Repo's and Balance Sheet Liability in Banks during the 2007 Crisis

[Dr. Jayanthi Sankaran]

Hi David,

In one of your videos covering repo's and the haircuts during the 2007 crisis, you mention that the cash borrowed appears in the Bank Balance sheet as a liability. This caused a run on the shadow banking system.

The example is a bank borrowing say $100 (the buyer of the repo) against a collateral bond from the seller of the repo (the lender). During the first half of 2007, the haircut = $0 and the cash borrowed = $100.

However, in the second half of 2007, the collateral which also included CDO's, ABS's and Corporate bonds had haircuts of 20% such that the cash borrowed by the bank = $80. This appears on the liability side as $80 which caused the bank to sell assets which led to a price lowering which led to a cycle of selling assets causing a run on the shadow banking system.

What I don't understand is that repo's are overnight borrowings and so are they shown in the Bank Balance Sheet as a liability of $80? Or are these term repos?

Thanks!
Jayanthi

 

[brian.field]

Jayanthi,

Chapter 12 in Tuckman's text can help with this question. It is assigned in Part II.

 

[Dr. Jayanthi Sankaran]

Thanks Brian - will look it up!

Best
Jayanthi

 

[brian.field]

Did you see David's recent post?
http://ftalphaville.ft.com/2016/03/22/2157236/something-very-significant-is-happening-in-repo/

Something very significant is happening in repo
Izabella Kaminska

Author alerts

|Mar 22 17:38|Comment|Share

Thespike in US Treasury bondfails to deliver, which startedearlier this year, is something we’ve been watching closely.

It’s fair to say we’re now at a significant milestone and the story is beginning to go mainstream.

From the WSJon Tuesday:

Settlement failures in Treasury repurchase transactions in March hit their highest level since 2008, underscoring concerns on Wall Street that trading conditions are apt to deteriorate in even the most-liquid markets under the acute stress evident early this year.

Almost 13% of Treasury repos through primary dealers in the week ended March 9 included a failure by one party to deliver securities as promised, according to the latest data available from the Federal Reserve Bank of New York. That is up from 2.7% last year and the highest ratio since 2008, saidJoseph Abate, an analyst at Barclays PLC.

Over at ADMISI Paul Mylchreest has dubbed it a $450bn plumbing problem:



So what’s going on and how to fix the problem?

Again, the standard response from industry is that this may be down to the unintended consequence of increasing the cost of renting out banks’ balance sheets.

In that regard, here’s Darrell Duffie, Dean Witter Distinguished Professor of Finance at the Graduate School of Business at Stanford University,writing in Forbes earlier this month:

In the case of repurchase agreements, known as repos, the “rental fee” for balance sheet space has been sharply increased by regulation. On a typical repo trade, a bank lends cash to a counterparty who secures the loan with bonds, say treasuries.The treasuries received by the bank are then usually financed by the bank itself on another repo, typically at a lower financing rate.

The bank profits from the difference between the two repo rates.

Absent capital requirements,this repo intermediation trade is almost self-financing because the bank passes the cash from one counterparty to the other, and the treasuries in the opposite direction.If a counterparty fails, the position can be liquidated with very low risk to the bank because it is almost fully secured or over-secured by cash or safe treasuries.

Aside from any needed addition to regulatory capital, this trade causes almost no net change in the bank’s balance sheet, so the rental fee for space on the bank’s balance sheet is almost zero.(It is not literally zero because there is a small risk and there are also short time lags between inflows and outflows of cash and treasuries.)

Breaking that down a bit: banks used to see a point in holding large market-making inventories of bonds because financing these assets was as easy as convincing legacy creditors that a greater pool of assets would more effectively protect their claims. In practice, this was achieved by diluting the claims of new subordinated creditors or equity investors.

It was this ‘safety’ injection into the highest echelon of the creditor base, which rationalised the speculative nature of the asset purchase.

Now, as Duffie observes, for as long the mark-to-market profit (or paper profit) of these arrangements exceeded the associated wealth transfer to high-order creditors, the trade appeared viable to everyone. [Ed - this issobitcoin.]

Not so much, however, if and when the paper profits began to undershoot the size of the wealth transfer. Something which in turn increased the chance of a rush to the exit, which in turn destabilised the house of cards the scheme was based on.

There’s no better way to think of this than as a system which uses the dilution of its lowest-order and freshest creditors (a.k.a depositors)/equity holders for the purpose of funding its risk-taking activities.

Put even more simply:before the new regulatory regime was instituted, assets were financed through the creation of additional subordinated claims on the bank’s overall capital base on the expectation that these assets would outperform in the long run, boosting the capital base for everyone. If the assets didn’t perform, the lowest creditor class would get hit disproportionately instead.

When bankers refer to balance sheet “rental fees” what they’re really talking about as a consequence is the risk and expense of creating unfunded liabilities (a.k.a money).

In the old regulatory system the cost of this was zero because all it required was a discrete reshuffling of the creditor claimant hierarchy. Under the new regime, however, even the lowest order liabilities have to be treated on much more equal terms.

Hence the balance sheet rental fee.

Regulators have restricted(very intentionally) the amount of ad hoc capital-structure shuffling a bank can get away with for the purpose of creating additional subordinated claims on its capital base. And they’ve done this in the name of depositor protection. As a consequence, if a bank is to take on longer term risk in the form of asset purchases it must now be compensated by the amount it costs it to introduce new claimants into its capital structure on a more equal and fair weighting with legacy creditors.

How this fits in with repo, of course, is that repos have historically been “self-financed” by the banking system — “self-financing” in this a case is a euphemism for the actual capital base of the entire banking system.

In that sense, the repo stock kinda represents the monetary float — the bit of the collective banking balance sheet which can and is continuously double-counted for the sake of financial lubrication, much like a continuous game of musical chairs.

As long as the music plays and the players keep going round (and the expectation is that the music should never ever stop anyway), banking theory dictates it matters not whether there’s enough chairs for everybody because nobody needs the chairs right now anyway.

And even if the music were to stop, the expectation is that everything would square up eventually because there isn’t a single player in the game who hasn’t pre-agreed his position in the claimant queue for the remaining chairs in the system (even if he didn’t know it).

And besides, as Duffie implies, bankers like to think that apart from small temporal inconsistencies related to time-lags between inflows and outflows of cash and treasuries, everything always squares up perfectly anyway. (Which inadvertently means the time lags in the clearing system ARE the risk in the system. Though, for more on that seehereandhere.)

Which brings us back to the current spike in repo settlement failures.

If these fails represent anything at all it’s the crystallization of the time-lag risk in the system which, in turn, provides a formal snapshot of the value being double-counted in the underlying capital base of the banking system.

In the context of a regulatory regime which, in the interests of depositor claimants, was set up very specifically to penalise the double-dipping tendencies of banks this arguably equates to a worrying signal from the markets.

Namely, that the efficiency of the modern economic system has become entirely dependent on the banking sector’s ability to transfer risk to an unwitting public sector balance sheet and depositor class (i.e. those who can least afford it) as and when the system requires the risk to be absorbed (i.e. when legacy creditors’ paper profits are threatened).

If we, the depositors, get fussier about our rights and protections — becoming notably more resistant to taking risk overall — the economy as a whole may have no choice but to become less capital efficient. That in turn could reduce the system’s overallcapital utilisation rate

, making capital much more expensive for everybody.

In short, it could indicate that without unrestricted cheap access to the buffer stock of society, the financial system’s ability to absorb imbalances just isn’t there.

 

[Dr. Jayanthi Sankaran]

Hi Brian,

Thanks for sending David's post on repo's. Will look through it!

Best
Jayanthi

 

[Dr. Jayanthi Sankaran]

Thanks Brian - read the article - interesting and very timely!

Best
Jayanthi

 

[brian.field]

Somewhat surprising to me, there is no mention of LaR nor CFaR.

Brian