A challenge in the “blockchain in capital markets” space is representing national currencies on a ledger in order to settle trades or act as collateral. When dealing with the US dollar, the Fed is the ultimate authority in determining what constitutes and does not constitute a dollar, an authority it also delegates to participating banks. Thus, until the Federal Reserve implements a “Fedcoin” there is no way to truly represent dollars on a distributed ledger. What can be represented however are IOUs. Commercial money issued by credit worthy institutions can be tokenized and used on a distributed ledger. As long as the tokens are easily created and redeemed, arbitrageurs (and issuers) can ensure that they trade within a narrow band around the dollar.
The problem becomes even trickier on decentralized networks such a Bitcoin who, by their nature, aren’t prone to attracting the most credit worthy issuers. People have suggested engineering “StableCoins” on such networks, tokens capable of tracking the value of national currencies. Such contraptions are born out of the constraints faced by these anonymous decentralized networks and thus shouldn’t be relevant to private blockchain applications for capital markets which have no shortage of credit worthy issuers. Despite that, they regularly come up in discussions around such applications. This is a reflection of the fact that many people developing technical solutions in this space have a poor grasp on the different set of constraints and threats faced by financial institutions as opposed to the anonymous participants of a decentralized cryptocurrency.
Not only is the use of “StableCoins” dubious for capital markets application, it doesn’t make a lot of sense in decentralized networks either. There have been several schemes proposed. In many cases, they can be dismissed for not passing a simple test. As Arthur Breitman phrased it: “no convertibility, no parity”.
This is one of the strongest laws of financial economics. It’s a lesson that Argentina has had to learn several times. Pegs can not be obtained by saying so. Many early proposals for StableCoins involved issuing tokens and willing them into trading at parity with the US dollar by calling them dollars. In fact, this law is so strong that it even applies to dual listed companies. A famous example, cited in Froot & Dabora  is that of Royal Dutch and Shell. Though the two companies had merged, Royal Dutch traded at a 10% premium! This is despite the fact that both securities entitled their holders to the same voting rights and to receiving the same dividends. Their prices should have been identical, but without the possibility of convertibility, arbitrageurs could not create a single price. Arbitrage, not rational behavior, is the workhorse of the efficient market hypothesis.
Since the early “StableCoin” schemes that attempted to will parity into existence, two schemes have surfaced: The simplest one attempts to adjust the quantity of tokens in the network to balance the exchange rate. While this can and does achieve parity, it does so in a trivial and uninteresting way. Holders of the tokens do not wake up to see that the values of their tokens has increased or diminish – they are still worth $1 each – but they do wake up to find that the balance in their account has changed! All this achieves is a pointless change of numeraire. It might help with menu costs but it does not solve the problem of collateral. (Yes, these have been proposed, often under the guise that “but the price goes up, so your balance just goes up anyway”.)
The other, somewhat more sensible, approach is to use a cryptocurrency as collateral for IOUs. In this model, such an IOU can be redeemed automatically at market rate against a collateral pledged in the native tokens of the blockchain. We might imagine some form of smart contract where a bitcoin holder could lock up some bitcoins to back some USD IOUs. In this case, the collateral is pooled and there is a single type of IOU issued against this collateral. The smart contract would require an oracle to be made aware of the prevailing exchange rate when redemptions happens, but this is a solvable problem.
The real question is what happens when the value of the cryptocurrency drops with respect to the dollar. In principle this should trigger a margin call. But if no one is ultimately responsible for the liability, who will place this extra collateral?
If the collateral does not increase, the risk discount on those tokens will increase, pushing towards more redemptions. This, in principle will increase the amount of collateral. For instance, suppose that a bitcoin trades at $1,000 and that one bitcoin of collateral is used to issue $500 of debt. Assume that the pool of collateral contains 100 bitcoins and thus $50,000 of debt have been emitted (a steep haircut). If the price of Bitcoin falls to $800 but 25 bitcoins, or $20,000, worth of debt are redeemed, then the pool of collateral will contain 75 bitcoins (and be worth around $60,000) and back $30,000 worth of debt. We are back at the same ratio.
In principle, the risk discount on a token creates a mechanism by which the collateral can adjust to changing exchange rate. However, those redeeming their tokens may want to sell the cryptocurrency in order to convert it back into dollars, and this might depress the price of the currency further creating a vicious spiral leading to insolvency.
Such a scheme might perhaps be workable with a stable and liquid reserve asset. Indeed, when the CME acts as a central counterparty in the futures exchanges, they are essentially creating fungible “oil” tokens or “wheat” tokens even, in some cases, in the absence of physical delivery in the contract. However, in the case of cryptocurrencies, where the markets are shallow and the swings are wide, this amounts to a bad tradeoff where counterparty risk is mitigated and fungibility is obtained at the cost of a dramatic increase in systemic risk.
Though there exists many possible variations in the scheme, they all end up shuffling credit risk one way or another, but they cannot make it disappear. Nassim Taleb, the self-described risk practitioner has described these approaches as “gaming the moments”. Small variations are eliminated, but rare variations become potentially much larger. StableCoins (or Argentinean pesos) can remain stable for a while, but then tend to fail spectacularly.
Pegs can’t exist without arbitrage; arbitrage can’t exist without convertibility.