Before we get started, you can read more about the stuff I'm going to talk about here on the e$ home page <http://www.shipwright.com/>, in an article I wrote on digital bearer bonds in the Idees Fortes section of the last August's Wired, and in a debate I had on Wired's BrainTennis site <http://www.braintennis.com/> with Kawika Daguio of the American Banking Association. (Remembering, of course, due to the structure of the debate, I didn't get the last word :-))
Finally, practically nobody agrees with me on all of this stuff. In fact, there are plenty of people who stand up and shout at me about it when I present it. :-). But, as my friend Rodney Thayer says, "you're only as good as the people who yell at you."
There are two kinds of digital transaction settlement, just like there are in meatspace. The first kind is book-entry settlement, where offsetting debits and credits are exchanged between the buyer and the seller, using a clearing house or some other trusted financial intermediary. Checks, credit cards, and virtually all capital market trades happen this way. Then, there are transactions which settle with bearer certificates, issued by some body who wants to borrow or loan assets, and usually through a trustee, who "holds the stakes" on behalf of the certificate holders. Old-fashioned bearer bonds and bank notes, and modern dollar bills, postage stamps, and coins are all examples of this.
Since the advent of telegraphy, and then telephony, book entry transactions started to proliferate because you could transact business at a distance, and you didn't have to haul bearer certificates around for physical settlement. Mainframe computing automated book-entry transactions for even retail purposes, so that we got personal checks, mutual funds, credit cards, then, in the early 80's, money market accounts, and now, debit cards.
However, microprocessors changed all this. In 1986, now-Forbes-columnist Peter Huber wrote a report for Judge Harold Greene, the judge who broke up the bell system, on the effects of that breakup. Huber called the report "The Geodesic Network", because Huber noted that Moore's law, that is, the collapse of microprocessor prices by half every 18 months, was collapsing the telephone networks as well: from a hierarchical architecture where lines were relatively cheaper than switching nodes, to a geodesic structure, straight out of Bucky Fuller, where switches were becoming exponentially cheaper relative to lines. By the way, in the report, Huber said that the Regional Bell Operating companies should be allowed to compete in any market they wanted to, including information services, as long as they opened up the lowest level of the switching hierarchy, the local central office, to competition. 10 years later, they're finally getting around to it. :-).
Anyway, since our organizational structures map almost identically to our communications structures, we are evolving a geodesic economy, with geodesic capital markets to match.
Which gets us to the mother of all geodesic networks, the internet, a ubiquitous public network which enables ubiquitous public markets for everything you can move on a wire.
Old-style hierarchically organized capital markets relied on what would be considered insecure book-entries, but those book entries were moved around on very secure networks (NASDAQ Level 3, NIDS, SWIFT, for instance) which had stringent access control and strong biometric identity (everyone's fingerprinted, for instance). That way, if anyone made the wrong book-entry, you could send them to jail. Jail's probably the most important component. That's because book-entry settlement won't work without the threat of physical force, (preferrably from those lovable force-monopolies we call nation-states :-)), to prevent people from backing down on trades they don't like, a problem in financial settlement called non-repudiation.
Back in the days of bearer certificates, I handed you cash, which you could tell by inspection was real, and you handed me an ornate intaglio printed certificate, which I could tell by inspection was real. To resort to a double negative, non-repudiation is not a problem. :-). Fraud is a problem, but, in a bearer certificate market, you can destroy someone's reputation, literally killing them economically, if they commit fraud. Reputational sanction is very powerful stuff, if you have direct contact with that person on a daily basis, like people did in turn-of-the-century Wall Street. J. Pierpont Morgan, when asked by a congressional committee what the most important quality someone in finance should have, said, "Character. I wouldn't buy anything from a man with no character, even if he offered me all the bonds in Christendom."
Modern book-entry settlement solves the non-repudiation and fraud problems by, again, sending biometrically identified market actors, including you and me, to jail if we lie. Remember, you're sending insecure transactions, debits and credits, down very strong financial networks. On the internet, that's a problem, because the internet is, by definition, an insecure network, or it wouldn't be public. To quote Doug Barnes, who's the head of marketing for C2NET, a financial cryptography developer in Berkeley, when he spoke at the First Conference on Financial Cryptography, which held this February in Anguilla: "It's not a good idea to have 'and then you go to jail' as a terminating step in an internet transaction protocol."
So, how do you solve this problem of sending secure transactions down insecure public networks?
Well, you can create temporary proprietary private networks -- right down to biometric identity credentials -- on the internet, using strong cryptography. SSL, Netscape's Secure Sockets Layer protocol, is the way it's done at the moment, with an encrypted link between your browser and the server taking your credit card number. That's what that little key in the left corner of your Netscape window means. When it's not broken, you have a secure session. The putative next generation of this kind of technology, SET, the Secure Electronic Transaction protocol, developed for VISA and MasterCard, takes the secure channel one step further, and better integrates both biometric identity and non-repudiation features. Right now, the credit card consortia treat SSL transactions on the net the same way more risky mail-order and telephone order transactions are treated, with higher transaction costs. However, transactions done using SET will get the same transaction processing charges that face-to-face transactions get. That alone will probably make SET the dominant credit card handling protocol when it finally emerges from development.
You can also create a temporary proprietary "network" for checks, as well. The Financial Services Technology Consortium will be testing a prototype of its FSTC Electronic Check system this summer.
However, even with secure links to pass book-entries back and forth on, you still have the problem of the world-wide ubiquity of the public internet versus the requirement to send people to jail if they make the wrong book entry. In addition, there is no privacy once you're inside the system. Anyone with a wide enough subpoena, or, in the case of the Financial Crimes Enforcement Network (or FinCEN), an "administrative request", or worse, any clerk with the right access privileges, can see *everything* you buy. That's bad news for any of us who value our liberty.
The solution to all of this is, of course, is bearer certificates. And, fortunately, we now have digital bearer certificates.
In the middle 1980's, a cryptographer named David Chaum invented a cryptographic financial protocol, called blind digital signatures, which enabled, finally, the creation of digital bearer certificates. These certificates could be passed around on the internet just like those intaglio certificates of yore. I can hand you a digital bearer certificate, which you can inspect, on-line in real-time, to see if it's genuine. You can take that certificate (which, in the process of validating it, is now actually a reissued certificate which only you can spend) and hand that to someone else in exchange for something valuable, who can hand that to someone else, and so forth, until the final person in the chain goes back to the issuer and redeems it for, say, a dollar, in some other form, maybe in their bank account, or even physical cash.
You can use this cryptographic protocol to create bearer forms of practically anything financial. Digital bearer bonds, stocks, futures, options, foreign exchange settlement, macroproject finance bonds, even personal bearer notes instead of credit card charges. Some financial cryptographers joke about gold-denominated burmese opium futures, which gives a hint about just how powerful these cryptographic blobs of ones and zeroes could be someday as abstractors of value.
And, of course, you can create digital cash, which was the original idea for the use of the blind signature protocol. David Chaum founded Digicash, BV, in the Netherlands, to get out from under restrictive export laws, because cryptography is still considered a munition in the US. A few years ago, Digicash released a product called ecash, which is currently being issued by Mark Twain Bank (now Mercantile Bank) in St. Louis, Deutchebank, and banks in Sweden and Australia. Nomura in Japan just signed on to develop an ecash system of their own.
Unfortunately, the problem with the current market model for digital cash isn't only one of not enough merchants who accept digital cash, but one of market structure itself. Digital bearer certificate markets in meatspace usually have several players. There is the trustee who holds the money on behalf of the share or bondholders. There is the underwriter who issues the certificates, sells them into the primary market, and helps keep secondary markets for the certificates liquid. Finally, there are the people who are using digital cash to buy and sell things. In the case of digital bearer certificate technology, there is the inventor of the protocol, and there is the developer of the software which uses the protocol, as well.
Because there was no initial interest in blind signatures from the financial community when he invented them, David Chaum was forced to think of himself as everybody but the actual users of digital cash. Remember, the internet, such as it was, was government controlled, and Chaum had to market this technology in other forms, like smartcards. In fact, it can be said that digital cash really didn't *have* a very large potential market until it had a public internet to be used on.
Over time, Digicash figured out that it wasn't a trustee, or even an underwriter, as those were banking functions, and eventually sold their software to banks so that they could market their own versions of digicash under the ecash brand name. At that point, Digicash started booking it's first really serious revenue.
Unfortunately, the market isn't segmented enough yet, and, as a result, the market is stalled. Right now, trustee and the underwriter are the same entity, which permits someone on the board of a bank to object to who uses digital cash, which would be a preposterous assertion to uphold in a physical bearer certificate market, say for bank-issued notes. Ideally, what you want is a many-to-many relationship between trustees, who hold the money in a reserve account on behalf of the users of the digital cash according to a bond agreement, just like trustees do with bonds now, and underwriters, who would be responsible for marketing, issuing, validating, and redeeming those certificates on the net. When the money comes on or off the net, it goes through the trustee, who uses the book-entry transaction system of ATM settlement, or SWIFT, or Fed wires to get it where it needs to go. But while the money's on the net, it's the underwriter's responsibility. Biometric identity happens with the trustee where it belongs, and digital signature validation and reputation happens with the underwriter, on the net, where it belongs. Everybody's happy.
Unfortunately, spinning the financial functions back out into the financial system wasn't enough for Digicash, because it appears it isn't even a software developer either. The company failed to deliver software that the market could use without extensive customization for each bank, and, even when it stepped back a level of abstraction and started to develop software libraries that other developers could use, Digicash couldn't even do that very well, either.
Digicash just got another round of financing, and, in the shuffle, David Chaum had to sign over his rights to the blind signature patent to Digicash. One would hope that the new management at Digicash, and Nicholas Negroponte, it's new chairman, will understand what Digicash is: the world's best financial cryptography firm. With that in mind, Digicash should licence and validate the implementation of its patents and leave it at that. Digicash's business model shouldn't be J.P. Morgan, or Microsoft. It should be Thomas Dolby, Inc., who only licenses Dolby stereo equipment but never actually builds it. In the new economy, "wetware" is a better market than software, anyway.
There are several other kinds of digital bearer certificate protocols out there as well. Millicent, developed by Digital, is one specifically designed for microtransactions, for instance. However, the most interesting microtransaction protocol I've seen is one developed by cryptographers Ron Rivest and Adi Shamir, the R and S in RSA. They have developed something called MicroMint, which requires enormous computation resources to create the first microcoin, but practically none at all to create subsequent coins in the same issue. So, economics dictate that you issue as many coins as possible, which makes for *very* small denominations, maybe in the micro- or even picocent range. Possible applications for MicroMint, or, more probably its successor protocols, would be things like internet postage, or even micromoney-on-the-routerhead cash settled packet switching. Routers would be in the interesting position of auctioning switching to the highest bidder, and, eventually, saving enough out of operations to buy copies of themselves to handle switching loads.
This "micromoney as processor food" idea is at the very heart of the geodesic economy, I think. That's because you can have cash-settled realtime auction markets for anything digitable. And, since every kind of information from entertainment to genes to teleoperated surgery is digitable, very interesting things are going to result.
The problem of the distribution of intellectual property can even be solved by these kinds of cash settled, recursive auction markets. I create some new software, or a movie, or a book, or a song, or a gene, and I auction the first copy off to the highest bidder, or I conduct a Vigery auction, or whatever. Then I sell the second copy the same way. Every one I sold to has the right to do the same thing I did, and sell their copies. This kind of institutionalized piracy gets you two things. First, new stuff, and the people who think it up, make the most money. Second, because the network is geodesic, a swarm of many small market actors can get more marginal return on distributing an intellectual product quicker than a single hierarchically organized entity, like a publisher or movie studio, can.
In finance, this process is called disintermediation, where number of intermediaries are collapsed between the buyer and seller of a given asset. Actually, the idea of disintermediation is kind of like zero gravity. There is never a point where financial transactions, or any transaction, has no financial intermediary. On the on a geodesic public internet, what happens is more like what happens in space. Like microgravity, which is actually what people call zero-g, Moore's law creates an environment of microintermediation, where smaller financial intermediaries can process smaller and smaller bits of finance, if you will. That personal bearer bond you issued to pay for yesterday's lunch might have been bought by a syndicate of microfinance bots who live in the network just to make money on the money they loan to people like you. They, in turn, sell those microbonds to people who don't want their money to sit around in cash and make no interest.
Remember, for an industrial financial network to make this happen, the financial actors have to be *much* bigger. So big, in fact, that, not only does a many-to-one relationship exist between the borrower and the lender, but, the bigger the lender, the more borrowers it has, the more economies of scale it has. However, like pricing, where something on the net cost less the more it's resold, because transportation and storage costs are practically nonexistant, Moore's law on the internet turns financial intermediation on its head, and creates *dis*economies of scale, yielding microintermediation in the financial markets. Someday, the number of intermediaries between you and the other side of a trade will be one, but the number of *trades* you do for the same amount of money will exponentiate in number.
I even claim that in the same fashion, someday, nation states will go the way of monarchy, and be largely ceremonial in nature. Physical force, the thing that nation states have held a monopoly on for so long, will even be bought and sold in a geodesic auction market for, you guessed it, geodesic warfare. A brave new world indeed. There are hints of this in a paper titled, "The Mesh and the Net", authored by someone at the Army War College(?), which can be found on the e$ web-site <http://www.shipwright.com>. Imagine if a land mine was smart enough to let anyone with the right digital signature through, but nobody else, and you'll get the idea.
A geodesic economy won't happen because nation-states can't handle extraterritorial non-repudiation, or even fraud. It won't happen because digital bearer protocols allow completely anonymous transactions, for things digitable and, eventually, for things not so digitable. This, of course, deprives nation states of their ability to control business, and, more important, their citizenry. Nor will those nation-states be able to afford to enforce many laws when all their book-entry taxes, like those on capital gains, or on sales, or on income, eventually dry up and blow away. The joke a couple years ago on a Harvard Law discussion list was "What happens when taxes become a tip?". The joke is not so funny anymore to those people in government who now understand the punchline.
However, for the rest of us, all those things are effects, happy accidents, of something much more important. It is claimed, mostly by me, so far :-), that digital bearer settlement will result in hugely lowered transaction costs, perhaps as much as three or four orders of magnitude -- 3 or four decimal places to the left -- over book-entry settlement. Think about how your great grandfather, if he was in business for himself, conducted his business. That is, usually in cash. If, at the end of the day, he spent less cash than he brought in, he had a net profit. That was business in a nutshell. He didn't need a lawyer or an accountant to keep him out of jail on April 15. In a bearer certificate economy, you count up the bearer certificates you have on hand, equity, debt, cash, whatever, and that's what you have. Accounting is for management purposes only, to analyse and to forecast, but not for transactions. You don't have to keep seven years of accounting records for the tax man, because book-entry taxes exist as a *result* of book-entry settlement (which requires physical force to prevent non-repudiation) and not the other way around. If you don't *care* who you did business with, you don't need the local nation-state to hunt them down and put them in jail.
You don't care who you do business with, because you don't need biometric identity to conduct business on the net. You can "kill" the reputation of a digital signature which commits fraud against you, just by announcing the fact, and proving the fraud. All of this can be done without ever physically harming the person who controls that signature, without ever knowing, or caring, who they are. Since it takes a very long time for a new signature to acquire a trusted reputation, people will think twice about destroying the reputation of the signatures they control. As, J.P. Morgan said, it's all about character. Not about force.
The simple truth is, if book entry settlement goes away, so does, for lack of a better term, the book-entry state. And, frankly, all the things the nation state sells, including force, will eventually be sold, better, and cheaper, from much smaller private actors.
And now a little about the transnationality and potential macroeconomics of digital cash...
In a presentation to the Digital Commerce Society of Boston last year, Tatsuo Tanaka, of the Center for Global Communications at the International University of Japan, while a visiting researcher Columbia's Center on Japanese Economy & Business, delivered a paper on the transnationality of digital cash. He talked about how digital cash underwriters, after people started keeping money on the net by purchasing financial instruments, would be in a position to reduce the exchange fees because they would earn it back in interest on the reserve accounts the underwriter held against redemption. Over time, even this interest would be so much that competition would force underwriters to hold reserves against their digital cash outstanding smaller than the value of the cash on the net outstanding. This is, macroeconomically, increasing the money supply.
Actually, since I can issue yen onto the net from say, Chile, and back them up with reserves in dollars, I'm creating more yen to begin with, and there isn't much the Bank of Japan, or any other central bank (yet another hierarchical industrial construct ;-)), can do about it.
Central banks do a terrible job controlling their own foriegn exchange rates as it is, which is why Chile's had so much success with their currency board. It can be said that by paying so much attention to controlling the money supply for inflation, that the Fed itself is acting in the same mechanical fashion that Chile's currency board is, only with enormously huger overhead. :-).
So, like the nineteenth century American banks, and the "free" banks of Scotland, a digital cash underwriter on the net has the ability to literally create money, within the bounds of the market. Within the bounds of the market; which is exactly how the Banks of England, or Japan, or the Fed does. Reality, particularly economic reality, isn't optional. And, frankly, when there are enough digital bearer assets of other kinds on the net, the trustee banks holding the reserve accounts for those digital cash underwriters won't be banks of deposit anymore, which makes the world even that much *more* interesting.
David Friedman, philosopher, law professor, and anarchocapitalist son of Milton Friedman, lives on the net and thinks about this kind of stuff a lot. He said recently that he has now brought dad over to the idea that a bunch of electronically linked (dare I say geodesic?) free banks can behave much more economically efficient than any oligopoly of central banks ever can.
Which brings us back to Tatsuk Tanaka, who says that, eventually, all those fractionally reserved issuers of digital cash will suffer some kind of calamity, forcing a run on the entire system, and in the rubble, some kind of central bank of cyberspace will result, just like it did in all those turn of the century banking panics which, though J. Pierpont Morgan and his associates managed to handle them just fine, scared the Feds enough to, um, create the Fed, as it were, thus creating the financial ziggurat we all know and love today.
Frankly, I'd put my bets on a geodesic swarm of instantly communicating microfinance bots, than monthly Fed board money supply meetings, wouldn't you?
[ Back to the previous Rant ] | [ e$ home page ] | [ On to the next Rant ]