How Stable Are Stablecoins?

In the current war between stablecoins, one thing seems to be troublingly absent: actual numbers.

Everyone seems happy shilling their favorite project or the entire stablecoin movement, which isn’t wrong in principle. However, we still very much lack comparative, data-backed analysis of the current stablecoin ecosystem.

So we thought we’d have a look at the data, compare the most popular coins and discuss the results. Also, we created a short Jupyter notebook that lets you play around with our analysis, if you want a deeper understanding of the stablecoin environment.

Letting Loose

While not an exhaustive list, these are the stablecoins that we used in our analysis:

• Tether
• TrueUSD
• USD Coin
• Paxos Standard Token
• Dai
• Gemini Dollar
• BitUSD
• nUSD
• CK USD

To start with, we wanted to answer the most obvious question first – how stable are stablecoins, really?

Stablecoin volatility might sound like an oxymoron, but it’s very much a daily occurrence. We’ve witnessed an extreme example of that just last month, when Tether briefly lost its peg to the dollar and was at one point trading as low as $0.86 on some exchanges.

To find out which stablecoins are the most and least stable, we began by comparing the daily standard deviations of all coins.

For those unfamiliar, standard deviation is a traditional statistical measure of variance or volatility.

A low standard deviation means that the data points tend not to stray (or deviate) far from the so-called ‘expected value’. In case of our stablecoins, the expected value should be more or less = 1 USD.

Below are the daily standard deviations (in %) of stablecoins for the past year:

Even with recent turmoil, Tether remains squarely the least volatile stablecoin of 2018, deviating less than 0.5% – or less than half a cent – from the mean on a daily basis.

Conversely, BitUSD has proved the most volatile of the lot, with daily fluctuations of just over 2.5%. To put that in perspective, the next most volatile coin – CKUSD – has registered daily deviations of just under 1.5%

And while many continue to dispute the merits of centralized stablecoins, they have, on average, proved more stable than their ‘decentralized’ counterparts. In the end, not even Tether’s blatant lack of transparency and mounting gossip managed to move the needle by much.

To prove that, let’s look at standard deviations for the first week of November. The order of stablecoins is kept the same to avoid confusion:

As we can see Tether is now the second least volatile coin, which likely ought to be attributed to its recent turmoil. In general though, the observed stablecoins seem to remain quite stable even in the short term.

For additional context, we also looked at large outliers, or the total number of days that a coin’s price moved more than 5 cents either way.

The results are mostly in line with the assets’ standard deviations, with only 3 such days over the last year for Tether, 1 for USD Coin and Paxos, and a whopping 99 days for BitUSD:

Cause and Effect

Still, comparing stablecoin volatility is only half the story. There’s another equally important question that we need to answer – what causes this volatility in the first place?

Common sense tells us to focus on two usual suspects:

1. Volatility in a coin’s trading volume
2. General market volatility

First, we used linear regression to examine the relationship between volume and a coin’s daily standard deviation, on a scale from 0 to 1.

Simply put, the higher the number, the more we can blame a coin’s volatility on a strong change in its trading volume:

(For the advanced: The value we are looking at is the so-called r squared (R^2) which describes the amount of explained variance. The value lies between 0 and 1, where 1 means that our model explains 100% of the movement in the data. If you’re more used to interpreting correlations, the correlation is simply the square root of the R^2 value)

As we can see, the results are a mixed bag. The stability of some coins, like Tether, TrueUSD and CK USD, seems to be significantly impacted by a daily change in trading volume.

Others, like Dai, Paxos and BitUSD appear mostly unchanged.

Now let’s see what happens when we plot the coins against general market volatility. To simplify the analysis, we used Bitcoin as the proxy for the entire crypto market, as it still accounts for more than 50% of the current market cap.

Again, the higher the number, the more we can assume that the general market (Bitcoin) movement affects a stablecoin’s price:

Right out the gate, we can see most assets reacting more intensely to general market volatility.

With the exception of nUSD (which seems to be an outlier in both cases and warrants further research), no stablecoin stayed immune to daily BTC price action.

Finally, we used the combined model – essentially conjoining these two factors – to see if it can explain stablecoin volatility better than either of them alone. Here are the results:

The longer the purple line, the more our combined model successfully explains the volatility of a particular stablecoin.

As you can see, in 6 out of 9 coins, a mix of trading volume and general market movement can be blamed for ~80% of its volatility.

If you want to dive in even deeper, a more comprehensive data sheet can be found in the appendix.

How can I use this data?

All this being said, volatility is hardly the ‘end all be all’ when considering which stablecoin to buy. Based on your inclinations, you might even prefer a slightly more volatile asset if it means less long-term risk.

For example, Tether may be the year’s most stable coin, but many would still put it in the ‘most likely to implode under towering misuse allegations and FUD’ bucket.

Whatever the case, if you plan on using stablecoins for hedging, it’s imperative to understand an asset’s general volatility and its underlying causes.

For investors, there may be room for some interesting trading strategies built around comparative stablecoin volatility.

With exchanges like Binance now offering several stablecoins to trade, one might find it increasingly appealing to try trading differences between stablecoins. Here we could use our newly acquired knowledge to invest in low-volatility stablecoins, while trading outbreaks in their relatively unstable counterparts.

Also, since we’ve seen that most stablecoins are at least somewhat affected by general market volatility, it could prove useful to look for assets that react less to daily BTC price action, to maximize the spread.

Those involved in arbitrage are probably already aware of these possibilities. However, as the recent Tether drama demonstrates, it can’t hurt to keep track of the stablecoins available on the exchange of your choice, and the significant differences in volatility between them.

Appendix:

What’s your take on stablecoin volatility? What aspect of stablecoins would you like us to tackle  next? Join the discussion over at our Discourse!