Does the Hive Social Network Have any Impact on the Price of Hive?

Yesterday I wrote a post using Granger Causality Testing to examine if there is potentially a causal link between the price of Hive and the user activity on Splinterlands. The reason I had started thinking about Granger Causality Testing was after this exchange with @tobetada

I had a similar discussion with @gavvet many years ago. The fact that price drives user activity is clear as day, but as someone who obsesses over the statistics I remember in the early days of Steem there were times that user activity was growing in the absence of a price rise, only for a price rise to follow. After all, just because price can feed into user activity by fueling the reward pool, doesn't mean user activity can't also drive price by fueling speculation.

Let's examine the connection between the price of Hive and the number of active social media authors on Hive. This post will use the same methodology as yesterday, and if you wish to learn more about applying Granger Causality Testing, the information at Real-Statistics.com was invaluable for me.

Hive Authors vs Hive Price

Here are the first datasets we will consider, in the form of line charts.

Just as yesterday, these datasets are not stationary, so we instead look at the daily changes in these metrics.

Applying Augmented Dickey-Fuller Testing we confirm that these datasets are stationary.

Finally, we run the Granger Causality Tests.

Huh 🤔

All statistically significant findings are highlighted. The ones highlighted green are the strongest effects, and those not highlighted can be ignored as not statistically significant.

After 1 day, a change in active authors has already had an effect on the price. Then for the next few days it has no clear effect, until the second two weeks when an effect can be seen.

On the other hand, a change in the price has no statistically significant impact on user activity after the first day. After 2 days and over the first week, the effect of price on user activity grows sharply, peaking after 6 days. It continues to have a strong effect on user activity for upwards of a month at least.

Overall the effect of price on user activity is far stronger and goes on for longer than the effect of user activity on price, but both effects exist.

Steem Authors vs Steem Price

Let's also look at the pre-fork days of Steem for comparison.

Charts:

Daily Changes:

Augmented Dickey-Fuller Tests:

Granger Causality Tests:

The effect of price on user activity was even stronger in the days of Steem. A very strong immediate effect, which would continue to exist and drop off for well over a month.

Interestingly the effect of user activity on price, while similarly not as strong, continued for far longer during the days of Steem than it does in the period of Hive.

Conclusions

Changes in the price of Hive has a strong effect on user activity on Hive, especially from 4 to 14 days after a change in price. This effect was much stronger in the days of Steem.

Changes in user activity on Hive have an immediate effect on the price after 1 day, but not as strong as the reverse effect. After about a week, a change in user activity has a further effect for another 2 weeks. This effect also existed in the days of Steem, but continued for longer and was a stronger effect overall.

Compared to Splinterlands, changes in social media activity on Hive has a more immediate effect on the price of Hive, but in the long run Splinterlands has a larger impact overall.

---

My statistics and analysis posts take many hours each to research, chart and write, so if you find them valuable and of interest to other Hivers, I appreciate your support in sharing, commenting, and/or upvoting my work. If you're interested in these kinds of stats posts, click the 'follow' button on my profile, or subscribe to the Hive Statistics Community which features daily Hive stats posts from @arcange as well as less regular posts from myself and others.