Why letting Lehman go did crush the financial markets | FT Alphaville

Why letting Lehman go did crush the financial markets

For some time now, the folks over at Clusterstock – notably John Carney – have led a challenge to a particularly virulent piece of received wisdom: that the failure of Lehman was necessarily an inflection point that took the severity of the financial crisis to a whole new level.

And with that the implication that the government’s decision to let Lehman fail was, in itself, a failure.

Until now, that kind of debate might have seemed a little academic – a question for historians. But day by day; bailout by bailout, its pertinence to current events and future policy is growing: politicians and regulators are going to find themselves increasingly under pressure to account for the growing number of expensive opportunities they are being occasioned with to Save The World.

Loath as we are to turn again to the “Japanese Scenario” for appropriate lessons, it’s worth bearing in mind that in Japan, it was ultimately the weight of public opinion, as much as it was economic or financial considerations, that came to shape the way the crisis played out. Distaste for spending taxpayers’ money grew extreme: bailouts became taboo. The way Japan’s authorities consequently pussy-footed their way around problems rather than tackling them head on drew the crisis out for nigh on a decade – dare we now even say, two.

So the stakes are high, and Lehman makes for a perfect case study.

The broadest and most challenging question, we suppose, is whether in the long run, the whole banking system was set for failure anyway. Or to rephrase it: from a counter factual point of view, would a world in which Lehman was propped up necessarily be a safer one?

As the FT’s own John Gapper has argued, it would not. The locus of panic would simply have shifted onto the next institution:

I’m not convinced that, even if Lehman had been rescued, that would have averted the problem since the weight of selling and panic would have moved on to the next financial institution and then the next. So the world would probably have ended up in the same position it is in today, with a broad financial sector bail-out.

Wedded to this is the assumption that bailing out Lehman would have had nothing to do with actually cauterising the root cause of the crisis: the US housing market.


The collapse of Lehman did create panic among the world’s financial institutions, and it did significantly increase “risk” in the system, not just redistribute it. And it seems likely that it did indeed make it more reasonable to expect that other institutions would fail – because it created generalised panic in the funding markets which every bank – irregardless of their pedigree or resilience – was dependent upon.

After Lehman collapsed, Morgan Stanley and Goldman Sachs very nearly did too.

While a Lehman bailout would not, as Gapper, Carney and others have noted, have solved the solvency crisis that faces banks currently, it would have averted the extremely violent – though short – liquidity crisis that the financial world experienced in September and October.

Whether that perfect September storm counts as a hastening of the crisis which in the long run may come to be seen as a good thing, the jury is still out on. We here at FT Alphaville though, think that the damage it wrought – damage which totally caught the US authorities by surprise – should not be underestimated.

First though, a little more on that post-Lehman liquidity crisis itself – and whether, indeed, per Clusterstock’s latest post, it was caused by Lehman at all.

Stanford University’s John Taylor has authored an “event study” that suggests that it was not the inability or unwillingness of regulators to save Lehman over the weekend of September 13-14 2008 that led credit markets to seize up around the globe. “The Financial Crisis and the Policy Responses: An Empirical Analysis of What Went Wrong” demonstrates that the credit markets actually did not actually go into cardiac arrest after Lehman declared bankruptcy. Rather, it was the dithering and incoherent government reaction that brought on the crisis.

In support of that, here, from Taylor’s paper, is this key graph:

link to Libor-OIS crisis graph

It shows the Libor-OIS spread – a key measure of perceived counterparty risk in the market. You can see just how egregious that widening was, in context, by looking at the spread over a longer period. And as Carney notes, it apparently also shows that it was government dithering after the collapse, rather than the collapse itself, that prompted the widening. Writes Taylor:

On Friday of that week the Treasury announced that it was going to propose a large rescue package, though the size and details weren’t there yet. Over the weekend the package was put together and on Tuesday September 23, Federal Reserve Board Chairman Ben Bernanke and Treasury Secretary Henry Paulson testified at the Senate Banking Committee about the TARP, saying that it would be $700 billion in size. They provided a 2-1/2 page draft of legislation with no mention of oversight and few restrictions on the use. They were questioned intensely in this testimony and the reaction was quite negative, judging by the large volume of critical mail received by many members of the United States Congress. As shown in Figure 13 it was following this testimony that one really begins to see the crises deepening, as measured by the relentless upward movement in Libor-OIS spread for the next three weeks. Things steadily deteriorated and the spread went through the roof to 3.5 per cent.

The problem here is that the Libor component of the Libor-OIS spread, is not really a wholly useful indicator of the state of the credit markets. It’s a reality-based fiction based on the aggregated opinions of individual banks as to the cost of unsecured interbank lending. What it is not is an actual demonstration of market movements. It does not necessarily actually reflect the rates banks are lending to each other at.

More to the point, Libor is calculated from an aggregate of individual banks’ guesses as to what rate at which other banks are likely to lend to them. Libor is a proxy metric.

In the wake of a collapse like Lehman, there is thus naturally a margin for significant statistical lag with Libor: no bank would wish to stand out by submitting the highest number, for it would show them to be the most at risk of failing next. At individual banks, those responsible for guestimating the daily Libor figure they will submit to the BBA are indeed very wary of what the previous day’s Libor figure revealed. Libor has path dependency.

While this odd Libor psychology doesn’t wholly explain the lag identified by Taylor and Carney, there is more compelling hard evidence.

If Libor is a proxy, here’s some compelling hard evidence. In a technical sense, interbank lending is typically done so that banks are able to meet their solvency requirements at the close of their books each day. At the Fed, banks lend money to each other by transferring it between their accounts in order that they might meet their close of trade reserve requirement.

What’s telling then, is what happened to banks’ reserves held at the Fed immediately after the Lehman collapse.

US monetary base

The weekly data behind the above graph (monetary base, which is banks’ Fed-held reserves plus coinage) shows that in the five days following the LEH demise, banks more than doubled their cash held in reserve at the Fed – cash way in excess of their reserve requirements.

In other words, clearly the banks anticipated – or were already experiencing – an interbank lending collapse straight after Lehman, even if it wasn’t immediately shown in the Libor figures they reported. (One explanation for the discrepancy is perhaps that because no interbank lending was actually taking place, Libor calculations became even more path dependent than normal- the calculations had nothing else of empirical worth to be based on).

To boot, there are hard statistics on another, arguably even more important, source of short-term wholesale financing for the banks that dried-up straight after Lehman: the commercial paper market.

The most immediate disaster for banks after Lehman’s collapse was the failure of Reserve Primary – a huge money market fund which broke the buck on September 15th after it suffered losses on unsecured commercial paper it had bought from LEH.

What happened in the commercial paper market – or “money market” as it is colloquially often known- really shows the true scale of the Lehman disaster: an electronic run on the banks. In graphical form:

Money market fund assets

The red line shows the collapse – in the two days following Lehman’s bankruptcy – of the market for commercial paper issued by banks. Within a week, $500bn of short-term funding had dried up. Then there is the asset-backed CP market, which fared equally badly. Credit lines banks had supplied to asset-backed conduits became less reliable and so too, therefore, did CP issued by those conduits, in turn making it more likely that such credit lines would be drawn down upon – a vicious cycle worse than that seen for ABCP structures when the crisis first hit them a year earlier.

It was thus in context that the decision to allow Lehman to collapse was a failure. A Lehman failure didn’t have to spell disaster- it could, perhaps should, have occurred alongside an announcement of a generalised guarantee on money market funds – as well as a broad commitment from the Fed to extend its liquidity facilities. That such announcements in reality, came a week later was no good.

In the context of what happened with Bear Stearns too, not bailing out Lehman was a mistake. The Bear decision introduced huge moral hazard, as John Carney at Clusterstock earlier noted:

… the bailout of Bear Stearns had in fact introduced massive moral hazard into the markets, allowing investment banking executives and their boards to believe that they wouldn’t be allowed to fail.

Lehman’s failure though realised that hazard. This Bank of America graph is particularly revealing:


In the world before Bear, CP investors had their wits about them. The CP they bought from the ailing bank decreased sharply as concerns about its health grew. In the world after Bear, the opposite was the case. Investors were more than happy to buy Lehman CP: lulled into a false sense of security by a sort of faintly implicit guarantee from the US government against too-big-to-fail banks.


Arguing counterfactuals is always problematic. In a world in which Lehman had survived, would a TARP ever have made it through Congress? After a Lehman bailout, would AIG have ended up a victim: a bailout too far? (If so, the consequences would have been far worse.)

There is – in spite of all the above – a lot to be said for the fact that a bailout of Lehman would have led to much more protracted, if less severe, financial malaise.