Most public protocol arbitrage doesn’t exist

At CoinFund/Grassfed Network we pay close attention to “protocol arbitrage” as part of our generalized mining program. These are alpha-generating opportunities that allow participants to obtain some digital asset cheaply by interacting directly with a decentralized network, protocol, or mechanism. The participant can then try to sell that asset at market price to earn an arbitrage-like spread.

Many examples of “protocol arbitrage” today are in the form of cryptoeconomic mechanisms that live on open, public networks. For example, in the MakerDAO ecosystem, one can liquidate collateralized debt position smart contracts (CDPs) when their collateral falls in value beneath a threshold. The liquidator then earns a fixed 3% spread from the protocol by buying out discounted Ether collateral and selling it on the market.

If you’re willing to stomach the risks of nascent smart contract technology, this seems like a healthy spread. In fact, participants have done some impressive optimizations on CDP liquidation, sometimes transacting as much as $5 million and taking $150,000 in spread within a single transaction.

But if someone were to ask me whether CDP liquidation is a great business, I would respond that, emphatically, it is not. There are two reasons for this — the “traditional” explanation and the “cryptonative” one — and I will tell you both.

The traditional explanation deals with proprietary information. If the CDP protocol is public, then the lucrative opportunity engenders intense competition for the arbitrage’s spread across all participants globally. Even though the spread is fixed at 3% today, someone might be willing to pay more gas in order to front-run a competitor’s liquidation transaction and earn 2%. In this fashion, the spread erodes to 0% and the capital requirement for a participant tends toward infinity, while the strategy throughput (how many liquidations are available) is fixed or bounded. In other words, the traditional reason why public arbitrage opportunities are not good businesses (for most participants) is because the spread quickly erodes under direct competition.

But there is an even stronger, cryptonative reason why CDP liquidations are not a good business. Ultimately, Maker backs practical applications and user-facing technologies, but the possibility of liquidation creates massive holes in user experience. For this reason, there is demand for solutions that protect users from liquidations. These solutions today come in the form of cryptonative insurance and automated utilities which do precisely that — manage CDPs on behalf of users, preventing liquidation. So as it turns out, though open competition is a problem for Maker protocol arbitrage spreads, a bigger problem for spreads is that they may not exist in the first place.

In summary, “protocol arbitrage” is an extremely interesting area of public decentralized networks. But many such opportunities might have a short shelf-life because of the public nature of protocols, open competition, and unexpected developments in the markets, such as cryptonative insurance. As spreads erode, one must be a more and more capitalized participant in order to profit. As opportunities erode through automation, insurance, and other cryptonative mechanisms, the average participant will find it harder and harder to generate alpha.