Whether it’s the infamous Tether, a newer fiat-backed stablecoin, or innovative projects like Maker, almost everyone in the blockchain space recognizes the value and necessity of a safe, reliable stablecoin. There’s plenty of debate over which to use, but let’s take a more analytical dive into which stablecoin is best at actually doing it’s job, i.e. being stable.
(I won’t rehash another intro piece on stablecoins; if you want a quick rundown of their usage and history, I recommend this excellent article from Bitcoin Magazine).
First, the terms of this analysis. For this comparison I’ve selected 6 stablecoins based on the criteria that they are pegged to USD ($1.00), currently functional, and actively traded: Tether (USDT), USD Coin (USDC), TrueUSD (TUSD), Paxos Standard Token (PAX), Gemini Dollar (GUSD), and Dai (DAI). Tether needs no introduction; it’s the oldest and by far the most used stablecoin, as well as the most controversial. Four others (USD Coin, TrueUSD, Paxos, and Gemini Dollar) are also USD/fiat-backed stablecoins, banking on the value prop of “just like Tether, but not a scam” with increased transparency of their balance sheet and operations, and the ability to actually redeem the tokens for USD. Lastly we have Dai (DAI), the preeminent cryptoasset-collateralized stablecoin and part of the Maker system.
Since three of the stablecoins in our sample were only launched in September and October of 2018, our observation period will be the 60 days from October 31 to December 29. Two other active stablecoins are excluded due to lack of liquidity (bitUSD with $6K of daily trading volume, only on BitShares-related exchanges) or because in practice they aren’t functioning as a stablecoin (Steem Dollars have traded on average at $0.79 to the dollar in the observed period, and with volatility 9 to 10 times greater than any other stablecoin analyzed here).
Here is the state of stablecoin usage among our sample — average daily trading volumes in the past 60 days, 30 days, and the last day of the observed period.
Clearly Tether is still king with the lion’s share of stablecoin trade volume, despite questions about its balance sheet and a Department of Justice investigation into whether its founders have used USDT to manipulate the price of BTC. Tether’s dominance is not simply a question of liquidity/availability either, as all 6 of the stablecoins are traded on at least one Top-10 exchange by trading volume. But most-widely used does not necessarily mean best, so let’s see how the competition actually stacks up.
Which stablecoin performs the best?
First, that depends on how you define best. Is it the stablecoin that stays closest to its $1.00 peg? Is it the stablecoin that stays at or above $1.00 most often? Or is it simply the stablecoin that exhibits the least day-to-day volatility, staying true to its name? Lastly, how does each stablecoin perform in a market downturn, when retaining value is most important?
Metric 1: Deviation from $1.00 peg
The first measure looks at each stablecoin’s high and low price on each of the past 60 days, and records the most it deviated from its $1.00 peg in either direction.
Deviation = Max(High Price — $1.00, $1.00 — Low Price)
For example, on 12/28 TUSD traded at a high of $1.02 and a low of $0.99, so its deviation for that day is Max(1.02–1.00, 1.00-0.99) or .02. The chart below highlights greater deviations in darker red.
The average daily deviation for each stablecoin was pretty similar, $0.02 to $0.03 from the $1.00 peg. But we really start to see separation is the max deviations, where GUSD deviating more than $0.10 from the $1.00 peg on 5 of the 60 days, including an $0.18 deviation on 11/14.
Winners as measured by deviation from $1.00 peg: USDT, PAX, TUSD
Losers: GUSD, DAI
While deviation from $1.00 may be the best way to judge a stablecoin’s effectiveness in an academic sense, the reality is that most people holding stablecoins are just fine with deviation to the upside! A common use case is purchasing a stablecoin in a falling market, when simply maintaining value is the goal — any increase is the stablecoin is just a nice bonus. So with that in mind, a more practical judge of stablecoins’ effectiveness may be downside deviation from the $1.00 peg.
Metric 2: Downside deviation from $1.00 peg
As the name implies, downside deviation is simply our prior metric but measured only for deviation below $1.00.
Downside deviation = If Low Price < $1.00, Low Price — $1.00
With that change, we see a quite different result:
Among the fiat stablecoins, it’s Tether that now shows the most risk to the downside, dropping as low as $0.945 on November 4th. What happened to our previous offender, GUSD? Turns out it never dropped below $0.95; all those $0.10+ deviations we saw were to the upside, and it traded as high as $1.18. Among all stablecoins, in the past 60 days it was actually DAI that showed the most downside deviation, dropping to $0.931 on November 4th.
Winners as measured by downside deviation: PAX, USDC, TUSD
Losers: DAI, USDT, GUSD
You may have noticed that the average downside deviation is no more than $0.01 for any of the 6 stablecoins. It’s actually more common for stablecoins to trade above $1.00 than below! That doesn’t tell us a lot in terms of risk; it’s typically only when market conditions really turn south that stablecoins have trouble maintaining their peg. More on that later.
In the case of cryptoassets meant to be used as currencies, the most important measure may not be how close it trades to the dollar at all, but simply how stable its price is. For a merchant, being able to assume the same value of a stablecoin day-to-day is critical regardless of what the exchange rate is. Which leads us to:
Metric 3: Volatility of daily price changes
We’ll define stablecoins’ volatility as the standard deviation of daily price changes over the observed period, expressed on an annualized basis.
σ( Δp1…. Δp60) x √365
For context, I’ve also included figures over the same period for major currency pairs (EUR/USD, GBP/USD, USD/JPY), equities (S&P 500), and major cryptoassets (BTC, ETH) for good measure.
By this measure, all of the stablecoins are considerably more stable than the major cryptoassets, showing volatility between 1/10 and 1/2 that of BTC and ETH. And while the top stablecoins are still more volatile than fiat currency pairs, 9.2% to 13.1% is not a huge gap — with further adoption we could easily see stablecoin volatility reduced to that of major fiat currency pairs.
Winners as measured by volatility: PAX, USDT, TUSD, USDC
Metric 4: Market correlation in a downturn
Let’s return to the issue of how stablecoins perform in market crashes. It’s one thing for a stablecoin to remain stable in a calm or growing market, but most people turn to stablecoins in times of uncertainty or market declines. Of particular concern is how correlated stablecoin prices are with crypto markets, and whether they will decline in value at the same time as the crypto markets. Below is the downside deviation chart for the crypto markets’ 3 worst days in our sample:
This is where things can get really concerning — on 11/24 BTC is down almost 11%, but USDT and GUSD are also down 4% / trading at $0.96. Extrapolate that to a 30% market decline and you could lose 12% of your value holding a supposedly risk-free stablecoin. This correlation of declining prices is more common among risk assets; you may recall in the 2008 financial crisis when diversified portfolios of bonds, stocks, and real estate saw their values plummet as asset classes suddenly became correlated. But for an asset meant to maintain value during a “flight to safety” like a stablecoin, it’s disastrous.
Winners as measured by market correlation in downturns: PAX, TUSD, USDC
Losers: GUSD, USDT, DAI
And the winner is…
Based strictly on our quantitative analysis, USDC, TUSD and PAX all performed reasonably well when analyzed by deviation, downside deviation, volatility, and market correlation in downturns. It seems these newer-gen fiat-backed stablecoins have made material improvements over Tether, both in their actual performance and their transparency to users.
Despite not performing quite as well by the metrics in this analysis, DAI also gets credit for being by far the most successful crypto-collateralized stablecoin to date, and for generally maintaining its value with ETH collateral despite the ETH’s 90% drop in price since DAI launched in December 2017.
To be clear, this is only an analysis of quantitative performance; to accurately judge the best stablecoin, there are many qualitative factors to evaluate as well. For fiat-backed coins you need to judge the counterparty/credit risk of the issuer and the credibility of their financial audits. It’s also worth nothing that the regulated stablecoins (PAX, USDC, TUSD and GUSD) all appear to have backdoors for law enforcement to access and freeze tokens, which were required by financial regulators and were fully disclosed by the project teams.
As far as crypto-collateralized coins, there is always some level of risk of smart contracts being hacked or code otherwise malfunctioning. There is also the possibility that if cryptoasset prices fall too much too fast, even an over-collateralized stablecoin could break its peg. In the case of DAI, the protocol’s last resort of printing and selling MKR on the open market is interesting in that it effectively dilutes the “equityholders”, but it will only work in moderate or gradual market declines. Should Ethereum face an existential crisis that leads to a catastrophic drop in the price of DAI and its ETH collateral, no one will be rushing to buy an ERC-20 token representing the stability of a ERC-20 stablecoin! That’s like trying to raise an equity round for Bear Stearns in March 2008 — you’re only going to find buyers at fire sale prices. In fairness to the Maker team, they are close to launching DAI backed by multiple forms of collateral, which should help alleviate this issue, and more generally this risk should subside as crypto volatility decreases with adoption.
Another risk I don’t see mentioned enough is the natural tendency for stablecoin holdings to be concentrated at exchanges, since, obviously, that’s where the trading is happening. As an example, 73% of Gemini Dollars appear to be currently held at Huobi-controlled addresses (the top three addresses shown here). Now you are wrestling with the counterparty risk of both the issuing entity (Gemini) and an exchange holding enough of the stablecoin to negatively affect the price/stability (Huobi). Gemini and Huobi have (as far as I know) both been good actors in the crypto space thus far, so that doesn’t mean you should automatically wary of using Gemini Dollars, but it illustrates the potential risks of centralizing ownership of any coin or token. (On that note, does anyone have a thorough list of exchange wallets they can point me to? I’d love to investigate this further…)
Regardless of which stablecoin you choose to use, it’s great to see so many new stablecoin offerings and as a cryptoasset class they will no doubt continue to develop and improve. Ideally we’ll see a wide variety of offerings from highly regulated fiat-backed stablecoins, to crypto-collateralized decentralized stablecoins, and even algorithmic stablecoins (Basis may have folded for regulatory reasons, but others such as Reserve are coming). Here’s hoping a reliable suite of stablecoins will enable the next round of decentralized apps coming to market in 2019.