steps to extend its regulatory paradigm – designed, once again, to prevent bank failures – to non-bank institutions as well.  Such institutions include broker-dealers, which, as I noted earlier, have their own regulatory capital regime that is designed to manage, rather than prevent, failure in order to ensure the return of customer assets. In addition, Title II of Dodd-Frank was explicitly designed not to prevent failure, but instead to manage the liquidation of large, complex financial institutions close to failure – indeed, the very name of the Title is “Orderly Liquidation Authority.”

In light of this, a more cynical person might suggest that the Fed’s efforts to extend the failure-prevention paradigm of bank capital regulation to financial entities that are already subject to failure-management regulatory schemes implies an institutionalization of the concept of too-big-to-fail.  Good thing I’m not a cynical guy.

I digress, but it is important to remember that the Fed’s lender of last resort activity during the financial crisis came after its intervention in the Bear failure as well as its bailout of AIG.  To be fair, once the Fed resumed its traditional role as the lender, rather than investor, of last resort, it did so robustly.  By March 2009, the Fed had lent a staggering $7.7 trillion dollars to beleaguered financial institutions, including $1.2 trillion on one day alone on December 5, 2008.[8]  And you thought your holiday spending was high!

So what does all of this have to do with capital?  To answer that question requires a better understanding of the recent and disturbing fascination with imposing bank-theory capital requirements on non-bank institutions.  Here, the recent FSOC intervention in the money market mutual fund space is quite instructive.

In August 2012, a lack of consensus among the Commission on the best way to proceed with proposing reforms to our money market fund rules led to an ill-advised abdication of the issue to FSOC, which enthusiastically took up the cause, leading to an unprecedented — albeit invited — incursion into the regulatory purview of an independent regulator.  The result was the issuance, in November 2012, of a report entitled “Proposed Recommendations Regarding Money Market Mutual Fund Reform,” in which FSOC floated – pun intended – the concept of a “NAV buffer,” that is, a capital requirement for money market funds.[9]

As I delved into the issue of money market fund reform following my return to the SEC as a Commissioner, it quickly became apparent to me that, perhaps in the hopes of staving off more stringent regulation, the industry was coalescing behind a capital buffer requirement of approximately 50 basis points, to be phased in over a several year period.  For the largest money market funds, this would have resulted in an approximately 1 to 200 ratio – a $500 million buffer to support $100 billion in investments.  This would amount to chicken feed in any serious capital adequacy determinations.

The ostensible reasoning behind a capital buffer for money market funds is that it would serve to mitigate the risk of investor panic leading to a run on a fund.  Common sense, however, belies this notion.  Do we really believe that investor panic would be assuaged by the comforting knowledge that for every one dollar they had on deposit, the money market fund had set aside half a penny?

Common sense also leads to the conclusion that there is no reason to assume that this view of capital requirements as a panacea to mitigate run risk is limited to money market funds.  Indeed, the now notorious “Asset Management and Financial Stability” report issued by Treasury’s Office of Financial Research last September featured similar reasoning, as reflected in its implied support for “liquidity buffers” for asset managers.[10]

As I noted in my statement at last June’s open meeting at which the Commission voted to propose reforms to our money market fund rules, which by the way thankfully did not include a capital buffer, “It became clear to me early on in this process that the only real purpose for the proposed buffer was to serve as the price of entry into an emergency lending facility that the Federal Reserve could construct during any future crisis – in short, the “buffer” would provide additional collateral to facilitate a Fed bailout for troubled MMFs.”[11]

Indeed, some Fed officials and academics[12] have suggested as much.  In a speech delivered last February, New York Fed President Bill Dudley, while expressing support for the FSOC-proposed money market fund reform mechanisms of a NAV buffer and a “minimum balance at risk,” explained his concern that “even after such reforms, we would still have a system in which a very significant share of financial intermediation activity vital to the economy takes place in markets and through institutions that have no direct access to an effective lender of last resort backstop.”[13]  He went on to raise the possibility of expanding access to the lender of last resort to additional entities in exchange for “the right quid pro quo—the commensurate expansion in the scope of prudential oversight.” Arguing that “[s]ubstantial prudential regulation of entities—such as broker-dealers —that might gain access to an expanded lender of last resort would be required to mitigate moral hazard problems,” he concluded, “Extension of discount window-type access to a set of nonbank institutions would therefore have to go hand-in-hand with prudential regulation of these institutions.”[14]

Fed Governor Daniel Tarullo, on the other hand, indicated his discomfort with extending access to the discount window to non-bank entities in a speech last November, noting that he was “wary of any such extension of the government safety net.”  In the context of addressing the “vulnerabilities” of short-term wholesale funding, he stated that he “would prefer a regulatory approach that requires market actors using or extending short-term wholesale funding to internalize the social costs of those forms of funding”[15] – that is, an increased capital charge. In a different speech earlier last year, he cited the risk of “regulatory arbitrage” if increased capital charges were applied “only to some types of wholesale funding, or only to that used by prudentially regulated entities”[16] and concluded, “Ideally, the regulatory charge should apply whether the borrower is a commercial bank, broker-dealer, agency Real Estate Investment Trust (REIT), or hedge fund.”[17]

All of this adds up to a terribly muddled situation.  Is the Fed seeking to impose bank-based capital charges on non-bank entities in conjunction with granting them access to the discount window at the cost of submitting to prudential regulation, as Mr. Dudley suggests?  Or is the situation just the opposite, as Governor Tarullo implies – would those additional capital charges be intended to prevent non-prudentially regulated financial entities from ever relying upon, as Governor Tarullo puts it, an extension of the “government safety net” the discount window provides?

Put another way, is the goal to expand the Fed’s role by making it the lender of last resort to non-bank entities such as money market funds and broker dealers, or is it to use its Bank Holding Company Act authority and its role in FSOC to dictate capital requirements to non-bank entities in order to prevent those entities from ever gaining access to the discount window?

These are more than purely semantic questions, although semantics play a role: one man’s expansion of the Fed’s role as the lender of last resort is another man’s institutionalization of bailouts for failing financial institutions.

In my opinion, both Governor Tarullo and Mr. Dudley raise very good points that warrant a healthy debate.  The issues they raise, however, as well as the more general issue of how much capital is enough in the banking and capital markets, create a degree of confusion about the Fed’s role as the lender of last resort.  Should the Fed still perform that role?  If so, when and for what entities?  Does such lending, in fact, constitute a bailout?

All of these questions require answers as we debate questions of capital adequacy.  If we are to assume that the Fed will not, or cannot, expand its role as the lender of last resort to non-bank entities, including non-bank subsidiaries of bank holding companies, would it ever be possible to set capital requirements at a level that would guarantee avoidance of 2008-type scenarios?  I think not, even if we were to impose capital requirements of 100%.  To me, therefore, capital markets regulators simply cannot stray from the theory of capital as a tool to facilitate the unwinding of a failed firm with the goal of returning customer assets.

I certainly don’t want to leave the impression that I disregard the Fed’s concerns about capital requirements for bank affiliated non-bank financial institutions.  Indeed, it is my hope in the coming year to work with Commission staff and FINRA to begin an in-depth review of whether it would be appropriate to establish separate capital rules for bank-affiliated broker-dealers.  If we determine that such a bifurcated broker-dealer capital regime would be appropriate, however, any such regime would be based on the principles of our current program for broker-dealer net capital, and it would be crafted to stand on its own, without any reference to the discount window.  On this and related issues, it is far past time that the SEC play an active role in the policy debate in order to ensure the ongoing vibrancy of our capital markets.

Before I conclude, I’d like to make one final point that is obvious but still needs to be reiterated: the judgment calls regulators make in establishing capital rules incentivize regulated entities in a manner that inevitably results in unforeseen (although often quite foreseeable) externalities.  A classic example is the beneficial capital treatment provided to certain asset-back securities under what’s known as the “Recourse Rule.”[18]  The Recourse Rule, issued by the Fed, the FDIC, the OCC and OTS as a supplement to their implementation of Basel I, hugely privileged highly rated ABS as well as ABS issued or guaranteed by a GSE.  How hugely? Well, for every $100 of highly rated or GSE sponsored ABS, well-capitalized banks had to set aside $2.  This compared to a $5 set-aside for unsecuritized mortgage loans and a $10 charge for commercial loans or corporate bonds.  In other words, by holding ABS from these favored categories, banks could reduce their capital requirements by 60% compared to holding an equivalent amount of mortgage loans and by 80% compared to holding corporate loans or bonds.

I’m reminded of the ubiquitous TV commercials featuring children’s responses to simple questions: what’s better, bigger or smaller, faster or slower, more or less?  As the ads say, it’s not complicated.  Concluding that setting aside less capital was better than setting aside more capital, banks loaded up their balance sheets accordingly, and by 2008, a staggering 93 percent of all the mortgage-backed securities held by American banks were either GSE-issued or rated AAA.[19]  As noted in a 2010 report issued by the American Enterprise Institute, “If not for the recourse rule’s privileging of mortgage-backed bonds, the burst housing bubble almost certainly would not have caused a banking crisis. The banking crisis, in turn, froze lending and caused the Great Recession.”[20]

As the Recourse Rule illustrates, regulatory capital requirements play a tremendous role in incentivizing financial institutions’ holdings.  All the more important, therefore, that regulators use the right tool for the right job.  We rightly take great pride in our capital markets, the deepest and safest in the world.  We’re an entrepreneurial nation, and taking risks, whether with respect to investments or otherwise, is as American as apple pie.  Superimposing upon those markets a capital regime based on the safety-and-soundness banking paradigm, on the other hand, would be as sensible as orange pie.

Thank you all for your attention this afternoon.  I’d be happy to take questions.