When we hold crypto assets, "the team is doing things" is the confidence that "the price of the currency will take off in the bull market" and the bottom line of "continue to hold when trapped in the bear market".

But will "the team is working hard" really make the price of the currency rise more in a bull market? Will it be more resistant to declines in a bear market?

This article uses 10 years of historical data to tell you the answer.

The four bull and bear cycles of the Crypto market

Bitcoin's Genesis block was born in 2009. Its price has experienced multiple bull and bear cycles in the following 14 years, and industry narratives such as "ICO era", "public chain explosion", "Defi Summer" and "NFT wave" have emerged one after another.

For the convenience of analysis, this article defines 2015.07-2018.01 as the first bull market, 2018.01-2020.03 as the first bear market, 2020.03-2021.05 as the second bull market, and 2021.05 to present as the second bear market.

The first round of the ICO bull market from July 2015 to January 2018 was a long time ago, and there is too little data available to obtain rigorous results. Therefore, this article focuses on analyzing the last three cycles.

What factors can reflect "team work"? We found six factors

The vast majority of projects in the industry are based on blockchain technology, and the code is open source on Github (GitHub is a platform for code hosting and sharing).

Therefore, Falcon uses GitHub's six factors as quantitative standards to measure "what the team is doing", including: Star, Fork, Commit, Issues, Pull requests, and Watchers. The following are the specific meanings and types of the six factors:

Detailed introduction to the six factors of project GitHub data

The Github data of all the projects in this article can also be seen on Falcon's products. Visit the link.

Effective sample size and term explanation

The team counted the price trends of coins in three market cycles and their corresponding project GitHub six-factor data. After outlier processing, 81, 330, and 596 valid token samples were retained in the three market cycles respectively.

The following table will explain the terms:

In the first round of bear market (2018.1-2020.3), GitHub data had a certain anti-fall effect on the price of coins, but the effect was limited, which may be related to the small sample size.

Let’s start with the first bear market:

Descriptive statistics of the six factors of GitHub data and the rise and fall of coin prices:

The token data of the first round of bear market is relatively scattered, which is consistent with the characteristics of the early rise of the crypto market. The standard deviation values ​​of the seven statistics during this period are far from the average, indicating that the previous prices and GitHub data of different currencies are quite different. The GitHub attention of more mature tokens such as Bitcoin and ETH at this stage is extremely high, but the attention and developer contribution of many emerging currencies on GitHub are relatively low.

Statistics of the prices of coins whose price drops in this interval are less than the average drop (in bold black) and their corresponding six factors of GitHub data:

The gray grid represents the tokens that are opposite to the market trend. We believe that the nature of such tokens is relatively special and needs to be comprehensively analyzed in combination with the market situation. There is only one binance-exchange in this interval. Looking at its six factors of GitHub data, the star and fork values ​​are in the top 10 of the statistics, but commit, issues, pull_requests, and watchers are extremely low. This is mainly because the bnb token only had the attribute of "platform currency" before 2019, and no "public chain" attribute, so the code is not open source. In the second half of 2018, the market focus was on the platform currency sector, and bnb had a high increase and was resistant to declines in this cycle. For this coin, only the star and fork factors of the six factors of GitHub data have a certain correlation with price.

Among the tokens whose price drops are less than the average, 40% of the tokens have a GitHub factor in the top 10 of the statistics, while the remaining tokens have a generally lower GitHub situation. It is preliminarily inferred that during this period, the GitHub factor has a certain positive effect on reducing the price drop, but the effect will not be particularly large.

The second round of bull market (2020.3-2021.5) Github projects with more activity rose more in the bull market

Descriptive statistics of the six factors of GitHub data and the rise and fall of coin prices:

The token data of the second round of bull market is relatively concentrated, and the maturity and prosperity of the crypto market have improved. * The standard deviation statistics of the 7 statistics in this interval are close to the average value. Compared with the statistics from 2018 to 2020, the sample data in this interval is more concentrated. Combined with the actual market situation, on the one hand, the token market has developed more maturely in 2020, and the tokens that emerged in 2018 have achieved certain development in this interval, and their corresponding fundamental GitHub data have generally increased significantly. On the other hand, with the development of the market, the number of tokens issued in this interval has increased significantly, and with the increase in the number of reference samples, the concentration of data distribution has also increased further.

Statistics of the prices of coins whose price increase in this interval exceeds the average increase (black bold) and their corresponding six factors of GitHub data:

Among the 330 data, 11 of them have a price increase that exceeds the average, and 5 of them have six factors of GitHub data that exceed the average, accounting for about 45%. It is preliminarily inferred that the increase in GitHub data has a certain correlation with the increase in coin prices , and the specific correlation is analyzed in the third part of the article.

The projects that did not rise but fell during the bull market are all those with very inactive Github development.

Abnormal currency price (coin price falls in a bull market):

Among the 330 valid samples in this period, the prices of 28 tokens fell against the trend, reflecting that these 28 tokens are very weak. At the same time, 90% of the GitHub data corresponding to these tokens are below the average and are generally close to the minimum.

The third bear market (May 2021 to present) GitHub's more active projects have made a certain contribution to resisting the bear market, but their role is still not very large

Descriptive statistics of the six factors of GitHub data and the rise and fall of coin prices:

The top 20 tokens and their other 6 statistics are sorted by star factor (the tokens in bold black are those that exceed the average value):

As the crypto market further develops, the token data in the second round of bear market is more scattered, which is speculated to be related to the further differentiation of industry gaps. * The standard deviation values ​​of the 7 statistics in this interval are quite different from the average value, indicating that the token data in the second bear market stage is more scattered. In 2021, the token market is still in a period of vigorous development. More and more people are pouring into the token market. People first target the token projects with better development and more mature in the market. The corresponding GitHub attention of such tokens is as high as tens of thousands of times. However, for the emerging tokens in this period, it still takes time for the public to be familiar with them, and the attention and development level are naturally much lower.

Combined with the statistics of the top 20 tokens in the star data, it is found that the tokens whose six factors rank above the average in GitHub data have certain similarities in statistical rules, and it is inferred that there is a high correlation between the six factors. At the same time, it is found that the tokens with particularly high rankings in the six factors of GitHub data are all relatively mature tokens, and their issuance period is basically between 2015 and 2018, such as bitcoin, ETH, and dogecoin.

Abnormal currency prices (coin prices rise in a bear market):

There are 28 anomalies in the data of 596 tokens, among which 6 tokens have more than one factor of GitHub data exceeding the average, accounting for 28%. According to the table, it can be inferred that the increase of GitHub data has a certain contribution to the bear market resistance, but its role will not be particularly large. Such a strong price advantage of such a currency is mainly determined by factors of other categories.

How do we quantify the correlation between the GitHub factor and price? Which coefficient should we use to make the judgment?

In the above, through simple statistical analysis, we found that Github data plays different roles in bull and bear cycles.

So how do we quantify the correlation between the Github factor and price?

The QQ graph uses the sample quantile as the horizontal coordinate and the corresponding quantile calculated according to the normal distribution as the vertical coordinate, and represents the sample as a scattered point in a rectangular coordinate system. If the data set follows a normal distribution, the sample points will be a straight line around the diagonal of the first quadrant. It is more reasonable to use the Pearson correlation coefficient to analyze the data set that follows the normal distribution, and it is more reasonable to use the Spearman correlation coefficient to analyze the data set that does not follow the normal distribution.

The results of the six-factor QQ plots for the three intervals are as follows:

As can be seen from the table, the sample points of the six factors Star, Fork, Commit, Issues, Pull_requests, and Watchers in the three intervals are not distributed around the diagonal of the first interval, that is, they do not obey the normal distribution. The correlation analysis of the six factors and the token price will be judged based on the results of the Spearman coefficient.

The first bear market (2018.1-2020.3): Due to the sample size, the correlation between the GitHub factor and the coin price is limited

Correlation table between six factors and currency price increase:

The five factors of GitHub data have a positive effect on the resistance of the coin price to a bear market. It is easy to see from the table that the correlation coefficients of star, fork, issues, pull_requests, watchers and price are all around 0.260, and all show a significance level of 0.05, which statistically indicates that the five factors are positively correlated with the coin price.

In this interval, the commit factor has no significant relationship with the price increase. The correlation coefficient between commit and the price increase or decrease is -0.032, close to 0, and the P value is 0.776>0.05, indicating that commit has no correlation with price.

The correlation results of star, fork, issues, pull_requests, watchers and price are consistent with our previous judgment, that is, there is a certain positive effect. We know that the correlation will not be too high, but the correlation of 0.260 is meaningful for our subsequent research on the trend of token prices and the construction of related factor strategies. The result of commit is slightly inconsistent with the previous article. We initially concluded that it was due to limited sample data. In the second and third intervals, we collected more token data and will further examine the correlation between commit and price.

The second bull market (March 2020-May 2021): The more active GitHub is, the more the price of the coin will rise

Correlation table between six factors and currency price increase:

In the second round of bull market, as the validity sample increased from 81 to 330 , the correlation between the six factors of star, fork, commit, issues, pull_requests, and watchers and price increased significantly, and the correlation was around 0.322, significantly higher than the correlation mean of 0.260 in the first interval, and at the significance level of 0.01. Among them, the correlation between star, commit, and watchers and price was as high as 0.350. All six factors in this interval were positively correlated with price, which seemed to confirm our speculation that commit was negatively correlated with price in the first interval, that is, the sample data was not large enough and was affected by individual extreme values.

The GitHub factor is time-sensitive in the second bear market (May 2021 to present)! It is still significantly correlated with the price of coins in the bear market, but it is not necessarily resistant to declines

Correlation table between six factors and currency price increase:

For the third interval, the number of valid samples increased to 597. Compared with the first interval, the correlation between the six factors of star, fork, commit, issues, pull_requests, and watchers and price increased. Under the significance level of 0.01, the mean correlation was 0.216, slightly higher than 0.205 in the first bear market, but significantly weaker than the correlation of 0.322 in the second interval.

We believe that the six factors of GitHub data are positively correlated with the increase in coin prices, but they have a certain timeliness!

That is, the six factors have stronger predictive power and contribution to the rise and fall of coin prices in a bull market, but their effectiveness is relatively weak in a bear market. The coin prices in a bear market are more affected by other major factors (such as volume-price factors, market sentiment and other alternative factors). GitHub data is only part of the fundamentals and plays a relatively limited role.

Conclusion

Through the above content, Falcon summarizes the conclusions of this article:

1. With the development of the Crypto market and the prosperity of the industry developer ecosystem, Github data and currency prices are increasingly showing a strong correlation.

2. From an investment perspective, you should invest in projects that are actively developed on Github and avoid projects that are not actively developed on Github.

3. In a bull market, the more active the project on Github, the higher the increase; in a bear market, the more active the project on Github, the more resistant it is to decline.

4. The correlation between Github and coin price is significantly higher in the bull market than in the bear market.