[Cryptography] Fun and games with international transaction settlement (was Re: IBM looking at adopting bitcoin technology for major currencies)

[Robert Hettinga]

The way interbank clearing and settlement evolved is by the exchange first of pieces of clay and paper promising money to each other, and then having accounts on each other’s books, the balances of which changed after authenticated messages were themselves exchanged. 

Then financial entities kept accounts at a third party, a clearinghouse, to make settlement between all parties in a market net out faster. Today those clearinghouses for currency usually are called "central banks". Hence the “central” part, oddly enough. For securities that happen at, oddly enough, clearinghouses. Sensible, those capital markets operations types are. When I was paying attention this stuff, the New York Stock Exchange cleared at The Depository Trust Company. Before London’s Big Bang, custodial banks cleared their LSE-traded securities against each other without a clearing house. Then they built an obscurely-named database mirror called “CREST”, and set up a company, sensibly called “CRESTCo”, to run it. IBUC, a company I founded, did stuff with CRESTCo for a while. We wanted to hook up a Chaumian blind-signature to CREST and issue certificates on the net backed up by securities held in a custodial account in London. They paid us to help figure out if it was possible. Yes it is. Or was, before 9/11. Interest, such as it is, is now disclosed.


To continue the example, then. I send you, a fellow trusted financial entity (for various values of trust, financial, and entity) a piece of paper (or clay, in the cuneiform days…) which you believe to be true, to hand some asset (grain, gold, cows…) to somebody else with the promise that I’d pay you back. You take the okay, clay, and give the person offering you those same assets, as promised. At some point, you and I settle up, or, more likely, you just pass it on in exchange for something else of equal value discovered in a price-clearing auction of some kind, agreeable to both parties, maybe that person will come back to me, or someone they’ve traded with will. Note that there is no interest in this transaction. Okay, there’s a spread, taken off the top, but the actual claim on the traded document is perpetual unless stated otherwise. It is, in financial terms, a zero-coupon (interest) perpetuity.

Oddly enough, that’s what fiat currency is. It makes financial professors laugh in recognition, but that’s what “money” is. If I hand it to you, and you hand it to somebody else, and, some body eventually hands it back to me, I have to pay up, or all of the *other* pieces of clay, or paper, or base metal, or bits, become worthless. See “trust” above. It isn’t the asset that matters. It’s the fact of it’s exchange that matters. Like the old joke about a can of beans, those beans are for *selling*, not for eating.

At *exactly* the same time people started writing financial messages and trading them around like they were real stuff, people had actually invented writing to keep track of that very stuff. Usually in lists, oddly enough. Registers, say. Even though “bank” means bench (“bancos”, where the Venetians traded currency), banks have more often than not been people who kept people’s stuff and issued them either paper, or coins, or clay, representing other people’s claim on that stuff — oh, grain, say — or just kept lists, which those people, or others with their permission, could claim later. 

This usually happened at a premium for traded tokens, or a discount, if they just moved money around on the books of a bank. That second bit is because banks, by modern definition, at least, are people who lend out more than they are technically responsible for paying back. Rothbard, and the usual Real Libertarians out there, thought this stuff, fractional reserves, was Evil, but it seems to work, so we keep it. Problem is that bankers have socialized their collection process, in exchange for, shall we say, a contingent claim on that bank’s revenue and assets by local force monopolists. That’s called “debasing the currency” in most quarters, particularly if said force monopolist is responsible for defining what that currency actually *is*.


So, now along comes financial cryptography. We have a class of financial cryptography protocols that mimic moving those tokens around. Chaumian blind signatures being the Big Bang of Financial Cryptography for just this reason; and Bitcoin, which amounts to registered unforgeable certificates with a distributed, and also unforgeable, certificate registry. 

Lots of people have messed with book-entry financial cryptography, by the way, from Eric Hughes’ “encrypted open books”, to Wayner’s translucent database stuff, but, in general, just shoving encrypted instructions over a public internetwork to execute, clear, and settle a given transaction between my financial entity, your financial entity, and some third party clearinghouse — SSL, in other words — seems to be good enough for book-entry transactions right now. The big bang for *that* stuff was probably the telegraph, and certainly the hollerith card. Probably not much more to wring out of book-entry settlement that won’t also happen elsewhere at the same time.

Right now, Nakamotoan (heh...) registered digital certificates are *threatening* to be three orders of magnitude cheaper than SSL-encrypted financial instructions to internet-fronted financial entities wired into proprietary networks on the back-end. They’re certainly cheaper than remittance networks, which is all that remains of legacy telegraph operator Western Union, for instance. I like three orders of magnitude (cypherpunk eyes roll everywhere) because it’s a big scary number, and nothing else. It’s a placeholder for real data to be used later to determine the boundary layer between the adoption of one financial settlement technology and another. Implicit in that claim is that book-entry settlement (my bank, your bank, and a clearinghouse swapping debits and credits) was cheaper than Walter Wriston’s Singapore At 3PM, where couriers, armed and otherwise, walked physical paper bonds and bearer shares between buildings on the same block to literal cages in each other’s offices all at once. I used to work in one of the last of those cages, BTW, in Chicago, when I was a, um, punk, myself. :-) I also lived in a hole. In the road. And had to fill it in the morning. But that’s a Yorkshireman story for another time.


So, can finance use the blockchain, that distributed certificate register of, um, registered certificates, for something else? The Etherium guys think so. One of them was up on Dave Birch’s podcast a few months ago talking about something that would make Hernando de Soto (no, not that one, the newer one…) get a big case of the weepy glees: an unimpeachable distributed registry of land titles. Which would be Really Fucking Cool. (An actual term of financial operations art, that, here’s hoping it passes the censor’s muster here…)

Certainly, modulo whatever problems they may or may not have in delivering code, the Etherium guy has the financial elephant by the proper set of protuberances. In my former professional opinion as a former clerk in a cage. Among other things.

Couple those unimpeachable registries of title to other things, like, oh, all physical assets either before or after manufacturing, or in transit, or in storage, or on a retail shelf, and you can turn right around and create an *index* of those assets. Strangely enough, you can then turn around and calculate a *price* for that index, which, oddly enough, is what a central bank purports to do when they claim they’re “controlling inflation”. Modulo socializing other “externalities” (Hayek’s "last refuge of the dirigistes”) like “employment”, whatever that is, or monetizing national debts into quantitatively-eased phlogiston, or ether, or whatever Special Sauce they call it this week. Okay, so prices aren’t actually calculated, they’re discovered, but accumulating those discovered prices into an index gets you… well, you get the idea.


Some Idiot :-) talked about this nonsense, among other things, in various venues, public and private, fifteen years ago or so, taking Gene Fama’s and Fisher Black’s name in vain in the process. Essentially you can trade that index for other assets *as if* it was a fiat currency, using — these days, time flies — a blockchain as a clearinghouse, and the index as the stated value. These indices would be competitive, being thought up by Various Bright People Like Yourselves Here, Only Financial Ones, and anyone could use an index to issue, heh, zero-coupon perpetuities, denominated in assets contained in that index. 

Mass hysteria. Cats and Dogs living together. Somebody else here, *cough*, talked about “Gold-Denominated Burmese Opium Futures” once, in a ballroom long ago and far away. From me. Now, anyway. Kinda like that, only weirder.

This *kind* of thing, I believe, is what IBM is on about here, I think. Mostly the subsidiary bits, not the “and then you create a currency without the state, maaan” bit. Because That Would Be Just Crazy, If Not Actually Morally Wrong, okay?

In surfing, I believe they call kind of euphoria a “stoke”.

Sounds vaguely familiar to everybody here, I’m sure. Particularly the part where you have to pry some enthusiast or other off your leg before they get *too* excited…

Cheers,
RAH