Shadow banking’s enduring perils

Five lessons from the last crisis — for managing the next one

In the immediate aftermath of the global financial crisis, most people thought that shadow banking was all in the past, and good riddance! Today, however, it is becoming clear that shadow banking is also in our future, even centrally so. The crisis was just one step toward that future, revealing weakness in order to force necessary restructuring. Of course the future won’t be the same as pre-crisis, but we can already dimly see its outlines emerging.

In retrospect, the Global Financial Crisis of 2007-2009 can be understood as a stress test of the emerging new system of market-based credit, so-called Shadow Banking, which collapsed under the stress. Central banks caught the falling knife just in time, and have been providing life support ever since (QE, ZIRP, NIRP), keeping the system alive while regulators on the one side and bankers on the other have been busy creating a new, and hopefully more robust, market structure.

Today, as the Fed celebrates “liftoff”, it is just barely possible to detect the outlines of the newly emerging market structure. It’s not a done deal by a long shot, but we can begin to connect some dots and to imagine how the new system might work once it is complete. Most important, we can begin to consider the possible resilience of that new system in the face of the next stress test.

Four dots are focusing my mind:

  1. “What Excess Reserves?” Pozsar’s Global Money Note #5
  2. RMA and SIFMA recommend Fed’s overnight bank funding rate as new Benchmark
  3. “Creators of First and Last SIV Back after the Crash.”
  4. “Market and Funding Liquidity: An Overview”

For me, shadow banking means “money market funding of capital market lending”, specifically global funding (in USD) of local lending (not necessarily USD). The process of financial globalization since WWII brought with it the separate globalization of capital markets (as asset managers overcame home bias in their search for risk-weighted return) and money markets (as borrowers overcome their own home bias in their search for the cheapest source of funding). From this historical perspective, shadow banking can be understood as merely bringing these two separate trends together, integrating global money markets with global capital markets. To the extent that financial globalization is in our future, so is shadow banking.

Pre-crisis shadow banking was a Rube Goldberg machine–remember the famous Pozsar map?–cobbled together out of repurposed bits and pieces lying around that had been originally created for other purposes. That machine was doing money market funding of capital market lending, but it was hard to see the forest for the trees. By contrast, post-crisis shadow banking is much more a creature of intentional design, by regulators as well as bankers, both trying to learn the lessons of the last crisis, inevitably imperfectly.

What follows is my list of lessons, and a preliminary assessment of the extent to which emerging market structure reflects those lessons. I offer both not as a definitive statement, but rather in the spirit of trying to see through the fog that always enshrouds the future.

Lesson One: Centrality of the Dealer Function. In a market-based credit system, the price of funding and the price of capital assets are both established in global dealer markets. Profit-seeking dealers offer trading options, to buy or sell, and absorb the resulting order flow on their own balance sheets. They may and do change the prices they quote over time, depending in part on their own net exposure, but the important point is that in doing so they make markets. This is important because a market-based credit system requires ongoing markets, i.e. market liquidity, but it does not require costless market liquidity. Regulators and bankers alike are groping their way toward mechanisms that will appropriately price market liquidity, in normal times as well as crisis times (as dot #4).

Lesson Two: Money dealing versus Risk Dealing. Money dealers buy and sell term funding while risk dealers buy and sell risk exposure. In principle both could be cash markets–say for prime commercial paper and mortgage backed securities. But they could also be derivative markets–say for overnight index swaps and interest rate swaps. (One way that liquidity is being priced is through deviation between cash and derivative prices.) The important point is that regulators and bankers alike are groping toward a system in which money dealing and risk dealing are separated legally and institutionally. You can do one, or the other, but not both. That means rather far-reaching institutional change, as existing firms shed some businesses and new firms rise up to fill the void (as dot #3).

Lesson Three: Matched book dealing versus speculative dealing. The rise of central counterparty clearing is, in effect, shifting the matched book portion of dealer positions onto the balance sheet of the central counterparty, leaving behind the (much smaller) speculative book portion. This shift is already rather far advanced in risk dealing, and is now under way in money dealing. Capital regulations apparently drove the former, while the newer liquidity coverage ratio is apparently driving the latter. The important point is that regulators and bankers alike are groping toward a system where liquidity risk and solvency risk are separated legally and institutionally, so that they can be backstopped separately as well.

Lesson Four: Dealer of last resort versus Lender of Last Resort. In the crisis, the profit-driven private market-making system froze up, both money dealing and risk dealing. In response, central banks expanded their traditional lender of last resort operations to backstop both term funding markets and capital asset markets, in effect becoming dealers of last resort. Subsequently, in their efforts to keep the system going during a period of necessary institutional change, they went even further, becoming in effect dealers of first resort, setting market prices rather than just backstopping them (QE, ZIRP, NIRP). From this point of view, the importance of “liftoff” is not so much shifting overnight rates away from zero but rather beginning the shift from dealer of first resort toward dealer of last resort. Regulators and bankers alike are groping toward a system where central banks serve as backstops to market-making that is done by others.

Lesson Five: Global money versus Local money. The eurodollar market is the global funding market, and the dollar is the global pricing benchmark. The Fed’s new Overnight Bank Funding Rate tracks eurodollar transactions as well as local Fed Funds, thus neatly bridging (by averaging) between the new and the old. Pozsar is right to emphasize the increasing irrelevance of the Fed Funds market, but I do not think the Fed will soon switch over to targeting OBFR, even though clearly global markets are already switching (dot #1, dot #2). So far as I can see, the hard “floor” on the local system is the Fed’s RRP facility, and everything else trades higher. Globally, the central bank liquidity swap lines create bounds on how far global dollar rates can deviate from local dollar rates, so providing elasticity at a price. Regulators and bankers alike are groping toward a system where global liquidity pricing disciplines net flows. The Fed is the ultimate global supplier of dollar liquidity, but that supply will not be costless, not in crisis times and also not in normal times.

Share your perspective