Basis (previously known as Basecoin) recently raised $133M to construct a supposedly price-stable cryptocurrency through a system that algorithmically manages the money supply. Unfortunately, to borrow a phrase from Nassim Taleb, Basis is a ‘Bob Rubin trade’ and should be treated with healthy skepticism. It effectively transfers tail risk to the people it is meant to protect (stablecoin holders) and rakes in potentially massive short-term profits before the system potentially collapses. This Bob Rubin trade works as follows:
- Sell a so-called ‘stable’ asset with significant hidden tail risk
- If things go well, collect short-term profits
- If/when things go bad, exit the market ⇒ losses left with ‘stable’ asset holders
Importantly, this does not mean that Basis is something malicious like a Ponzi scheme. The Basis team may honestly believe it’s a good idea. Unfortunately, the world is full of economic ideas that people thought were good but turned out to be too fragile. It’s possible to transfer risk without intending to simply through using bad assumptions, models, and design.
Of course, I encourage you to check out their design for yourself on their website www. basis.io. The following describes the way I see things. If anyone on the Basis team thinks I am wrong, please do reach out.
How Basis proposes to work, and how it can fail
Basis aims to control the coin supply such that the market price of the stablecoin remains close to its target of $1 USD. This involves dynamically increasing/decreasing the coin supply as demand changes. In a manner similar to fiat currency, trust in the system is required in the initial setup of the stablecoin to assert the prescribed target value. Following the initial setup, Basis proposes to dynamically increase/decrease the coin supply in response to demand in the following ways:
- Should the price increase above the target, the system creates new coins to increase the coin supply until the price regains an equilibrium around the target. These new coins are awarded as dividends to stakeholders in Basis (owners of their investment tokens), who can then sell them on the market.
- Should the price decrease below the target, the system attempts to sell bond tokens that incentivize stablecoin holders to destroy their stablecoins in exchange for future interest, which acts to decrease the money supply and bring the equilibrium price back in line with the target. This bond interest is paid the next time the system mints new stablecoins to increase the coin supply (importantly, this is not guaranteed to happen).
I see the following problems with this:
- The stablecoin is subject to catastrophic failure. The bonds are in fact very risky derivatives. If the market for these dries up, as is prone to happen during extreme events, the stablecoin collapses. Thus any growth in the Basis system is likely to be transient.
- Any profit from transient growth in the system is purely extracted by investors when they receive and sell new stablecoins. Stablecoin holders have no claim to the value of this growth as there is nothing collateralizing the coins. This is especially important if/when the stability fails. This fulfills the ‘Bob Rubin trade’ as there’s no reason to think the growth is lasting.
The stability of Basis will rely on always having enough agents willing to take the risks associated with each position in the network. I.e., when necessary, there need to be enough agents willing to take what is essentially a leveraged crypto position and/or buy the bonds with the promise of future interest. In a normal domain, there are probably such agents. But these positions make sense only if you believe that the system will work. The bonds’ future interest relies on the system recovering–i.e., the price rising above target again so that the system mints new coins to award bond holders. Agents may not be willing to bet on these bonds during extreme events that affect faith in the system or cryptocurrencies in general (these are the ‘tail events’ that I refer to). This self-fulfilling issue shouldn’t come as a surprise as the same issues affect central banks’ control of fiat currencies (although of course in that case, the currencies also have the backing of powerful governments and transient profits aren’t extracted by investors).
In essence, the Basis system is very similar to existing crypto-collateralized systems, such as BitUSD and MakerDAO’s Dai. Basis creates value out of thin air by prescribing that the coins have value whereas the crypto-collateralized version is based on a blockchain network that has already done this (note that this point isn’t criticism: everything must start at this point somewhere). In essence, Basis is sort of self-collateralized. The stakeholders who receive income when new coins are sold are in a similar position to the collateral owners in the crypto-collateralized system: they earn any profits from increased value in the system and face a loss of their input (the price they paid for the stakeholder position) should the system fail. The stablecoin holder, however, gives away all potential return from holding a crypto-asset in exchange for ‘medium’ volatility insurance. In particular, in tail events, the stablecoin takes on the direct tail risk of something like cryptocurrencies (in the case of Basis, this would be something like an unestablished altcoin as opposed to ETH in the case of Dai). As a result, the expected return of the stablecoin should in fact be negative as it is either worth the target price–e.g., $1–or, in the tail, significantly less.
Additionally, like all proposed stablecoins, Basis also relies on trusted oracles to provide real world price data, which could be subject to manipulation. This is a separate issue from its economic fragility; I won’t focus much on it here except for the following notes. There are ways to create decentralized oracles (I encourage you to read about Schelling games); notably, Basis mentions these in their whitepaper. Unfortunately, these can fall apart when collusion is possible. Collusion in a decentralized oracle would directly influence the market price of the stablecoin and would potentially trigger one of the previously mentioned tail events.
Attempts to patch the problems fail
The Basis team has attempted to patch the fragility problem by using bond price floors and bond expirations. I encourage you to read more about it here. The price floor is intended to stop more and more bonds from being created in a downward price spiral. The bond expiration is intended to eventually trigger bond prices to increase as it shortens the queue of bondholders waiting to receive newly minted stablecoins. Unfortunately, this fix doesn’t solve the fundamental problems:
- The price floor just means that the mechanism stops working entirely. If this is invoked, there was something causing massive downward pressure on the stablecoin price in the first place. As the bonds were the only mechanism addressing this, the price can continue decreasing.
- The expiration of older bonds will indeed tend to make new bonds more attractive. However, as a consequence, the expiration makes all bonds less attractive. These bonds are already very risky derivatives; this just makes them more risky. And so the market for them will dry up even faster during extreme events.
While I have no doubt that the Basis team has models that show their design can be stable, as they claim, the same thing has been said of nearly everything that eventually leads to some financial crisis. Such models should be taken with a grain of salt.
The Basis team further contends that bond expirations can lead to ‘graceful re-pegging’ of the stablecoin after bonds expire (e.g., 5 years later) in the aftermath of a catastrophic collapse. However, there are good reasons to suspect that this either doesn’t happen or that this is not ‘graceful’. At such a point, the market for the stablecoin may have completely dried up, in which case the direct reduction of coin supply by selling a new bond may not have an upward effect on price. In this case, an agent purchasing bonds (recall that this would additionally have to be above the bond price floor, meaning that the profit from this would be limited for the risk involved) would want to directly reignite the dead Basis system. However, if this is their goal, they may be better off starting a completely new version of Basis instead, in which they can own the investment tokens that receive newly minted coins (after accounting for any bondholders). On the other hand, a dead Basis market may be susceptible to cycles of pump-and-dump action, which may instead be the motivation behind an agent entering such bond positions. This is hardly the stability that stablecoin holders sign up for, and it’s unlikely that such behavior would instill further confidence in a broken system.
This isn’t to say that Basis can’t work, at least for a time. It can be stable within an assumed stable domain. However, it’s an inherently fragile design and is, in a probabilistic sense, long-term unstable. I appreciate that the Basis team seems to at least recognize some of these problems. Indeed, designing a robust stablecoin is very difficult. This said, however, Basis is very good for speculative investors at the expense of stablecoin holders despite the design’s fragility. This is my main criticism. It gives the possibility to make huge short-term profits (i.e., all of the short-term profit of launching a new cryptocurrency since all potential returns are being shifted to speculative investors) and then cash out the profits from expanding the coin supply before any crashes occur. Even if, in the end, speculators’ stakes in the platform become worthless, their profits can still exist if they exchanged them for other assets than the stablecoin. Perhaps this sheds some light on why VCs are willing to pour such a massive sum into this idea. There is a lot of profit potential with little downside for the company as long as people are willing to buy their product and take on the tail risk, at least transiently. But ordinary stablecoin holders will be left holding the bag when an eventual crash occurs.
Basis is a repackaging of an idea that has already failed
In fact, the basic designs behind Basis have already failed before. This is perhaps best exemplified by the failure of Nubits, one of the first stablecoins dating back to 2014. If Nubits becomes ‘undervalued’, it relies on (1) people being willing to go long the currency to ‘arbitrage’, (2) people being willing to lock up their Nubits in a bond-like instrument that promises future returns when the currency becomes overvalued next (should this even happen), and/or (3) being able to sell new NuShares (representing some sort of interest in the system that becomes profitable if the coin becomes overvalued again) to in essence recapitalize the system. Obviously, people may not be willing to take these positions if their faith in the future value of Nubits is shaken, as Nubits experienced in 2016 and is ongoing currently in 2018.
The stability mechanism works perfectly if prices get too high–just mint new coins and sell on the market and give profits to Basis investors. But it can fail if prices get too low. There is no reason to think that enough agents will be willing to take the required positions if confidence in the system is lost. This confidence can be rather whimsical and has caused Nubits’s liquidity crises. The main difference between Nubits and Basis is that Basis proposes to operate in an algorithmic/automatic way whereas Nubits relies on DAO-coordinated action. Is this enough to patch the problem? Maybe it helps–although this importantly remains to be seen–but it doesn’t prevent the root cause.
If we look back at past Nubits forums (circa 2014), we see that talk about Nubits sounds like current enthusiastic talk about Basis–and for good reason, the underlying mechanisms are nearly identical. Basis enthusiasts should be paying more heed to the current forums lamenting the failure of Nubits.
The stablecoin space going forward
In general, the stablecoin community seems to be caught in a cycle of rediscovering the same issues but then repackaging the same fragile ideas as solutions. The space needs a paradigm shift toward new and anti-fragile ideas. The existing stablecoin projects all work by directly porting old financial system ideas on-chain. Ultimately, I suspect that solutions to the hardest cryptocurrency problems–price stability as well as transaction scaling–will require new ideas that build uniquely on the context of blockchain. The purpose of my writing is not simply to criticize but to help facilitate this shift.
If this area of research interests you, please do get in touch. I’d love to get the ball rolling on new ideas. I am working on some ways to do this better (at some point, I may write a followup article introducing some of these).