I was reading Jonathan Stray’s account of Jon Hicks’s possibly fictional 1989 account of the how banking and modern money came to be, and it occurred to me that the bond market might be an alternative to the borrow-short lend-long model of fractional-reserve demand-deposit banking.
That is, if you have some money that you want to keep liquid but also earn interest on, one possibility is to deposit it in a bank as demand deposits. The bank lends (most of) it out at higher interest, keeping a small (“fractional”) reserve on hand in case some depositors show up one day to close their accounts, and pays some (lower) interest to you and the other depositors for the use of your money. (Modern banks have been able to mostly eliminate depositor interest, as their oligopoly on payment systems, and in some case government regulations, obliges people to deposit their money in banks even if they don’t earn interest.) The risk you take is that there might be a bank run — if too many depositors show up demanding their demand deposits at once, the first ones will get paid in full and the last ones will get their share of whatever is left over from the bankruptcy.
Given the existence of a liquid trading market, bonds, such as corporate bonds, might be a reasonable alternative. Instead of depositing your money in a bank account, you buy some bonds. Instead of withdrawing it, you sell the bonds on the market. Bond prices go up and down a bit — usually measured in basis points — and occasionally companies will go bankrupt (or countries will go into sovereign default) and again you’re just a bankruptcy creditor, but you can diversify to minimize this risk. Lending to companies is overall mostly profitable, and the interest rates are higher than you get with a checking account.
This is why money-market accounts are popular, but there’s no fundamental reason that retail investment in the corporate-bond market needs to be intermediated by banks or brokers; that’s just an artifact of the limits of 20th-century information-processing technology. Intermediating retail access to corporate bonds in this way just adds risk — your bank or broker is far more likely to go bankrupt than the entire basket of companies you lend to, and additionally they’re in a position to cheat you in a variety of ways, including bucket-shop tactics (that is, fractional-reserve banking, but in a way that is illegal in the US), front-running, and paying bonuses to their executives just before bankruptcy.
How would this differ in practice from a bank account? Blue-chip corporate and sovereign bonds typically fluctuate in value only by basis points, so the chances of a significant loss of capital in this way is fairly small, and you can diversify the risk across many different debtors. The big difference is what happens when there’s a “bank run”: in the money market, those lenders who are most eager to liquidate their holdings are often the ones who take the losses, while in a demand-deposit bank run, the most eager lenders are the only ones who get paid. A temporary selloff in the bond market doesn’t directly affect the companies whose bonds are being sold, unless they are floating a new issue of bonds at that moment, which they usually aren’t. If the selloff is indeed temporary, rather than a result of the debtor’s impending insolvency, those lenders who held during the selloff will preserve their capital (the precise opposite of a bank run), and those who risked buying the bargain-priced bonds are rewarded.
In the modern commercial paper market, though, clearing of trades is not instantaneous; the US corporate and municipal debt markets mostly clear via DTC, the Depository Trust Company, which acts as a counterparty to most transactions in corporate and municipal debt, permitting settlement to be delayed for two days or longer. As I understand it, the London Clearing House’s EquityClear SA, despite its name, plays a similar role in Europe. Such clearing houses centralize debt-transaction counterparty risk in much the same way that banks and bank clearing houses centralize counterparty risk for cash transactions; without clearing houses, buyers of debt would be taking the risk that sellers would take the money and run without ever handing over the bonds they were ostensibly selling.
Clearing house insolvency is a likely outcome of the next world war, however, and represents the kind of systemic risk that it’s very difficult to diversify away.
The obvious solution is to clear the trades with the Ethereum blockchain or something similar, so that the transfer of debt title from the seller to the buyer is a single atomic transaction with the transfer of funds from the buyer to the seller. This eliminates counterparty risk from the bond markets as long as the blockchain’s integrity remains secure.
The actual coupon payments on the bonds, or title to collateral to secure those payments, could be incorporated into the same smart contract, but this is much less essential — the lenders are at least theoretically aware of who they’re lending to and assuming the risk that that counterparty (the borrower, that is, the issuer of the bonds) will default. The whole purpose of the borrower issuing debt in the first place is so that they can invest the money thus raised, so you would expect that for most of the lifetime of the bond, they won’t have an account with enough money in it to pay off the bond. So there’s little point in making the coupon payments nominally automatic via a smart contract.