This is part three of a multi-part series about the deep and largely apocryphal history of collateralised lending.

Here are parts 1 and 2. For those who can’t be bothered to read parts 1 and 2, but still want to read part 3 — I mean, suit yourself — or who have read them but just forgot, here is a recap.

This Substack is reader-supported. To receive the best posts and support this struggling artiste, consider becoming a paid subscriber. Only half a pint a week and you get the satisfaction not just of feeling like a Renaissance patron: you get access to Premium JC and all kinds of neat stuff.

Recap

We started with the simple question — simple for securities lawyers and ISDA ninjas, that is: it might not occur to anyone else — “why do Americans prefer collateral pledges and Brits prefer title transfer collateral?”. In part 1 we covered the difference, in the JC’s argot, between “collateral” and “security”: “collateral” is something you hand over to your creditor. “Security” is an enforceable legal priority you grant over assets you do not hand over. These are quite different modes of performance assurance, especially when it comes to managing a creditor’s “cost of funding” indebtedness it is owed.

We are accustomed to thinking mainly about credit risk — lawyers are prone to catastrophising and never looking on the bright side — so in Part 1 we spoke a bit about a happier reason for taking collateral: funding optimisation. In other words, making more money.

In part 2 we will traced a “just so” history of financial collateral from no collateral or credit mitigation, to merchants knocking about on the high seas and off the Barbary coast — we spoke a bit about poor old Shylock in the Merchant of Venice — to customary liens and pawnbroking arrangements — this was all the excuse I needed to tell an amusing story about Jimi Hendrix, and I may well come back to it — and fetched up with an exposition on the development from the 19th century of modern financial markets — negotiable bonds, stocks, centralised markets with oddly gesticulating men in silly blazers and ladies scribbling things on blackboards — to electronic markets, clearing, custody, and the opportunity these modern collateral arrangements presented to lenders to better manage their funding costs. That is, make money.

We ended with an odd question: if borrowers were providing more moneys’ worth in collateral than the actual money they raised, could creditors use that collateral somehow to offset their funding costs of the very loan the borrower used to buy the collateral and did that mean, in a round about way, that borrowers were, in effect, lending to themselves?

In this part we will alight on the answer which, at the risk of spoiling it, we will discover to be Yes.

That is weird enough but we will see that, thanks to some odd wormhole or eddy in the flow of spacetime, this means a bunch of different financial instruments, that seem quite different from each other, seem to collapse down to a single wave form, and become the same.

We will compare collateral arrangements with the financial transactions they collateralise and see how —if viewed in purely economic terms — they are not really that different. Things get a bit strange loopy.

So, without out further ado, let’s play the third act of this odd movie.

The hidden financing in every swap

“What is dull is never done.”

—The Swappist Oath

But hang on, if a swap is a “self-funding” loan then, by ponying up collateral, isn’t the “borrower” lending to itself?

The answer is, in a sense, yes. But that is the very distinction between “borrowing” and “financing”.

Borrowing is the outright transfer of capital. The lender gives money, outright, and accepts the borrower’s unadorned credit risk for its return.

Financing is the temporary conversion of an existing asset into cash. You give the asset away in return for cash, against a promise to repay the cash against return of the asset. It is a “swap”, of sorts, even if not quite in the sense implied in an ISDA.

The financier’s risk is different: it has intraday exposure not to its financing counterparty, but to the financed asset. This usually has a greater value than the sum advanced: as much as 30% more. Only if the asset suddenly collapses in value — if it “gaps down,” in the vernacular — does the financier have any credit risk to its financing counterparty for repayment.

The financier, in the vernacular again, has “second-loss” risk to its counterparty.

So, to labour the point:

Lending is the outright, uncollateralised transfer or money from lender to borrower.

Financing is the temporary, and reversible, exchange of an asset for its cash value.

From one angle only — that of “an advance of money from one party to the other” — do they look the same. In the round, economically, they are quite different.

It is a fun exercise to bucket different financial products into loans and financings:

A swap as a financing

The mark-to market exposure under a ISDA Master Agreement, not counting its CSA, represents outright indebtedness. Not necessarily to the dealer: the indebtedness — the “Exposure” — can swing in and out of the money. In an odd way, a Swap portfolio resembles a revolving credit facility — an overdraft. The only difference — and it is structural and not economic — is that the “withdrawals” and “deposits” in a swap are beyond the control of debtor and creditor. They are generated by — they are “derivatives” of — market movements on transactions.

But swaps were not designed to be like overdrafts. They are meant to be “unfunded” instruments. They were designed, by the First Men at the dawn of the Age of Swaps, as paired, offsetting loans, the principal sums of which are meant cancel each other out. What is left is, literally, a “derivative” of the exchanged loans: the present value of remaining cashflows under one loan deducted from the present value of the remaining cashflows under the other. This is a neat expression of the market delta.

Since swap dealers are not meant to be borrowers nor lenders, and don’t take deposits to fund their operation they are not necessarily regulated or capitalised to take large debt exposures to their customers. It would be eye-wateringly expensive to do so even if they were.

So, dealers require the their swap exposures to be transformed from loans to financings that do not attract the same capital charges.

A CSA is designed to solve that problem. Economically it is the inverse of a portfolio of swaps. So, if a swap is an overdraft facility, so is a CSA: only its equal and opposite. Both are outright disposals of capital.

Of course, a CSA can only exist with an ISDA. A CSA existing independently is a nonsense. CSAs are sort of “ISDA-linked swaps”. Their “underlier” is the net Exposureof the very ISDA they are collateralising.

When we add a CSA to counteract an ISDA Exposure they cancel each other out: Exposure goes one way, variation margin goes the other, and — hey presto — there is no longer an outright disposal of capital. We have a financing.

Usually, a financing starts with a party putting up an asset to raise money. In a margined ISDA, that sequence is reversed — the dealer “lends” money, and the customer punts over an asset to collateralise it, and that makes it into a financing — but economically, they are the same, whichever comes first.

From this perspective, a swap looks like a margin loan from a dealer to a customer to buy a financial asset. Instead of giving money to the customer to buy an asset, the dealer uses it to buy the asset as a hedge, for its own account. The dealer now owns the asset, but pays away all the its economics to the customer.

It also looks a repo, which looks like a stock loan, which looks like PBr. Are all just different articulations of the same principle. Financing.

I feel like I have been labouring this point. Time to move on.

The financing parcel game

Triago: All the world’s a margin loan —Each man a mere financier:Who calls out wonkish marks andPledgès troth o’er something fancierWhate’er fetches upAnd passeth thy criteriaRehypothecate the lot, dear friend—In a flush of rehysteria.

—Otto Büchstein, Die Schweizer Heulsuse

Recall that what an outright lender cares about is making sure its cost of providing capital — the cash it has to stump up to make the loan — is less than the return it gets from lending it. Borrow cheap, lend expensive.

(Seems trite, I know, but does not seem to have occurred to the treasury department at Silicon Valley Bank.)

Generally, a bank funds its loan book by borrowing, from retail depositors, commercial paper, medium term notes, bonds raised in the debt markets — and repos — but that is to give away the punchline a bit so let’s park that.

The bank must pay interest on, and hold capital against, those “funding liabilities”. The difference between the amount of interest a bank can earn from lending money, and the amount it must pay to acquire that money, is its profit margin.

A bank that could find a way of lending money to X without first having to borrow it from Y would be in a rather special place. A place that would seem, rather, to transgress the laws of banking physics: how can you lend something you do not first have?

Like this. Step one: take financial collateral from your borrower thereby converting an outright loan into a financing. Step two: convert that financial collateral into money, and repay your lenders with the proceeds.

Behold: the beguiling magic of a margin loan.

By financing — that is, converting into money — the collateral it received from its borrower, a margin lender no longer needs to borrow the money it used to make the margin loan. The margin borrower is, effectively lending to itself through its collateral.

Selling the collateral outright would leave the lender with unhedged market risk to that collateral — it needs to return it the borrower later so if the price goes up after it sells it, it will book a loss — so the lender needs finances the collateral with repo or stock loan arrangements. As long as its repo counterparty is there, it is hedged against collateral price risk. Any margin arrangements on its repo will mirror the margin arrangements on its own margin loan. Economically, the margin lender steps out of the picture. Unless there is a disaster.

But hold on: how can you sell my asset —?

This is all well and good, but hold on: how are you meant to sell an asset that has only been pledged to you, and that therefore you don’t own? This is exactly the problem of the pawnbroker we saw in the last episode: unlike a right-handed guitar, strung upside down for a left-handed m****g genius, when the pledged asset is fungible financial collateral, the lender can sell, loan, or re-pledge it. This process is called rehypothecation — if you are American — or “reuse” if you are a Brit.

So we should talk a little about rehypothecation.

The first time you encounter it, it will seem utterly mad. This because, candidly, it is utterly mad. It is one reason why the British view of collateralisation, by title transfer, makes more sense than the American view that collateral should be transferred by pledge.

It is a piece of American conjury, designed to overcome the difficult fact when you provide collateral by pledge you give up possession, but not ownership. You grant a lien over an asset that remains yours.

This is rather like a pawnbroking arrangement: when Private James M. Hendrix of the 101st Airborne pitched up with his right-handed Stratocaster, strung upside down for left-handed m——g genius, his pawnbroker lent against the guitar and had good credit protection if Private Hendrix failed to return, but until he was obliged to, the pawnbroker had to sit on the Stratocaster. It could not sell it, pledge it, or raise money against it.

That buggers up the opportunity for funding optimisation. So some bright spark contrived a pledgee’s right to “rehypothecate” a pledged financial asset: this is the right to sell it, on condition that you are liable to buy it back whenever its owner wants it back. This is conceptually possible because of the unique nature of modern financial instruments: they are inherently fungible — individual securities of the same ISIN can’t vary in quality — and in the modern era, being mere electronic impulses in a digital menagerie they don’t really have individual corporeal identity at all — and lastly, the market is liquid, meaning securities can, in most cases, be readily acquired. The only variable is price, and that is an explicit risk the rehypothecator runs.

In the real world of real, perishable, scuffable, dentable, string-it-upside-down-for-a-left-handed-m——g-genius-able assets this would not be possible. Each asset is unique. Its nature and value are coloured by the unique path it takes through spacetime and the vicissitudes it encounters along the way.

None of that matters for financial instruments. They are abstract concepts. They can’t perish, be scuffed, dented or strung upside down for left-handed m——g genius. So the idea of rehypothecation makes a smidgen of practical sense. A scintilla. A wafer-thin-mint of sense. But no more. It works, as long as you don’t think too hard about what, in theory, it means.

A rehypothecated asset might as well have been title-transferred. To all intents and purposes, it has been. Rehypothecation converts a pledge into an outright title transfer, only with annoying formal, and quite moot, perfections and legal ninjery surrounding it.

British and American means of posting collateral are, to all practical intents, the same. You title transfer to the dealer, the dealer punts it out into the market to raise money against it. This it uses to repay its own treasury department, from whom it borrowed the money to buy the hedge in the first place. You don’t owe the asset. You have a claim against the dealer for an equivalent asset.

Is everything a financing then?

The implications may take a while to sink in. A margin lender is, basically mirroring its customer. It rehypothecates posted collateral into the market. More than likely, the lender’s market counterparty will do the same thing. There unfolds a chain of back-to-back financings. The collateral assets fly around the financial markets, like neurons in a brain. Rather the way cash flies around the economy. The greater the velocity, the more productive everything is.

Could this sort of thing happen indefinitely? Yes.

Does it? Yes.

Who knows where it all ends up? Search me.

Much of modern finance is an elaborate chain of, effectively, margin financing arrangements. This is a weird and confronting idea when you first encounter it.

But it explains a lot.

For there are pawnbrokers and mechanics, and then there are twenty-first century multinational financial institutions.

As markets evolved from enterprising men on bicycles discounting jewellers’ bills in lower Manhattan, and as money, credit and systems digitised, capital efficiency has become ever more imperative. As markets grew, crashed and recovered, so did regulations. Each crisis begat new rules; each new rule focused more intently on capital regulation.

As the computerised market sped up, so too did the regulatory response. There is a curious, ironic feedback loop, where technology enables greater complication, which enables more risk-taking, which triggers collapses, which triggers regulation, which triggers innovation in means of optimising capital efficiency, which enables more risk-taking — you can see where this is going. Basel I was 30 pages. Basel 3 runs to several hundred.

All of this momentum pushed financial institutions to the same place: capital is expensive, balance sheet is precious, and assets must therefore be put to work. There’s some irony here. Regulations designed to reduce systemic risk incentivised institutions to take it: to sweat every last drop of funding from their operations — which only increased the speed and rate of flow. No longer would yards and yards of customer assets be left to sit quietly in dusty custody accounts, any more than customer cash is left to sit in a safe.

Everything that could be reused to raise cash would be. Longer, less visible financing chains emerged. Participants grew ever more interconnected. There was greater complexity. Assets circulated with ever-greater velocity.

Has this reduced systemic risk? Or simply shifted it? Great question. It gets answered in real time.

See also

* Variation margin: deep history part 1

* Variation margin: deep history part 2

* A swap as a loan

Thanks for reading! This post is public — but took quite a long time to write! If you enjoyed it, tell your friends. If you didn’t, tell someone who you don’t like. Who knows? they might enjoy it, and if they don’t? So much the better.

This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit jollycontrarian.substack.com/subscribe