Progress and Confusion
Chapter

8. Shadow Banking as a Source of Systemic Risk

Editor(s):
Olivier Blanchard, Kenneth Rogoff, and Raghuram Rajan
Published Date:
April 2016
Share
  • ShareShare
Show Summary Details
Author(s)
Robert E. Rubin

My colleagues on the panel are highly respected and deeply experienced experts on central bank issues. In this chapter I express my views as a practitioner, and those views are based on my experience at Goldman Sachs, where I had responsibility for some, and then all, of the firm’s trading and arbitrage activities; my years in the Clinton administration; and my present activity advising several investment organizations and as a continuing participant in the national policy dialogue. I will focus on three issues: the possibility of market excesses, the likelihood of future market and financial destabilization, and systemic risk in shadow banking. A few other issues I cannot take up because of time limitations I will list anyway, because they are on my mind and because I think they are important to anyone involved with investment or policy.

These issues include the monetary policy risks in having expanded the Fed balance sheet so greatly; the fallacy, in my view, of the argument that these risks can be largely avoided by tightening through increasing interest rates on excess reserves or using reverse repos—what I call the magic wand argument; the substantial limitations on the ability of econometric models to predict future economic conditions or responses to policy; and finally, the daunting and hugely consequential question of where the combination of dysfunctional government in the United States, the euro zone, Japan, and elsewhere on fiscal and structural issues joined with expansive monetary policy will lead for the currencies and for financial and economic conditions in the United States and around the world.

Now I will turn to my three focal topics, starting with excesses. I would guess that we would all agree that there has been a substantial reaching for yield along the risk curve for some years now, owing to the low interest rates on US Treasuries, which I think were predominantly a result of multiple factors other than QE2 and QE3, which I think had limited effect on rates. Having said that, I do think QE2, and even more QE3, heightened the reaching for yield by creating a sense of comfort in the financial arena that the Fed could and would keep rates down.

I don’t have a judgment as to whether this extensive reaching for yield has, in fact, led to excesses. But that is a realistic possibility, with the US stock market at roughly all-time highs, the revival of covenant light and even noncovenanted lending, the vast increase in fixed income ETFs, euro-zone sovereign debt in the troubled countries selling at what to me, at least, is inexplicable yields on a risk-adjusted basis, and much else. And if there are excesses, they will inevitably fall of their own weight at some unpredictable time.

However, even if markets are not broadly in excess now relative to long-term fundamentals, they will be periodically and unpredictably in the future. In my view, markets are a psychological phenomenon in the short term, oscillating between the fear and greed rooted in human nature, while on average reflecting fundamentals over the longer term. Market-based financial systems over the centuries have always experienced periodic excesses on the upside and then, in reaction, overshooting on the downside, when excesses inevitably fall of their own weight. And I don’t see any reason to think that all of human history with respect to market-based financial systems should change now. Thus, I think there is a high likelihood of periodic market and financial system destabilization at unpredictable times in the future.

Moreover, future destabilization may begin—and, almost by definition, will begin—in unexpected places. Goldman Sachs was almost destroyed by the unexpected bankruptcy of Penn Central in 1970. We then put in place measures to prevent any similar situation posing such a dire threat again. But I have never forgotten the comment of John Whitehead, one of our senior partners. He said that our actions would prevent a future Penn Central–type crisis for our firm, but that the next crisis would come from some totally unexpected place, and beyond that, there would surely be future crises. If you look at the history of market-based financial systems, that is what has happened repeatedly.

I don’t believe that regulation will ever succeed in preventing excesses, and therefore significant market and financial cyclicality, as long as we have a market-based financial system. But regulation can and should try to reduce the probability of excesses, the likely severity of excesses, and the market and economic effects of downturns when they occur, such as by reducing the vulnerability of organizations involved with the financial system through constraining leverage.

The reforms put in place in response to the financial crisis—for example, the Consumer Financial Protection Bureau, the measures on derivatives, and increased capital requirements—have presumably in large measure accomplished those purposes with respect to banks, other than for the too-big-to-fail issue. In my view, nothing yet done or proposed would solve this important issue under conditions of serious systemwide duress.

More recently, attention has increasingly turned to shadow banking, and so will my remarks. The shadow banking system has long existed, but now many functions of traditional banks are rapidly and substantially gravitating toward the shadow banking world, in part because of the increased constraints on banks and dramatic changes in technology.

For example, market-making, and therefore market liquidity, has always been greatly facilitated by market-makers creating profit-seeking constructions around positions acquired through market-making activities. With proprietary trading now barred for banks, and with the great difficulties in distinguishing between market-making and proprietary trading, bank market-making and market liquidity have declined substantially. And both market-making and now, at an incipient level, the provision of capital to meet primary issuance demand are moving to multiple other platforms, including hedge funds, broadly defined.

Similarly, the increase in capital requirements has deterred various kinds of smaller and medium-sized lending by banks, and this regulatory constraint, plus the capacity for credit evaluation and for interaction with borrowers created by big data and other technological developments, is leading to rapid growth of this type of credit extension by nonbanks, including private equity funds and organizations established for these purposes.

More broadly, the shadow banking world involves a vast array of institutions, asset classes, and activities, and it is growing rapidly. There seems an enormous potential for systemic risk in this world. Let me briefly mention three examples. If excesses develop in financial assets, sooner or later those asset prices will destabilize, and hedge funds and other asset managers, who are highly sensitive to short-term results, could engage in a rush to the exit with those assets, and even with good assets, in order to increase liquidity. That could trigger broad financial market duress. Moreover, while leverage obviously exacerbates the pressure to liquidate, that can occur even when leverage is limited.

Another example is the vast increase in the size of fixed income ETFs. They promise constant liquidity, but the assets of those funds might not be salable in an orderly fashion in times of significant market stress. Heavy redemptions could thus lead to highly destabilizing market dumping. As just one more example, there are new regulations with respect to derivatives, and while they are useful, I don’t think they get to the heart of the matter. The outstandings in derivatives are vast. Under normal conditions, that all works. But in times of market stress, correlations move in all kinds of unpredictable ways, and many participants find they have risks different from, and multiples greater than, they had expected. And this could lead to serious market and economic duress, and to unexpected counterparty credit failures. I published a book in 2003 in which I suggested that margin and capital requirements be greatly increased for all users of derivatives—a view I had held going back to my days at Goldman Sachs—and I still think that is critically important, to reduce derivative use, to provide a larger cushion, and to better protect against systemic risk.

I could go on endlessly listing asset classes, types of organizations, and types of activities in shadow banking that could generate systemic risk. Moreover, that situation is not limited to the United States. In today’s instantaneously interconnected world, systemic risks in shadow banking abroad can rapidly and powerfully affect us.

One regulatory answer put forth in response to this systemic risk in shadow banking is macroprudential regulation. I interpret that term, in this context, as financial regulations for shadow banking analogous conceptually—though not necessarily with respect to specifics—to financial regulations for the banking system. However, despite work going on with respect to some aspects of shadow banking, in very large measure, as far as I can see, this challenge still needs to be met. In my view, meaningful macroprudential regulation requires meeting three hugely complex and highly time-consuming challenges:

  • 1. Creating a comprehensive catalogue of the shadow banking world, including asset classes, types of organizations, and types of activities. Today, to the best of my knowledge, no one comes close to having identified the full reach of shadow banking or the systemic risks it poses, and doing so will involve a lot of ambiguities and uncertainties. Moreover, while this project would be a monumental undertaking for the United States alone, it should also, perhaps as a separate undertaking, at some point, include the significant shadow banking systems elsewhere.

  • 2. Developing a menu of tools to address the possible systemic risks posed by the shadow banking world. This menu might include, as examples to consider but not recommendations, capital changes, margin requirements, leverage constraints, and possibly position or concentration limits of some sort, or measures tailored to particular shadow banking issues. At best, these tools will be quite imperfect, given the complexities of the various aspects of shadow banking and the complexities around the effects of regulatory measures under conditions of systemic stress. In this context of regulation, while asset classes and organizations are important, I would probably focus especially on activities because systemically risky activities can occur within institutions that are sound and cause serious trouble in the financial system. For example, a large asset manager could be strong enough to withstand very large shocks but have derivative activities, hedge funds, ETFs, or money market funds that could, under stressful conditions, trigger or contribute to serious risks for the financial system.

  • 3. Devising a plan for effective implementation and coordination of macroprudential regulation, once the first two challenges are met. This third challenge is daunting, given our large number of regulatory institutions and the Federal Stability Oversight Council’s limited powers. It becomes even more daunting once shadow banking outside of our borders is considered.

In theory, the Office of Financial Research, housed in the Treasury Department, is empowered to undertake these studies. But as a practical matter, far greater resources will be needed, and other regulatory institutions, especially the Fed, could contribute capacity. Perhaps the best way forward is first to recognize the imperative for these studies, and then have the FSOC develop a collaborative plan to get it done. But that, I suspect, is a lot easier said than done.

I think the timeframe for accomplishing this whole project will be lengthy. However, each of the three components could be divided into subsets, in order to move forward in some areas while continuing to work on others. Also, even a strong program would provide far from perfect protections, given the uncertainties and unpredictability in markets, the wide swath of shadow banking, the enormous complexities with respect to each of the three steps, and the inevitability of unforeseen developments.

Thus, I believe, contrary to the views of many, that the Fed should take systemic risk into consideration in monetary policy decisions, even though excesses and bubbles are impossible to identify with high convictions except ex post. That doesn’t mean that the formal criteria should be expanded from the current two, combating inflation and promoting full employment, but simply that systemic risk should be considered.

I’ll make one final observation: all judgments relating to markets and financial systems are obviously about probabilities, though, as Stan Fischer once said to me, while every thoughtful decision maker recognizes that, relatively few have deeply internalized that probabilistic mind-set and actually operate that way. Moreover, the judgments about probabilities with respect to markets, policy responses, and the like should be treated as having a great deal of uncertainty about them, and that uncertainty itself should be appropriately weighed in decision making. Thus, creating strong regulatory protection, and measuring costs versus benefits, should allow for the possibility that developments, especially under stress conditions, will be at substantial variance with the most likely projected case and may even be outside the range of projected possibilities. Effectively regulating shadow banking is imperative, and it is a complex challenge.

    Other Resources Citing This Publication