Exploratory Analysis – How does the Price of SBD impact Power Ups and Steem Power Investment?

The price of SBD has been above $1 for some time now and despite efforts from some to peg SBD to the US$, it seems to be more following the price of STEEM.

That being said, while the price of SBD is greater than $1, many people opt to take post payout 50% 50%. By doing this and using the SBD to purchase STEEM and then power up, work out more profitable than taking payment 100% STEEM Power.

Based on this I carried out an exploratory analysis. Starting with one question, each answer lead me to another question. And by the end, I had proved some thoughts I had and also found out some very interesting information.

As so not to distract from the findings, details of how the analysis was preformed including the query’s is at the bottom of this article. Below you will find each question I asked of the data and what the data returned

A. What % of rewards claimed were made up of Vests, SBD and STEEM?

B. Can I see this over time please?

C. . Wow that’s a dramatic drop in the % of rewards taken in Vests at the end of December. What was going on with the price of STEEM and SBD at the time?

D. So the price of SBD went up and at the same time the % of rewards collected in SBD also increased dramatically. As the price of SBD began to fall, the % of rewards collected in SBD also fell. This indicates as expected, that there is a correlation between the price of SBD and the % rewards claimed in SBD. Can we see this correlation?

E. Almost 70% of rewards claimed were in SBD and there is a correlation between the price of SBD and the % rewards claimed in SBD which can be seen clearly by this massive curve. Is it possible there is more SBD leaving Steemit than being powered up?

F. From the table it seems that there is less % of the SBD claimed powered up when the price of SBD is high however we can also see that there is always more vests Powered up than claimed in SBD. From this it appears there is not more SBD leaving Steemit than being Powered up. Can we see a negative correlation chart to prove there is a drop in Power UP as % of SBD claimed?

G. The curve shows negative correlation between the price of SBD and the Vests Powered up as a % of SBD claimed. There is a cluster of power up when SBD was below $1 where up to 2000% of the SP earned was powered up. This then tails off as the price of SBD rises.

But if the vests powered up is always greater than the SBD claimed, how would this data look if we compared power ups to total vests claimed?

H. Very surprised with the above table. As the price of SBD rises the Power up as a % of Total vests claimed also increases. Can we see a correlation chart for this?

I. there is a cluster when SBD was just under $1 where the Vests powered up as a % of total vest claimed ranges up to 100% however in general we can see a positive correlation between the increase in the price of SBD and the total Vest claimed. What about the total vests powerd up, does this have a correlation with the price of SBD?

There is a cluster on the right with vests powered up of under $2 and less than 1bn vests. There is also a cluster of under 5bn as the price of SBD rises. But there is no correlation between the total vests powered up and the price of SBD.

Conclusion

If there is a positive correlation between the price of SBD and the Vests Powered up as a % of total Claimed vests, however we have established that there is a negative correlation between the price of SBD and the Vests Powered up as a % of SBD claimed, what can we establish about additional investment in Steemit while the prices of SBD go up?

The table above looks at both May 17 and Jan 18 as the power up values are similar. However there has been a 332% increase in the price of SBD during this period. There has also been an increase in the Total Vests Claimed of 215% and a massive increase in the SBD vests claimed of 4231%. But no increase in powerups.

With a total increase of 215% in vests claimed, if the level of investment on steemit remained the same we would also see an increase in the power up of the same %. This would suggest to me that there is 215% more vests leaving steemit in Jan 18 over March 17 and would therefore indicate that people are less likely to invest in Steem Power (and long term position) on Steemit when the price of SBD is high. I think both Witnesses and Whales need to take note of this data and if they don't like the trend, its really time to act. With the still uneven power distribution on Steemit, action by these groups have a major impact on the values above and how people will react to investing in the platform.

I think it would be interested to look at this data against the price of STEEM. I know that the price of SBD and STEEM have been following each other, but for a long time SBD was $1 or less and so it was more profitable to take payment 100% Steem Power. I wonder if the price of SBD was pegged to $1 and STEEM was higher would there be a better incentive to keep your investment in Steem Power?. The 13 weeks power down does meant that you have to stay for a while, but factoring this out, I wonder how this would play out. Maybe my next exploration analysis will look at this.

What stood out to your from this data? What further questions does it lead to? Does this give reason to be more positive about pegging SBD to $1? Please do comment below and let me know.

The Data

Data was taken from the Steemsql database using the following queries

Select * From TxClaimRewardBalances (NOLOCK)

Rewards can be paid out in STEEM, Vests and SBD. As these are all different units, it was important for the analysis that all units be converted to 1. For this analysis I have taken VESTS to be the main unit of measurement.

In order to convert the STEEM and SBD to Vests, required additional data. I was unable to find historic data on the VESTS to STEEM conversation so I had to calculate my own. To do this I used the following query

Select * From VOFillVestingWithdraws (NOLOCK)

This query returns the VESTS withdrawn and the STEEM received. This data can then be used to calculate an average rate for VEST to STEEM over time and on given dates.

This VESTS to STEEM rate can then be used to convert any STEEM values to VESTS.
I also need to convert SBD to VESTS. This was a little more complicated. First I needed historic prices of STEEM and SBD. Then using these prices, I could get the STEEM to SBD ratio. All SBD can now be converted to STEEM, and then using the STEEM to vest above we can also convert the SBD to vests.

To gather data on historic SBD and STEEM pricing I connected to coinmaketcap using power BI.
The M queries used were

let
    Source = Web.Page(Web.Contents("https://coinmarketcap.com/currencies/steem/historical-data/?start=20130428&end=20180423")),
    Data0 = Source{0}[Data],
    #"Changed Type" = Table.TransformColumnTypes(Data0,{{"Date", type date}, {"Open", type number}, {"High", type number}, {"Low", type number}, {"Close", type number}, {"Volume", type number}, {"Market Cap", type number}})
in
    #"Changed Type"

And

let
    Source = Web.Page(Web.Contents("https://coinmarketcap.com/currencies/steem-dollars/historical-data/?start=20130428&end=20180423")),
    Data0 = Source{0}[Data],
    #"Changed Type" = Table.TransformColumnTypes(Data0,{{"Date", type date}, {"Open", type number}, {"High", type number}, {"Low", type number}, {"Close", type number}, {"Volume", type number}, {"Market Cap", type number}}),
    #"Renamed Columns" = Table.RenameColumns(#"Changed Type",{{"Close", " SBD Close"}})
in
    #"Renamed Columns"

As everything is now in VESTS, these values were used to compute the % calculations shown in the analysis above. To calculate the power ups I needed some new data.

The sql query used was

Select *
From TxTransfers (NOLOCK)
WHERE [type] in ('transfer_to_vesting')

However I noted from this data that there were some transfers of vests made from one account to another that did not equate to Powerup. To filter these out I ensure that the transfer sender was the same as the recipient.

---

Posted on Utopian.io - Rewarding Open Source Contributors

---