Reward Pool Analysis (Why your Rewards are Dropping and When it May Turn Around)

in #steemit7 years ago

Gold Rushes Come and Go

The value of rewards on the Steem blockchain has been volatile since the beginning. In July of 2016 huge rewards were given out which sparked the interest of thousands in Steem. After Hard Fork 19, there was another spike in rewards as the increase in voting power and change in rewards pool management unleashed a virtual gold rush. Now that the rewards are decreasing, there is a rising concern that something is going wrong ... let's take a look:

Rewards Pool Balance and SP Payouts

The above chart was limited by the data I was able to pull mostly from @jesta's api (thanks!). I have tracked SP payouts by the blockchain back to the beginning of April when the rewards pool was still filling after Hard Fork 18. While I was not able to find actual rewards pool data from that period I will assume that the rewards pool was gradually filling to the point where my data picks up on 24 May. I've also plotted the daily active users but we'll get to that later.

Reward payouts are controlled by the size of the rewards pool and cumulative claims (votes) against the rewards pool over the past 30 days called 'recent_claims'. Remember, that the blockchain releases a fraction of the rewards pool each day depending on the volume and weight of votes. We can simply this to the following formula (this was outlined by @bilphil in this post).

reward_per_rshare = reward_balance / recent_claims

Back to the above chart. The month of May was fairly calm with slowly increasing SP rewards as the rewards pull filled and user growth was low. In June, user growth increased which increased the pull on the rewards pool and caused it to start draining (remember that recent claims looks back over a 30 day period so it reacts fairly slowly). This is how the system is designed and it worked correctly as it prevented existing users from having their rewards quickly diluted by the influx of new users 👍.

Then we come to Hard Fork 19 where votes are now 4x more powerful and the recent claims variable was manipulated to purposely partially drain the rewards pool (~1.6 million STEEM was a little overkill so I agree with this call). On the above chart it is easy to find where HF 19 hit (look for the SP Payout spike and change to the rewards pool rate of change, 20/21 May ). The second spike you see on the SP payout line is the 'gold rush' as users started madly posting to take advantage of the higher payouts (it occurs 7 - 8 days after the HF due to payout delay).

The declining rewards experienced by individual users from the gold rush forward were influenced by a combination of three factors (I'm ignoring the declining price of STEEM because that's covered in my Daily Investor Report: increasing recent_claims, decreasing rewards pool, and increasing daily users and posts to spread the rewards over. If you look at the above chart you can see the three factors converging (not actually a relevant fact but I'm a chart nerd so I thought was cool while I was writing this).

The natural next question goes something like:

Thanks for telling me why I'm getting less, but when do things get better?

Using Regression Analysis to Predict the Future

Before we dig into the we have to start with the universal rule of predictive charting:

Past performance does not guarantee future results.

I've gathered data on the rewards pool and recent_claims over the past 3 weeks and used that to produce a regression analysis to project the future path of both values. The projection shows that recent_claims will top out around 29 July (2 weeks from now). While I discovered that the rewards pool is declining is currently declining in a logarithmic decay I expect it to break this trend close to time that recent_claims tops out. A logarithmic decay would not stop going down which has not happened in the past and while I will keep an eye on it, I don't expect it to happen this time.

Therefore, I would expect rewards to be lowest around 28 July

Note: The above analysis is based on Steem Power payouts. The primary factor impacting total rewards is the price of STEEM and the SBD which is currently controlled by the greater Crypto trends.

Discord Chat:

You can reach me on Discord on my private server or through private message to LeongKhan


Sort:  

Humbling stuff. If minnows right now think it's hard to earn steem, just wait a few more years until the reward pool is even lower and there's 10x as many users to split it between...

We might be at the beginning of the transition away from Steemit as a major income stream opportunity, and towards the "micropayments / beer money" phase. Eventually, if there are gonna be millions upon millions of users, only a very scarce few will earn serious money from the blogging...

Content will need to be even better, more engaging, worth viewing on its own merits, as the shift takes place.

The great thing about the Steem blockchain is we can adjust the variables it runs on to compensate for growth. I think we will eventually increase the inflation rate to compensate for splitting between an increasing number of users.

IMHO, Steemit is at a critical juncture that will either see massive adoption or high attrition. The excitement that drives people to post, vote and comment is the financial reward (obviously). The community is great but the underlying drive has to remain or people will leave. As minnows discover they can't get more than a few cents per post without some great luck, they might drop out. Investors like myself looking to buy 10k steem start second guessing the long-term viability of the platform.

Bottom line is user growth drives increasing steem valuation. But the only way to get user growth is if new users have a chance to earn. We are near the tipping point of either greatness realized or the cliffs of doom.

Isn't it inevitable that, as the site has more and more users, the average payouts will be much lower?

Otherwise the site has to cap at a small size of creators... you can't have infinite creators and great payouts.

Perhaps the answer is in making the site work even when payouts are low.

Interesting read. If the prediction is correct, remains to be seen.
Especially for the longer term development. The number of active users is rising, and people are hoping for user numbers in the double or triple digit million range, like you see it on facebook or YouTube. But since the growth of Steem is limited to near 10% per year by the Steemit master plan, this will mean the reward pool has to be shared between those millions of users.
That would mean the single voting power value drops to next to nothing.
I wonder how this will play out in the future.

That is a possibility. If the total reward pool remains the same, then it may cause rise in Steem price. Steem to the moon!

There are two solutions to the scaling issue:

  1. The inflation rate could be reset to the original 100% per year.
  • Pro: Increased distribution and voting power of individuals
  • Con: Dilutes early adopters
  1. Rely on the increase in STEEM value that would come with mass adoption to compensate for the smaller STEEM rewards. The math helps show it:
  • $1 STEEM at 20,000 daily active users
  • x 10,000 for 20 million active users = $10,000 STEEM
  • While a new user's (15 SP) vote is worth $0.002 today, that would be $24.10 today, without vote dilution from increased voting
  • vote dilution would likely knock this back down to about the same level so a new user's vote would still be worth $0.002 but be a much smaller fractional value of actual STEEM
  • Pro: Rewards early adopters
  • Con: Hard for new users to gain SP without buying in at high prices

I'm sure there will be a discussion in 6 months to a year about how to solve this issue. I'm currently for option 1 but I will have to review the facts again when we get to that point.

I think both options are questionable at best. The first one I would even concider suicidal. While keeping up the fassade inside the community, it would destroy the faith - and the value of Steem - in the outside world. Doubling the amount of Steem every year will lead to a hyperinflation after some years. The old chess board dilemma, you know.
And hoping that Steem rises in value... Well, hope is always good. But its not much of a plan, really.

Luckily we don't have to hope for the price of STEEM to rise with growth of the user base. We have pretty good evidence of that.

https://steemit.com/cryptoreport/@leongkhan/steem-fundamental-analysis-crypto-report

I do not believe that crypto currencies are out of the woods yet. All crypto, not only Steem. Apart of some systemic problems inside the cryptos, there is a growing hostility of the central banks and other powerful entities in the financial sector towards cryptos. And they have the means to give cryptos a real hard time - if neccessary, by declaring it illegal.
So, I dont take a bright and happy future for cryptos for granted yet.

Or Steemit finally implements ads and start pushing different types of monetization ways and making money, which would naturally grow the value of the platform.

Very true, while I doubt Steemit itself will implement ads, several of our partner projects may and that could drive value into the entire blockchain.

Great piece of work Leong, so a combination of getting past Bip 148 and flattening decay. What pattern would you expect after these events ? Could they incrementally increase the inflation on a month by month basis rather that make a decision 6 months from now ?

I think it could actually be better to tie inflation to recent_claims which would essentially stabilize the rewards pool.

Good information, thanks!

nice info mate.

Very interesting :)

Upvoted and resteemed. Thank you for the knowledge :)

good post
plise followback

Nice analysis! Is this data based on Steem or USD...just wondering if the crypto recession impacts this data.

All the data is based on STEEM.

Gotcha...this is good stuff, will follow for more!

Nice to see you finally found time to publish your findings leongkhan :)

Thank you for putting such a huge effort into this. It is really insightful.

@ronni

Thanks! Too many projects ... too little time :-p

very interesting post and i hope the programmers will look into this from the early stage in order to balance the payment between new and existing users

My money is on the external factors too, so I'm betting on a mid-august lift.

Because stupid segwit fear. Mostly.

In general though, our stars and analysis of the internal factors totally align.

This post has been ranked within the top 50 most undervalued posts in the second half of Jul 15. We estimate that this post is undervalued by $14.07 as compared to a scenario in which every voter had an equal say.

See the full rankings and details in The Daily Tribune: Jul 15 - Part II. You can also read about some of our methodology, data analysis and technical details in our initial post.

If you are the author and would prefer not to receive these comments, simply reply "Stop" to this comment.

As payouts decline, do you think it will eventually lead to a decrease in the total number of posts? And will it create more of an incentive to focus on quality over quantity?

Yes, decreased payouts do tend to decrease posting. I've not really seen that it tends to increase quality though, some authors hold their best content for when rewards are higher.

No one is in business just to spin their wheels. Money talks and BS walks. The post rate will go down.

The big push to get users in here, at least from YouTube to DTube centers around freedom of speech. Same with FaceBook. However, the incentive of getting paid for your posts is also pushed, quite openly. I would say earning for your content tilts the scales lower the other two.

Most people come over here to "get their feet wet", so to speak and I think after the infatuation phase goes away and they see little to no ROI for their efforts, they simply abandon ship back to there their friends are. If one is little better than the other, and the other has your friends, where would you go?

Freedom of speech is great, but with no audience? You might as well not speak at all.

That said, minnows tend to find other minnows, so the future is not too bad. It all depends on how popular you get I suppose.

Thanks for your hard work and analysis. Followed for more!

Hello! I just upvoted you! I help new Steemit members! Upvote this comment and follow me! i will upvote your future posts! To any other visitor, upvote this post also to receive free UpVotes from me! Happy SteemIt!

Thanks for posting this. Its very informative.

That first chart looks distressingly like one of our (US) various insolvent pension funds (maybe Illinois), draining off a large capital stockpile by more than earnings every year.

Could it be that a large part of this rewards pool drain is coming from the whales who were involved in the no-voting project before the HF, but are voting now, increasing the "cumulative claims"? Thus, resulting in more total vests claims in the reward pool per day, necessitating diving into the capital "savings" of the pool?

"Therefore, I would expect rewards to be lowest around 28 July"

Thanks for the concrete prediction. Looks like solid analysis. Will be good when they start trending back up.

I'm sure the no-whale-vote experiment was a factor in because it added to the impact of the increased voting power. However, that is already accounted for and should still level out in the next 2 weeks.