There Is An Absolutely Solid Collateral Case For What’s Driving Curve Inversion(s) [Part 2]

Eurodollar University’s Making Sense; Episode 40; Part 1: The Inside Story of Lehman, Collateral, and Why Always Premature Celebrations

There are, as I started, so many “ins and outs” to the whole that its difficult to parse the details and distill them down into a basic synopsis. Still, if forced to do it, the Lehman Brothers saga might best be summed up as follows:

A couple hundred billion in “secured loans” secured by collateral JPM knows doesn’t have much of a functioning market. Does JPM, having done the Fed a huge favor with Bear, really want to be in that kind of a situation? Would it want to risk even the slightest chance to be near the edges of anything resembling that situation?

No. It would demand, and begin seizing, only the best collateral (collateral calls – and not just to Lehman) to make any estimated or even perceived potential shortfall. And that’s exactly what it did – at the same time everyone else was doing mostly the same thing.

The nearly useless Financial Crisis Inquiry Commission gathered a wealth of information and internal communications (its only use; the conclusions reached merely ended up mimicking the prevailing establishment blame anyway, subprime mortgages, despite all its findings proving beyond doubt that wasn’t the actual problem), and specifically mentioned how in the week before the “bank’s” failure Lehman had been pressed by all of its repo counterparties to post more collateral via JP Morgan as main triparty custodian.

On Monday, September 8, more than 20 New York Fed officials were notified of a meeting the next morning “to continue the discussion of near-term options for dealing with a failing nonbank.” They received a list documenting Lehman’s tri-party repo exposure at roughly $200 billion. Before its collapse, Bear Stearns’s exposure had been only $50 to $80 billion. The documentation further noted that 10 counterparties provided 80% of Lehman’s repo financing, and that intraday liquidity provided by Lehman’s clearing banks JP Morgan, primarily could become a problem. Indeed, JP Morgan, Citigroup, and Bank of America had all demanded more collateral from Lehman, with the threat they might “cut off Lehman if they don’t receive it.”

The utterly useless Federal Reserve believed a new crisis policy put in place after Bear would adequately answer for this particular snafu should the situation ever arise, something called the Primary Dealer Credit Facility (PDCF). When, in fact, the PDCF merely demonstrated for the nth time how little policymakers understood the system officials constantly told the public they would ably defend (to understand more about why the PDCF was ill-suited, thus ill-fated, the complete story behind it and Lehman’s triparty collapse is here).

OK, but what is triparty repo?

The majority of repurchase agreements, really collateralized short-term lending, are done between just two counterparties – bilateral and bespoke. That’s the repo we can’t really see because other than the two counterparties involved, we really won’t know anything about the transaction unless it happens to take place across quarter-end or some other reporting period date.

In this bilateral shadow repo, the two parties have to find each other – the one with spare “cash” and the other with some form of collateral the cash lender finds amenable to its risk-aversion (can it reliably liquidate the collateral should the cash borrower default, and in what likely price range?) Typically, they already have done business and know each other.

Triparty repo is when some custodian bank stands in between the cash lender and cash borrower, doing so to more efficiently match one with the other (for a small fee) and maintaining the books and upkeep on each’s behalf. Rather than customized one-to-one, cookie-cutter mass repo across the whole range of potential participants.

For most of the repo market’s history, that was JP Morgan (along with Bank of New York).

As triparty custodian, however, this third participant though in between counterparties who bear (pun intended) ultimate recourse does take on some risks. The Fed’s PDCF did little to mitigate them:

What really happens is quite different; was quite different. Again, even the hint, the smallest perception of being stuck with collateral that may not be reliably liquid, investors pull, they demand instead only collateral that is reliably liquid – and not just from the one questionable dealer – leaving the triparty custodians (where there even is one) to begin protecting themselves from any fallout of what is instead an interbank run on collateral.

As with its securities lending data I discussed in yesterday’s Part 1 for the collateral case, the Federal Reserve also through its New York branch keeps tabs on triparty repo. Unlike its securities lending, it took GFC1 for FRBNY to begin doing so (an all-time “oops”). Then also like the securities lending figures, the Fed has little to no idea what the data really means.

I went over these numbers last year, so if you need a refresh the links are below.