Plain vanilla bank runs are as old as fractional reserve banking itself, and usually happen just before or during an economic and financial collapse, when all trust (i.e. credit) in counterparties disappears and it is every man, woman and child, and what meager savings they may have, for themselves. However, when it comes to shadow bank runs, which take place when institutions are so mismatched in interest, credit and/or maturity exposure that something just snaps as it did in the hours after the Lehman collapse, that due to the sheer size of their funding exposure that they promptly grind the system to a halt even before conventional banks can open their doors to the general public, the conventional wisdom is that this is a novel development (and one which is largely misunderstood). It isn’t.
As the NY Fed’s blog (whose historical narratives are far more informative and accurate than its attempts to “explain away” the labor force participation collapse) recounts, the first tremor in the shadow banking system took place not in 2008 but some 250 years ago… during the Commercial Credit crisis of 1763, whose analog today is the all too shaky and largely unregulated core shadow banking system component: Tri-Party Repo.
From the NY Fed blog, by James Narron and David Skeie:
Crisis Chronicles: The Commercial Credit Crisis of 1763 and Today’s Tri-Party Repo Market
During the economic boom and credit expansion that followed the Seven Years’ War (1756-63), Berlin was the equivalent of an emerging market, Amsterdam’s merchant bankers were the primary sources of credit, and the Hamburg banking houses served as intermediaries between the two. But some Amsterdam merchant bankers were leveraged far beyond their capacity. When a speculative grain deal went bad, the banks discovered that there were limits to how much risk could be effectively hedged. In this issue of Crisis Chronicles, we review how “fire sales” drove systemic risk in funding markets some 250 years ago and explain why this could still happen in today’s tri-party repo market.
Early Credit Wrappers
One of the primary financial credit instruments of the 1760s was the bill of exchange—essentially a written order to pay a fixed sum of money at a future date. Early forms of bills of exchange date back to eighth-century China; the instrument was later adopted by Arab merchants to facilitate trade, and then spread throughout Europe. Bills of exchange were originally designed as short-term contracts but gradually became heavily used for long-term borrowing. They were typically rolled over and became de facto short-term loans to finance longer-term projects, creating a classic balance sheet maturity mismatch. At that time, bills of exchange could be re-sold, with each seller serving as a signatory to the bill and, by implication, insuring the buyer of the bill against default. This practice prevented the circulation of low-credit-quality bills among market participants and created a kind of “credit wrapper”—a guarantee for the specific loan—by making all signatories jointly liable for a particular bill. In addition, low acceptance fees—the fees paid to market participants for taking on the obligation to pay the bill of exchange—implied a perceived negligible risk. But the practice also resulted in binding market participants together through their balance sheets: one bank might have a receivable asset and a payable liability for the same bill of exchange, even when no goods were traded. By the end of the Seven Years’ War in 1763, high leverage and balance sheet interconnectedness left merchant bankers highly vulnerable to any slowdown in credit availability.
Tight Credit Markets Lead to Distressed Sales
Merchant bankers believed that their balance sheet growth and leverage were hedged through offsetting claims and liabilities. And while some of the more conservative Dutch bankers were cautious in growing their wartime business, others expanded quickly. One of the faster growing merchant banks belonged to the de Neufville brothers, who speculated in depreciating currencies and endorsed a large number of bills of exchange. Noting their success (if only in the short term), other merchant bankers followed suit. The crisis was triggered when the brothers entered into a speculative deal to buy grain from the Russian army as it left Poland. But with the war’s end, previously elevated grain prices collapsed by more than 75 percent, and the price decline began to depress other prices. As asset prices fell, it became increasingly difficult to get new loans to roll over existing debt. Tight credit markets led to distressed sales and further price declines. As credit markets dried up, merchant bankers began to suffer direct losses when their counterparties went bankrupt.
The crisis came to a head in Amsterdam in late July 1763 when the banking houses of Aron Joseph & Co and de Neufville failed, despite a collective action to save them. Their failure caused the de Neufville house’s creditors around Amsterdam to default. Two weeks later, Hamburg saw a wave of bank collapses, which in turn led to a new wave of failures in Amsterdam and pressure in Berlin. In all, there were more than 100 bank failures, mostly in Hamburg.
An Early Crisis-Driven Bailout
The commercial crisis in Berlin was severe, with the manufacturer, merchant, and banker Johann Ernst Gotzkowsky at the center. Gotzkowsky’s liabilities were almost all in bills of exchange, while almost all his assets were in fixed capital divided among his silk works and porcelain factory. Berlin was able to mitigate the effects of the crisis when Crown Prince Frederick imposed a payments standstill for several firms. To prevent contagion, the prince also organized some of the first financial-crisis-driven bailouts after he examined the books of Gotzkowsky’s diverse operations. Ultimately, about half of Gotzkowsky’s creditors accepted 50 cents on the dollar for outstanding debts.
Meanwhile, banks in Hamburg and the Exchange Bank of Amsterdam tried to extend securitized loans to deflect the crisis. But existing lending provisions restricted the ratio of bank money to gold and silver such that the banks had no real power to expand credit. These healthy banks were legally limited in their ability to support the credit-constrained banks. To preserve cash on hand, Hamburg and Amsterdam banks were slow to honor bills of exchange, eventually honoring them only after pressure from Berlin. The fact that Amsterdam and Hamburg banks re-opened within the year—and some even within weeks—provides evidence that the crisis was one of liquidity and not fundamental insolvency.
The crisis led to a period of falling industrial production and credit stagnation in northern Europe, with the recession being both deep and long-lasting in Prussia. These developments prompted a second wave of bankruptcies in 1766.
Distressed Fire Sales and the Tri-Party Repo Market
From this crisis we learn that it is difficult for firms to hedge losses when market risk and credit risk are highly correlated and aggregate risk remains. In this case, as asset prices fell during a time of distressed “fire sales,” asset prices became more correlated, further exacerbating downward price movement. When one firm moved to shore up its balance sheet by selling distressed assets, that put downward pressure on other, interconnected balance sheets. The liquidity risk was heightened further because most firms were highly leveraged. Those that had liquidity guarded it, creating a self-fulfilling flight to liquidity.
As we saw during the recent financial crisis, the tri-party repo market was overly reliant on massive extensions of intraday credit, driven by the timing between the daily unwind and renewal of repo transactions. Estimates suggest that by 2007, the repo market had grown to $10 trillion—the same order of magnitude as the total assets in the U.S. commercial banking sector—and intraday credit to any particular broker/dealer might approach $100 billion. And as in the commercial crisis of 1763, risk was underpriced with low repo “haircuts”—a haircut being a demand by a depositor for collateral valued higher than the value of the deposit.
Much of the work to address intraday credit risk in the repo market will be complete by year-end 2014, when intraday credit will have been reduced from 100 percent to about 10 percent. But as New York Fed President William C. Dudley noted in his recent introductory remarks at the conference “Fire Sales” as a Driver of Systemic Risk, “current reforms do not address the risk that a dealer’s loss of access to tri-party repo funding could precipitate destabilizing asset fire sales.” For example, in a time of market stress, when margin calls and mark-to-market losses constrain liquidity, firms are forced to deleverage. As recently pointed out by our New York Fed colleagues, deleveraging could impact other market participants and market sectors in current times, just as it did in 1763.
Crown Prince Frederick provided a short-term solution in 1763, but as we’ll see in upcoming posts, credit crises persisted. As we look toward a tri-party repo market structure that is more resilient to “destabilizing asset fire sales” and that prices risk more accurately, we ask, can industry provide the leadership needed to ensure that credit crises don’t persist? Or will regulators need to step in and play a firmer role to discipline dealers that borrow short-term from money market fund lenders and draw on the intraday credit provided by clearing banks? Tell us what you think.
* * *
Fast forward to today when we find that the total collateral value in the Tri-Party repo system as of December amounts to $1.6 trillion.
… or 10% of US GDP. What can possibly go wrong.