When Alt Season? A Lookback on 2020

The research below is an excerpt from our Q4 Quarterly Report in partnership with eToro! Read the full report here:

The TIE & eToro Q4 Quarterly Report

As arguably the most exciting year in crypto comes to an end, we wanted to take a moment to look back at the data and ask one age-old question: “when alt season?”.

Here’s an interesting question. If on January 1, 2020, you put all of your money into bitcoin, or into the top 100 altcoins equally, which would have made you richer by the end of 2020? The answer can be seen in figure 1, a figure that will be reappearing throughout this article. Surprisingly, altcoins were a more lucrative investment until December 18th. During bitcoin’s recent monumental rise, it has just barely outpaced the gains of the alts.

After looking at this figure, the question any active trader would be asking is how can one predict when to move from alts to bitcoin and vice versa. If you started with bitcoin, with only two swaps the entire year - bitcoin to altcoin in May, and back into bitcoin in September - the 3x or 4x gains seen by either investment grows to over 7x.

The goal throughout this article is to explore some potential drivers of two market conditions - so-called “BTC season” and “alt season.” Of course, this is all speculation - hindsight is always 20/20 (pun intended). It is interesting to think about the potential influences on bitcoin and altcoin performance as we consider actionable insights going into the new year.

Figure 2 shows the relative trade volume for bitcoin and altcoins (on average). Relative trade volume is an interesting metric - it is not scaled by the market cap of the coin. Rather, it is a measure of out-of-the-ordinary volumes for each individual currency. Specifically, it is the standard deviation of trade volume over time, respective to each coin. The way to interpret the purple line is high values (values above the horizontal black line) indicate when BTC has abnormally high trade volume compared to the alts. Conversely, values below the black line correspond to times with abnormally high trade activity across the 100 altcoins.

If it looks like bitcoin is trading abnormally high, you should own bitcoin, and if alts are trading more, you should own the alts. This makes intuitive sense; if alts have more trade volume, they have more demand, and thus drive price up. However, this could be a classic chicken or egg situation - is trade volume increasing because price is increasing?

Perhaps the signal can be identified with something a bit more subtle - something that is not on every trading chart. Using data from The TIE, figure 3 shows the ratio between the number of tweets about bitcoin versus the number of tweets about each altcoin per day. This suggests that alt season may be defined by the relative number of tweets about altcoins. This metric seems to not only capture the May and October swaps, but also potentially the short alt season in late February.

This signal mirrors the relative trade volume profile shown back in figure 2. Relative tweet volume is indeed highly correlated with relative trade volume, suggesting a general dynamic between these three variables - when people are tweeting about cryptocurrencies, they are also trading them, which causes their price to increase. The order of these influences is still somewhat ambiguous. However, there is no reason to believe that there is not positive feedback between all of these variables.

Our previous research has shown that listings tend to have a large impact on price. Using news collected through The TIE’s SigDev feed, we compared how the number of listings announcements per day could be used as a potential signal. There is a clear increase in the number of listings in late summer - the same period of time in which altcoins outperformed bitcoin the most significantly. Furthermore, spikes in this signal, like those seen in mid-May, may also be a signal for alt season. This makes sense - as altcoins become more widely available (listed on more exchanges) demand for these assets is increased.

Lastly, figure 5 shows the number of news articles being published directly from crypto projects (ex. The Ethereum Foundation publishing an article on Ethereum). Spikes in this signal seem to correlate with starts of alt seasons. This makes intuitive sense; these announcements usually pertain to big developments or partnerships for crypto companies - events that typically signify an increase in value. We can run some backtests with this idea to see if it is profitable.

Figure 6 shows the potential gains increase to nearly 5x with a simple trade strategy; if the number of news articles directly from crypto projects increases quickly (increase by an average of 15 per day with a 7-day moving average), swap into altcoins for two weeks. Otherwise, stay in bitcoin. These exact parameters (15 news articles/day increase, 7-day average, two week swap period) are just minutia. The finding that swapping holdings into altcoins when their news was high was generally more lucrative regardless of strategy specifics.

The timescales of these swaps are also quite interesting. In this example, there are only four swaps, each only two weeks long. This means that 85% of the time is still spent simply holding bitcoin. Still, this small time spent holding altcoins increases gains from 3-4x to over 6x.\ The specific timing of the swaps based on news is highly significant (p-value = 0.005).

This means if one were to randomly swap four times throughout 2020 from bitcoin to altcoin, they would only beat this strategy five out of 1,000 times - which is not likely. We simulated this to see how one would perform random swapping from bitcoins to altcoins; indeed, out of 10,000 trials of “random swapping,” only 50 happened to outperform this strategy (Figure 7).

2019 was a simpler time. Bitcoin pretty much always outperformed altcoins - quite a long bitcoin season. This raises an obvious yet important point:

Economic systems (especially crypto-economics) are highly nonlinear. This means that one set of relationships that work today might not work tomorrow. 2021 may be a whole new beast with even more complicated dynamics. Hopefully these insights spark some ideas on unique places to find indicators for bitcoin booms and busts.