Jump to content
opaopa

Does xRapid effect volume? Analysis.

Recommended Posts

20 minutes ago, opaopa said:

Conclusion
We cannot refute our hypothesis. With such a short period since release, weak assumptions and a methodology that is questionable at best, it is not certain that the indications from this analysis has anything to do with xRapid at all. But one can hope. :)

Solid conclusion. Thanks for the extra effort. Really apreaciate this over all the hype and hysteria. ☺️

Share this post


Link to post
Share on other sites
1 hour ago, opaopa said:

With the free availability of data on open blockchains there is a surprising lack of analysis on xRapid adoption. From reading on this forum I have no clear answer whether adoption is leading to any volume effects. I hope to reach an indication of the answer to that question.


Hypothesis
With adoption of xRapid the volume discrepancy between working days (mon-fri) and weekends (sat-sun) will increase for XRP.
The hypothesis rests on the assumption that the behavior of businesses adopting xRapid will follow a pattern which will increase trading volume during weekdays  when the businesses are active, but leave a smaller effect on weekends when the business are closed. This is a weak assumption which should be considered when interpreting the results.


Methodology
Data sets have been downloaded from Coinmarketcap and consists of daily observations of  volume. The data is categorized as occurring on either a weekday or a weekend, category 1 and 0.
For each week, category 1 volume is compared to category 0 volume. Category 0 is normalized by multiplying the volume with 2,5 (5 day workweek = 2,5*2 day weekend).
This means that each weekend is compared to the previous 5 workdays. A number is calculated to show how much bigger is workweek volume compared to weekend volume. If the 7 days all would have the same volume, the number would be 1.


Results

image.png.721983c4e026238c0707dc77b3dc98ef.png
 
xRapid announcement is represented with the red dot in the chart. We can clearly see that workweek-to-weekend volume has increased since release and that last week in fact was a record discrepancy for the year. However, last week (Monday) was also the USDT “crash” which contributed to huge volumes for cryptocurrencies in general. Therefore, we would like to control for such events and isolate the xRapid effect if possible.


Control – BTC
Methodology
The same datasets as for XRP is downloaded for BTC. Each weeks workweek/weekend volume for XRP is divided with the corresponding workweek/weekend volume for BTC. Through this manner we can negate the effect of market wide events such as the USDT-scare (resting on the assumption that the USDT-scare affected XRP and BTC volume equally, which again is a weak assumption)


Results
image.png.4fb4688060a4eee4a2ba61d0a2ed00b9.png 
We observe that the variance week-to-week shrinks after the controls. Last week is still a big break away from the trend over the last 6 months, however it is not record-breaking for the year.

Conclusion
We cannot refute our hypothesis. With such a short period since release, weak assumptions and a methodology that is questionable at best, it is not certain that the indications from this analysis has anything to do with xRapid at all. But one can hope. :)

Need more time rerun in a month. Great job btw!!

Edited by pumpndump

Share this post


Link to post
Share on other sites

Thank you. Sharing our analyses like yours we can mutually benefit ourselves by learning more from each other.

I think lack of analysis on xRapid adoption is still mostly caused by lack of publicly available information. Yes, we have CMC but their published data that is an aggregate from several crypto exchanges. At the moment xRapid is still utilized in production usage in Bittrex and Bitso as far as I'm concerned. Thus, I think we should get the trade data from Bittrex and Bitso alone in order to analyze xRapid adoption properly. If we could get that data then we could be able to create methods how to pair transactions (USD/XRP & MXN/XRP) by their value between those aforementioned crypto exchanges.

I really like the proposed hypothesis but I think it might be premature because many early xRapid remittances are initiated by foreign migrant workers who send portion of their paycheck back to their loved ones. I'm afraid that many of those migrant workers might send their money back to home during the weekends when they go shopping and have more leisure time than during weekdays when they are working.

Share this post


Link to post
Share on other sites
Guest
2 hours ago, opaopa said:

With the free availability of data on open blockchains there is a surprising lack of analysis on xRapid adoption. From reading on this forum I have no clear answer whether adoption is leading to any volume effects. I hope to reach an indication of the answer to that question.


Hypothesis
With adoption of xRapid the volume discrepancy between working days (mon-fri) and weekends (sat-sun) will increase for XRP.
The hypothesis rests on the assumption that the behavior of businesses adopting xRapid will follow a pattern which will increase trading volume during weekdays  when the businesses are active, but leave a smaller effect on weekends when the business are closed. This is a weak assumption which should be considered when interpreting the results.


Methodology
Data sets have been downloaded from Coinmarketcap and consists of daily observations of  volume. The data is categorized as occurring on either a weekday or a weekend, category 1 and 0.
For each week, category 1 volume is compared to category 0 volume. Category 0 is normalized by multiplying the volume with 2,5 (5 day workweek = 2,5*2 day weekend).
This means that each weekend is compared to the previous 5 workdays. A number is calculated to show how much bigger is workweek volume compared to weekend volume. If the 7 days all would have the same volume, the number would be 1.


Results

image.png.721983c4e026238c0707dc77b3dc98ef.png
 
xRapid announcement is represented with the red dot in the chart. We can clearly see that workweek-to-weekend volume has increased since release and that last week in fact was a record discrepancy for the year. However, last week (Monday) was also the USDT “crash” which contributed to huge volumes for cryptocurrencies in general. Therefore, we would like to control for such events and isolate the xRapid effect if possible.


Control – BTC
Methodology
The same datasets as for XRP is downloaded for BTC. Each weeks workweek/weekend volume for XRP is divided with the corresponding workweek/weekend volume for BTC. Through this manner we can negate the effect of market wide events such as the USDT-scare (resting on the assumption that the USDT-scare affected XRP and BTC volume equally, which again is a weak assumption)


Results
image.png.4fb4688060a4eee4a2ba61d0a2ed00b9.png 
We observe that the variance week-to-week shrinks after the controls. Last week is still a big break away from the trend over the last 6 months, however it is not record-breaking for the year.

Conclusion
We cannot refute our hypothesis. With such a short period since release, weak assumptions and a methodology that is questionable at best, it is not certain that the indications from this analysis has anything to do with xRapid at all. But one can hope. :)

Just a suggestion if you are looking at correlations and significance. You could start by stating a null hypothesis and then using a standard statistical test to scientifically confirm the significance of any correlation. 

Good work btw.

it would worry me if xrapid showed no significant change in volume. It would mean as a community we have probably misunderstood something fundamentally.

Share this post


Link to post
Share on other sites

Nice thread - this is a great example of why I enjoy xrpchat so much - people like yourself, who put some thought and energy into answering the same questions that many of us have. I'll continue to wade through all the bs, fud, and fomo posts on these boards for precisely this reason. Love it - Thx!

Share this post


Link to post
Share on other sites

Thanks in general for the kind words and thank you for the criticism. This is a half assed approach to a scientific method and that's basically the level of complexity and the amount of work I'm comfortable with. If anyone want to build onwards it would be interesting.

Will be sure to update this post anyway in the future.

Share this post


Link to post
Share on other sites

Excellent idea and work.

There was also an interesting idea that the exchange volume delta should be twice the XRPLedger delta if XRapid is a significant component of volume.  This is because a xrpaid transaction involves a buy and a sell on an exchange with a movement via XRPLedger.  So for each transaction twice the XRPLedger volume will be on-exchange volume.

Edited by Tinyaccount
Added a clarifying word

Share this post


Link to post
Share on other sites

×