Coin Listings: Is the Pop Real?
The year is nearly at an end, and one of the major trends of 2021 has been the explosion of DeFi. We’ve seen countless projects launch this year, and token prices have soared as DeFi adoption skyrockets. For many projects, getting listed on an exchange like Binance or Kucoin is a watershed moment, as many believe the leap from a DEX to a centralized exchange will push prices up. That got me thinking: is it a sound strategy to invest in coins newly listed on centralized exchanges?
To figure this out, I analyzed data from Binance and Kucoin. Binance is one of the biggest centralized crypto exchanges, with daily spot volumes of ~$20–40B per day in December. Kucoin, also one of the top exchanges in the world, does daily spot volumes of ~$2.5–4.5B. Looking at historical data from both exchanges, I evaluated how a simple trading strategy would perform if it went long on newly listed coins. I retrieved the listing dates for all the coins listed on Binance and Kucoin in 2021. Then I used CoinGecko to get their daily average price over the past year. Here are bar graphs of the price as a ratio of the listing price.
Woah! very strong evidence that there exists a Binance listing bump. Kucoin also has a listing bump but it’s not as big or as sustained as Binance, which makes sense as Kucoin lists more coins and does less volume. Another tidbit about this graph is that, on average, coins tend to spike the few days after they are listed. While that’s not true in every case, generally coins return to about their listing price after about 15 days.
So here’s an interesting strategy proposal based on these observations: as soon as an exchange announces they are going to list a coin, buy that coin, then sell it one day after it gets listed. Anecdotally, I’ve noticed coins tend to get announced about 1 day before they actually are listed. So I buy on day -1 and then sell on day +1. If I were to employ this tactic on Binance I would have a 1.34 return over two days. Quite good! On Kucoin, it’s not quite as good but still a healthy return at 1.199. But still, a 34% and a 20% return on average over 2 days is incredible!
Now of course, one issue with this line of thinking is that this data is likely to be heavily skewed. Taking a look at the distribution of prices one day after listing, we can see this fairly obviously. Most of the coins are centered around one with a few outliers way to the right. SHIB, for example, went up by a proportion of 2.25 the day after it was listed on Binance. Certainly an outlier in the data and not the norm.
So instead of looking at the mean, let’s look at the median. This will give us a much better sense of how each coin does individually.
Now Binance and Kucoin look very different when viewing it this way. Binance’s median still has a jump after one day while Kucoin’s essentially breaks even. What does this mean about our strategy? Well, now we can look at the median return for buying on day -1 and selling on day +1, which is still about 1.171 and 1.0219 for Binance and Kucoin respectively. Note that on Kucoin about 50% of coins had negative or 0 returns.
Another factor that may affect this strategy is what was going on in the market when these coins got listed. For example, early in the year the market was on a tear and prices were soaring, but right around May the crypto market crashed pretty much across the board. If you group the day after return by month you can see the trend.
The coins listed on Binance in those middle months when the market was down performed much worse than the bookends of the year during bull markets. The coins listed on Kucoin, however, seem much less dependent on market factors than those on Binance — but there are still peaks at extreme moments in the markets such as in February.
One final warning I would give about these conclusions: the data may suffer from survivorship bias. It’s possible that coins were listed and then promptly delisted from these exchanges, and because of the way I collected this data, I would not capture that. I looked for examples of this happening on Binance and I found only four such coins, so my guess is that this does not happen very often. I suspect that delisting happens infrequently, but be warned, this could positively skew the results.
-Noah, Trading at Dexterity Capital
