Last week, the stablecoin project Basis (formerly Basecoin) announced a $133 million investment from top tier investment firms. It is one of a number of stablecoin projects that aim to produce a token with a stable value relative to USD.
In this post, I will discuss the economic issues with the stabilization mechanisms proposed in their whitepaper, and also identify some characteristics that do contribute to stability. While I use Basis for concrete discussion, the same ideas apply equally to other stablecoin projects to the extent they share the relevant characteristics.
Using Loans To Pay Rent
The whitepaper proposes to even out price fluctuations in token price by expanding Basis supply when price rises above the intended peg, and issuing bonds when price drops below. Although their sale contracts supply, bonds are a future liability that not only matches the amount of contraction but exceed it by the bond yield. Should the price drop again and bonds are required in the future, future bond purchasers, being lower in the repayment queue, will demand higher yields to offset the greater risk of default or expiration. This liability diminishes confidence in the system, creating downward pressure. If the currency price fell below the peg due to investor confidence waning, extra liability gives them more reason to be wary.
Basis proposes to address this issue by allowing the currency to depeg instead of issuing ever more bonds and ever higher yields. However, nothing comes for free. An efficient market will factor in all characteristics of the system, whether it’s chance of depegging or bond expiration, so an action that does not create value is unlikely to result in a net increase in prices. Depegging is obviously undesirable to token holders. Adding to the bond queue causes a heightened risk of depegging, which makes the currency less desirable to hold, still contributing to downward pressure.
Intuitively the downward pressure should exceed upward pressure from supply contraction due to the net loss in the bond yield, although we cannot prove this apriori. If upward pressure does not exceed downward pressure, the price may stay unpegged and possibly oscillate. Even if upward pressure wins out and the peg is restored, the presence of a bond queue is a drag on the future ability of the system to stabilize itself. Present price stability, if achieved, comes at the cost of increased risk of future instability.
Saving up in times of plenty to be used in times of need would be a sound idea, but that’s not what happens here. In times of plenty some of the excess is spent as dividends, and in times of need a loan is used that requires repayment with yield. A similar personal finance strategy is employed by many who end up with questionable credit scores.
It may be reasonable to borrow against the future in circumstances where future volatility is expected to be less than present volatility, but that won’t be true in the long term equilibrium state where a baseline amount of market volatility remains.
“Nothing comes from nothing. Nothing ever could.”
— The Sound of Music
Self Buybacks Do Not Raise Prices
Another way to analyze expansion and contraction mechanisms is by analogy with company stocks. If per-share price rises due to market forces, there are mechanisms a company can employ that will decrease its per-share price. One is to pay out dividends from the company’s cash reserves, which reduces market cap and per-share price by a corresponding amount because an efficient market will account for the company’s reduced cash holdings. This option is not viable for a token, which has no money in the bank. Instead, the token can expand supply, which in stock terms would be a stock dividend. Number of shares increases while market cap remains the same, which lowers the per-share price. By this analogy, we see that indeed per-token price will be lower when Basis supply is expanded and distributed.
Now what happens when per-share price goes down? Basis issues bonds to buy back its own tokens. This is similar to a share buyback for stocks. However, in general share buybacks do not drive up per-share price. This is a tricky point. Share repurchase is a way to return capital to investors, not a mechanism to drive up prices. The number of outstanding shares is reduced, but the company must spend its reserve capital to do so, resulting in a lowered market cap corresponding to the amount repurchased. Any movement in per-share price is due to the market reacting to the signals a buyback sends.
Basis issues bonds instead of using reserve capital. Assuming liability is much like reducing assets, except with the extra cost of a yield. Companies may prefer debt if they have better use for the capital, or if they are trying to dance around taxes. These reasons do not apply for a token. This analogy suggests that issuing bonds is more likely to lower token price than raise it.
What does raise per-share price is a reverse stock split. For example, if the system converts every 2 outstanding tokens into 1, then per-token price will double. However, this would only stabilize the nominal value of each token, which may be useful in certain scenarios such as keeping menu prices steady that are denoted in number of tokens. If the goal is for token owners to be able to convert their holdings to a stable amount of USD, then this mechanism does not help.
Security vs Currency
We have argued that bonds are unlikely to provide stability, but that’s not to say Basis will be just as unstable as other tokens. Let us continue by disregarding bonds and considering the two token system that remains.
I discussed in a previous blog post that many tokens serve a dual role as a security-currency, and by considering an abstract pure currency token, we were able to apply existing models to each type of asset separately. This currency separation is exactly how Basis and Share are structured. There is no expectation of appreciation of Basis, and nor does it reward holders with dividends.
From the previous analysis, we said that for a pure currency token that has only currency like characteristics, its market cap Mc can be modeled as
Mc = Y / V
where Y is the income generated by users via use of the token, and V is the income velocity of money taken to vary slowly and within a narrow range across time. For a stablecoin where use of the token is not required by any platform, users demand the token only to the extent they prefer it over alternatives. They may prefer the token over others due to stability characteristics, or they may prefer the token over fiat due to ease of access. Competing alternatives have a limiting effect on Y and therefore Mc.
Splitting of security and currency may naturally isolate the price fluctuations of securities, part of which are due to investor sentiment about the future, from the currency token. Volatility in currency exchange rates are lower than for equities. Furthermore, the expansion mechanism proposed removes any hope of appreciation in Basis due to growth in Y, so sentiment about the future may have less effect on price, except with regards to the risk of dropping below the intended peg.
What About Share Holders?
Since Share has only security like characteristics, we can say, as for stocks, that its value is based on the expected profit stream. Dividends come from expansion of Basis. In the early stages, we can optimistically assume system adoption and hence demand for the token to grow. Once user adoption reaches equilibrium, it would be reasonable to expect usage to grow only alongside the economy of its user base. The world’s developed economies grow at a real rate of about 2% per year, so let’s say there’s a supply expansion of Basis by 2% per year. This is also the rate of currency expansion I applied in my previous post. The same calculation shows that we can expect the market cap of Share to be Ms = 0.52 Mc, or about half of the market cap of Basis. This is a maximally optimistic scenario that assumes no loss due to bond yields.
We have not yet factored in any transaction fees and expenses such as mining costs. Their analysis would be the same as before so I will not revisit it here. I do note that it would be unrealistic to expect the 2–3% transaction fees charged by credit card companies. These fees are not charged solely for value transfer, but for a retail service stack from merchant to consumer. Credit cards offer concierge service and auto insurance among a range of benefits. A blockchain does not even have a 1–800 customer service number.
Nor would a currency that charges such high fees be any good as a currency. The service provided by a token compares better with Fedwire, a low cost bulk transfer infrastructure. Banks use the system and tack on their own, much higher fees in a retail wire service. On 2018–04–17, bitcoin earned 0.000147% of its market cap in transaction fees, which extrapolates to only 0.05% per year.
Summary
While stability will not come from the bond mechanism, some amount of stability is expected to come from isolating security value and capping appreciation of the currency token, removing reasons to hold the currency that are sensitive to future expectations. Downside protection is not guaranteed, but supply expansion should make the intended peg an effective cap on token price. While the price is unlikely to be stable, there is reason to believe price fluctuations to be milder than for uncapped security-currency tokens.
