The Age of the Shadow Bank Run

The introduction to my column is this:

I RECENTLY asked a group of colleagues — and myself — to identify the single most important development to emerge from America’s financial crisis. Most of us had a common answer: The age of the bank run has returned.

I would like to see more discussion of how the permanently high demand for T-Bills as collateral will affect the U.S. economy:

Another feature of this new order is that more and more financial transactions will be collateralized with the safest securities possible: United States Treasuries. Demand for them will remain high, and low borrowing costs will ease our fiscal problems. Still, the resulting low rates of return serve as a tax on safe savings, encourage a risky quest for yield and redistribute resources to government borrowing and spending. It isn’t healthy for the private sector when investors are so obsessed with holding wealth in the form of safe governmental guarantees.

The bottom line is this:

The core problem is that the growth of short-term credit has been outracing our ability to protect it, and since 2008 most investors have realized that these shadow-banking transactions are not risk-free.

I didn’t have space to discuss whether this was a corporate governance issue or a moral hazard issue.  Under one view, managers/CEOs could purchase capital insurance to plug the runs, they just don’t have the incentive to do so.  The downside simply isn’t that bad for them.  Under another view, the market for “runs insurance” creates too much moral hazard to be feasible, or to some extent the market exists (CDS, etc.) but it just pushes the problem back another level and may even make matters worse by creating another level of credit.  A third view is that the collateral behind these short-term loans is somewhat of a farce, since it (sometimes) has value problems precisely in those world states when it needs to be called in.  It is probably a bit of each.

The conclusion is this:

In short, no promising financial path is before us. It’s good that the American economy seems to be recovering, and this may shove some problems into the future. But banking and finance remain a mess at their core. Welcome to the 21st century.


"I would like to see more discussion of how the permanently high demand for T-Bills as collateral will affect the U.S. economy:"

Demand isn't that high, its mostly the fed buying the bonds. This isn't real demand.

I don't work with this data regularly, so perhaps I am misreading But it looks to me like the banking system is now holding primarily government debt - far exceeding private lending. Personally, I see this as a major problem. If only the government can readily obtain credit, then the only source for private investment is retained earnings. The prospects for future growth are for the public sector and very large corporations. I'd say that is a fairly skewed pattern and worrisome.

Yes, you are misreading the tables. Read the second table. Private debt held is still much larger than government debt held.


You could say that previously,

Everyone was overleveraged.

Maybe it is something fundamental and things like central banking, the FDIC, etc. are just correlations. Finance is probably more important than we though it was, but less important than we think it is.

The piece was somewhat whiney and weak on evidentiary standards. Besides, what's the point predicting doom unless you offer some constructive sugesstions?

I am not competent enough to judge Tyler's analysis, though it strikes me as more than plausible, but I see nothing wrong with predicting doom without giving constructive solutions. Sometimes one can acurately see doom without knowing what specifically to to do to avert disaster. A doctrine that did not allow one to warn without supplying a countermeasure would be insane in say a fire or other actual physical disaster.

And since the problem seems to be too much reliance on government securities as collateral, I would suspect that would seem to hint at a possible direction in which to begin searching for solutions.

What kind of head in the sand approach is this? Dont raise a problem unless you have a pre-built solution for it?

"to some extent the market exists (CDS) but it just pushes the problem back another level and may even make matters worse by creating another level of credit"

exactly. We saw this in Greece, when huge haircuts were declared not to be defaults, and we can (could?) see it here if enough CDSs are triggered that the counterparties are rendered insolvent.

Which reminds me of something that has always bothered me -- what good is the FDIC (essentially CDS bought by the bank for the savings account holder with the federal government as the counterparty) if many banks fail at once?

Except that on the Greek default CDSs were triggered:

For non-finance-insiders who would like a pretty comprehensive readable description of the non-depository (so-called "shadow") intermediary system, the Federal Reserve paper Tyler links to in paragraph 4 of his column is the best I've seen.

The Times figured I'd spent enough time with his article already and wanted me to log in...can you provide that link?

You write "It now seems that the 21st century will resemble the 19th and early 20th centuries, with periodic panics and runs on financial institutions,"

Why not go back to doing what we did in the mid to late 20th century? Is this the argument of Caballero, to extend government protection as the did in the mid to late 20th century?

"That would check the problem, but at what cost? In a larger financial crisis based on insolvency, our government would face intolerable financial burdens, as happened in Ireland when its government guaranteed bank debts." Is this what was said about the FDIC back in the mid to late 20th century? No because the banks paid for the FDIC, not the government.

This would seem to be common sense unless one was instinctively anti-regulation and anti-government.

Or the problem may be that this is a feature of a zero or low inflation economy (come to think of it, weren't financial panics quite common in the lanscape back on the 'good old days' of money backed by precious metals?) Perhaps some updating of Milton Friedman is in order here. Inflation is everywhere and always due to money printing, financial panics are likewise everywhere and always due to lack of money printing. Perhaps we should seek a 'normal' level of inflation of 2-4% rather than seeking 0%-2% as the measure of success.

"Still, the resulting low rates of return serve as a tax on safe savings"

Really? How is this a "tax", and not absolutely normal supply-and-demand?

Seems like bank runs are inherent to fractional reserve banking. There would be no runs if banks weren't creating money out of nothing.

I don't know if any alternatives are better, but it seems like an important distinction.

If you didn't have fractional reserve banking you'd still have liquidity runs IMO. Key example, Greece and Italy. Yes yes Europe has fractional banking but from Greece's POV it doesn't. It's stuck with whatever amount of Euros the ECB thinks is proper to have and the ECB bases their decisions on the economy of Germany and France, not little Greece or even Italy. Did that stop a liquidity crises? No Both countries were not threatened so much from their current level of gov't spending but from the willingness of the market to 'turn over' their old debts as they matured into new ones.

Imagine if your bank suddenly told you after five years of you paying your mortgage that it was going to 'call in' and not 'renew' your motgage.....instead of simply continuing to make your mortgage payment you'd have to pay off the entire balance at the end of the month or find some other bank willing to create a new mortgage for you. That would be a liquidity crises and there's no particular reason to think it couldn' t happen in a world of zero fractional reserve banking.

I'd say that's a little backwards. Bank runs happen because banks are allowed to make credit, but not to print currency. If banks were allowed to actually print money, there would be no bank runs. This seems to me to be an obvious solution that we universally dismiss without properly considering.

You realize that there were in fact bank runs when banks were allowed to print their own money?

Only works because you eventually reach the point where no one would want the damned stuff.

Tyler, why don't you talk about this?

Report of the Committee on Honors and Awards (Bernheim). Bernheim explained that
nominations for the Clark Medal were solicited from economics department heads of major
research universities, from the Executive Committee, and from former Clark Medal winners.
The Honors and Awards Committee (Bernheim (chair), Banerjee, Card, ..... Electoral College,
VOTED to award the 2010 John Bates Clark Medal to Esther Duflo.

This takes me by surprise. They have been colleagues since her arrival at MIT – he was one of her PhD supervisors before she became a professor. In 2003 they founded (and still co-head) the Abdul Latif Jameel Poverty Action Lab (J-Pal), the MIT centre where anti-poverty initiatives are studied. They have lived together for 18 months.

Have you read the stuff of Morgan Ricks? He's a young law professor at Harvard Law and he's written a ton about this.

His bio is here with links to papers:

This seems to be one of those moments where the ground is shifting a bit for economics and Tyler, thanks for getting to such a base level before the moment passes. Having said that, while my response may seem a bit brutal, in no way is it meant to be snarky or some badass attitude (today at least) but out of genuine concern:

To the private sector - wake up and smell the coffee: what does it mean that the public wants to hold wealth in the form of safe governmental guarantees? Does it have to do with the fact that half of the populace cannot afford your most important product offerings? And to government, before you feel like the victor in this moment: You still want to raise the rent on everything so as to pay for knowledge services with the wealth of manufacturing and production. You have held the private sector in the vise grip of constant upward expectations. No amount of raising the bar is going to make it possible to pay the bill you have created. Services are a different form of wealth and have to be understood as such. People in the private and public sectors are trying to make knowledge accessible. Give people the chance to use these mass quantities of knowledge wealth, not just store it as another passive consumer purchase unless they actually get a job. Because if it does not get used, day in and day out by all of us, it will eventually fade away.
In other words, allow the individual to actually be a source of knowledge wealth, so that economies can finally be stabilized.
I am not trying to be an irritant I just want the world to be a better place.

Why does everything Tyler ever writes make me want to go buy gold bullion and keep it in my closet?

Ah, that's not what he's saying.

He's saying you should buy Treasury bonds.

Here's something I think would be interesting to know: When you net out all the tail chasing within the shadow banking sector, how much non-financial sector credit is funded from non-financial sector savers through these intermediation chains? And how would you divide this credit into categories that roughly distinguish the combinations of intermediation (credit, maturity, liquidity) being provided to the savers?

Everyone plays the curve. Eliminate money market funds ability to hold anything other than treasuries unless there is a $ for $ liquidity facility from a fed member bank. Oh btw, the capital charge for the liquidity facility should be a direct credit substitute charge not zero.

It isn’t healthy for the private sector when investors are so obsessed with holding wealth in the form of safe governmental guarantees.

What does "safe" mean in such a context? It seems that it clearly doesn't mean what "safe" usually means.

Why do so many people still assume that US Treasuries are the world's safest investment, when the Fed has already purchased half a trillion dollars' worth of them with (effectively) printed money?

It seems obvious to me that the Fed's willingness to do this makes clear: as soon as the "other shoe" drops in the US mortgage collapse (meaning, all of America's biggest banks are made to admit on their balance sheets that most of their loans are non-performing and they are insolvent), the Fed is going to pay off depositors by the printing press, avoiding a run on the banks by causing a runaway inflation of the dollar.

If I had money to invest I'd be looking for an investment that won't be hurt by collapses of both the dollar and the Euro. I can only think of one, and I wouldn't be surprised if the US makes it illegal again.

If there's 'runaway inflation' then the loans won't be 'nonperforming'. Either people will pay off their mortgages with inflated dollars or foreclosed houses will sell for more giving banks less of a loss on their loans. Since the banks will be making money...or less losses....why would the Fed have to go wild printing money to 'pay off depositors'? Also when exactly is this 'other shoe' going to drop? The subprime mess ended in 2008. Are you saying banks are holding mortgages that haven't seen a payment for four years now but are 'counting' them as perfectly good loans? I don't buy that accounting standards are that bad! Or are you saying that people who've made mortgage payments for the last four years, thought the worst recession since the Great Depression, are suddenly going to stop paying their mortgage once the unemployment rate falls? That latter doesn't seem to follow very well, if its the former then banks seem to have treated their depositors just fine since 2008 despite holding all these bad loans.

+1 to Boonton. JDGalt is about 3 years late with his 'all the banks are really insolvent, all loans are worthless' routine.

"in short, no promising financial path is before us. It’s good that the American economy seems to be recovering, and this may shove some problems into the future. But banking and finance remain a mess at their core. Welcome to the 21st century."

Not good news for the AD/AS model?

The AD/AS model is baloney. It doesn't make proper predictions very often, but its used because its one of the few models that can easily be explained in a classroom. Its a terrible, horrible model, but one that will not go away, due to the need to teach macro to millions of business students every year.

What seems to be missing is an observation of effects of the "Return on Capital" and its sibling "Return on Invested Capital" in productive enterprises. It is noe secret that the rates of return have been in a steady downtrend for over 50 years (from something over 6.2% down to the current 1.5% [approx]). The surpluses of large enterprises are not being redeployed into productive assets, and to the extent represented by "cash" are vulnerable to erosion.

As noted many, many years back by Walter Lippman and others, the popular and political concepts have that the increasing "complexity" of our financial systems and large enterprises can only be "balanced" for "stability" by increasing the authority (and complexity) of government power and actions (similar to the responses of 1913, 1920 and the 1933-40 eras).

Therein lies the obstruction to the activities and innovations that could provide diverse responses to diverse situations that constitute the finanacial impasse cited.

As a corollary: We are probably reaching the point where the complexity of political intrusions into commerce and financial transactions creates a response of a larger, more complex, more integrated (entangled?) private sector financial systems that are in turn exceedingly difficult to "manage" (if indeed management in any true sense is possible at all). This generates great difficulties in obtaining reliable and sufficiently current information necessary to support the elements of "Trust" essential to credit as an operational function of the several economies.

I am an investor.

I am sitting on a relatively high portion of cash right now. One of the fundamental reasons for moving to cash was that I concluded that the financial statements of many American companies, and in particular the banks, are completely untrustworthy. if you want people like me to buy corporate bonds sending the bankers responsible for the crisis to prison and outlawing naked derivatives would be good places to start.

Fortunately, though, being untrustworthy is itself untrustworthy. Suppose we take for the sake of the argument that all financial statements are lies. If you previously felt they were all true then shifting to cash quickly makes thought the world was safe and predictable and found out it was the opposite. Once that shock wears off, though, it makes sense to move back into companies. Why?

Well even if you can't trust financial statements the fact is there are profitable companies out there. If you can't go by their statements you will have to go by other means such as observing their ability to spend vast sums of money without ending up in bankruptcy (they have to somehow be earning money too then to pay the bills!). Note that if you tossed all your money into stocks at the deepest part of the recession, maybe when Obama just took office, you'd have brought the Dow at something like 7000 and now are enjoying 13000 or so.

Boonton 1, Joe Smith 0

Joe, there's plenty of good companies to own that have nothing to do with banking and pay fat dividends, proving real earnings.

The reference is to "Benford's Law." Current reports of large financial institutiond seem to show significant deviation from the historic norms.

Though I can't track my references at this moment, A surprising deviation from the general statistic rule of the frequency of in the intial numerical sequences of entries in the financial statements of large financial enterprises has been occurring and increasing. For example (but not the actual case), whereas ordinarily 78% of the line entries in a financial statement might begin with the numberal [1], now only 60% do so. This shift in numeric frequencies is thought to be indicative of changes that are not clearly apparent.

The numeric frequency phenomenon has a name, which escapes me now, but fits with the Fibonacci studies.

This may be some indication of "chaotic" effects from blending the results of dis-similar classes of financial operations, rather than segregating them, or more clearly indentifying the modalities of aggregation.

Comments for this post are closed