Jump to content


  • Content Count

  • Joined

  • Last visited

  • Days Won


jbjnr last won the day on January 21

jbjnr had the most liked content!


About jbjnr

  • Rank

Recent Profile Visitors

The recent visitors block is disabled and is not being shown to other users.

  1. The payments from ripple wallets examined in this thread were mostly sent to exchanges where they are sold for $$$. The funds are sometimes sent back in the form of USD.bitstamp on the ledger, but mostly we lose track of them as they will go via the traditional banking system back to someone's account. Please note that some updated figures were included in a later thread - and there might be a another, but I don't fully remember. I will of course do a prediction of Q1 2019 soon.
  2. Dear Bob and other xrpchat readers, Your terms are very fine and like everyone else, I'm happy to accept them. Thank you very much for your contributions so far - I looked at your profile and it says you joined on Mar 5th and have "won the day" 9 times since then. Today is March 15th and by the end of the day, you'll probably have made it to 10 days in a row. I wonder how long you'll be able to keep it up Please try! At the end of 2017 I was attempting to write an article about the valuation/price of xrp based on the activities of market makers, and I abandoned it when the price went through the roof over xmas 2017, thinking "Who cares any more!". Since then, I've looked at the half written paper a few times and wondered if I should attempt to finish it. I have been busy with other work and not made progress on it. I was considering putting an invite out on xrp chat to find someone to read what I've written and tell me if it's total balls or not (that is quite probable since I have not studied economics, have no knowledge of market making or other financial activities). If it isn't nonsense then I'd want a co-author (or more than one) - preferably someone who understands financial markets to help me finish it. Having read some of what you've written in the new members' thread, it seems like you have probably seen numerous models of this kind - it is a laughably simple model, and it does have problems (for example speculators hoarding xrp would make it difficult for market makers to provide enough liquidity to prevent an exponential rise in xrp price that would inevitably result in a crash. Since the model is so simple, I assume it's rubbish, but I have not seen it considered elsewhere (though I have not been thorough in my literature searches of this kind of material - and that's why I need help. I just don't have the time to do it properly). The basic prediction is that there is an essentially linear relationship between xRapid payment volume and price (ignoring speculators etc). Your posts have contained a number of comments that have made me consider if now might be a good time to look again at the model and I wonder if anyone reading this who has experience of market-making/foreign exchange markets might be interested in taking a look at it and reviewing it with the intention of helping me improve it, or point out the obvious flaws so that I can abandon it. If anyone is interested in assisting, then please PM me - with the proviso that, (unlike Andre Geim), I do not wish to work with my Hamster (or other fictional authors!). I will be very busy for the next month or two and probably won't make anything available just yet as I'd like to spend some time reviewing it after a long break during which I've thought of a couple of new sections that should be added that are currently missing, but I'm posting this now as it might spark interest in someone who has expertise in this field. Thanks.
  3. The remains from February have been re-escrowed and it's a surprise. Only 600m has been put back under lock and key. We can't jump for joy that sales are going up just yet, as in the past they have done a second escrow a few days after the first, but this represents a change in pattern and it can't be a bad thing (for the last 6 months, 800m has bee re-escrowed). It looks like distribution is increasing - it might be a result of overall volume increases recently and daily sales going up correspondingly, but looking at the total transfers for Q1 so far on the daily sales, there is only 118m in my total column, so it does look like there are some private sales/distribution going through. (Recent stuff like listing on coinbase etc might have been responsible for some large purchases). Let's hope that this indicates an upswing in activity that is maintained long term. EDIT : 118m in $$$ for Q1 2019, but total transfers from wallets is 300m xrp and the wallets that do daily transfers to exchanges are up by about 50m xrp by the looks of things, so daily sales can account for approx 50m and the rest is somewhere else. I'll do the numbers a bit more thoroughly later.
  4. This seems to me to be a quite extraordinary statement to make as a tagline for a conference presentation. I have not heard anyone from ripple say anything quite as inflammatory as this in the past (in terms of xrp use cases). I think I see why they don't need Corey Johnson doing their PR strategy any more - they have entered a new phase in the push for adoption and the gloves are coming off. Why did this not make more of a splash in the news? Am I misreading it?
  5. Escrow returned from Jan 2019 is 800m. No significant change from the last few months and nothing interesting to report.
  6. As a quick follow up - I did a bit of googling and found this article https://medium.com/@danielcimring/downloading-historical-data-from-coinmarketcap-41a2b0111baf that contains some code to scrape historical data from coinmarketcap. I have updated my scripts to use this data source. If I replot the Q4-now sales from RD1,2,3,9 then I get a graph that looks like this (black line is coinmarketcap data, blue line is summed sales from ripple shifted by 1 day, red line is xrp price) The correlation between ripple sales and coinmarketcap volume data is very good. I still see a 1 day offset between exchange volume and ripple sales (implying that their sales for today are based on yesterday's volume), but now the normalization coefficient is 279.02 which corresponds to 0.358% of the market volume (ripple quoted 0.16% in their Q4 market report). So I'm now only a factor of 2.2 away from their data instead of a factor of 10x. The correlation coefficient is up to 0.91 for the Q4 data which is not bad. I will keep my eye out for other sources of data to close the gap.
  7. Can you elaborate on this a bit further please? Did you spend time looking at wallet transfers to come to this conclusion, or was there some other information that you found?
  8. Marvellous response. No real answer to that is there! Some people like to relax by watching sport on TV. I'm not one of them.
  9. I agree with all you've written. I played around looking for sources of data that I could use to try to correct the x10 difference between my volume and ripples assumed volume but decided to go ahead and use the current figures for now (otherwise the Q4 report would come out before I was finished). I wanted to try to guess first to see how close I can get. I suspect that the data ripple are using and what I need to look into is available from https://developers.ripple.com/data-api.html#get-external-markets the external markets API. I have tried a few simple queries, but not tried hard enough to do some proper data mining and finding the numbers I need. Ideally I want someone else to collate all the data for all XRP currency pairs and sum it up and then I can just work with that. It was only this weekend that I noticed that the 0.17% figure was mentioned in the Q3 report - previously I had only paid attention to the sales volume and not % of total. I thank @Hodor for bringing this to my attention in some unrelated post. I will investigate other sources of volume data as I'd like to get my numbers as exact as possible. I will of course post it here if and when I make improvements
  10. Programmatic sales look to be slightly higher, but not by a huge amount. OTC sales might be off the charts. I suspect that most of the 'missing xrp' is part of R3 settlement that was previously classified as undistributed and now distributed, but sadly we are not privy to this knowledge to we can only speculate. No doubt when the Q4 report comes out we'll have something new to learn. I'm not sure if founders' wallets are part of the undistributed category or not. To answer some more of these questions I think I need to use the bigquery dataset/database and do a bit more rigorous digging (and adding up), but I've not had time to look into that as yet. Too much work to do. It may be that some of the founders' sales are part of the missing xrp category. As noted by multiple people - not knowing which accounts truly belong to who makes it very difficult to be certain of anything. It does look as though Jed has either stopped his sales or is using different wallets than before. Someone else posted here about that - apologies I don remember who it was. I can't see any sales of his after Jan 11th. All these small changes are clues that things are happening behind the scenes I'm sure.
  11. What's interesting is not that they're selling more than they're making out, but they are distributing more than they are selling. I want to know where this is coming from and more importantly where it's going to. This is part of ripple's internal activities that are building the markets that will be needed for the IOV and the potentially revolutionary changes that are going to happen in the money markets if xRapid takes off as we all hope it will. My investigation is just my way of poking through the numbers to see what I can find that helps me understand the xrp ecosystem. Ripple would be foolish to reveal too much of what they are doing at the moment as they need to keep the competition guessing and not overtaking them in terms of technology or strategy. You might think that I am concerned or disturbed about ripple selling so much xrp, but if you read the post in my previous thread, I said (twice) that "ripple are looking after us". By selling tons of xrp they are creating the markets we need for global adoption. Recent threads on this forum have gone into very long winded rants about the way xRapid operates, about the way money is valued and about the way financial markets work. Ripple clearly understands all this very well and also that giving away xrp doesn't have the same effect as selling it. Individuals and Organisations purchasing xrp with skin in the game is what we need to build liquidity. The slow but steady increase in volume we're seeing and the increasing distribution are part of this long game. I'm just trying to find out how the game is evolving as it happens.
  12. No, I'm not of that opinion. Ripple don't have to publish any of their sales or distribution figures, but they choose to do so anyway, so I feel they're being very open. What I find intriguing is that despite the fact that we know how much they are selling and what we are looking for and that the ledger is entirely public, it's still remarkably hard to track down transactions of interest and account for movement of funds.
  13. Dear XRPChat, As the Ripple Q4Markets Report is imminent, I'd like to present an update to my previous analysis of ripple sales (in thread linked here) With a focus this time, on Q4 sales from 2018-10-01 to 2018-12-31. Part I The previous findings were that Ripple has been selling approximately 3% of daily volume and this trend appears to have continued without any significant change. I shall confine my plots to the period 2018-04-01 (2018-Q2) onwards, as this appears to be a time when a strategy shift occurred in the xrp sales and data appears consistent after this time. The up-to-date plot of daily sales (from the account previously referred to as RP2) to tagged (presumed to be exchange) accounts is and the correlation to daily volume is Which maintains the finding that around 1/35 = ~3% of daily volume is being sold off with the previous days volume (shift =1day) used as the reference. NB. Using the time period 2018-04-01 from cryptocompare.com to today as reference. I had previously hypothesized that ripple are selling xrp not only from their warchest of escrowed tokens, but also from founders/staff accounts since the declared sales of xrp in the Q2/Q3 reports do not tally with the summed totals I found in sales from RP2. This seems entirely reasonable and I decided to investigate this further. Two further pieces of information are useful. The first is that as pointed out in a recent comment by hodor - in the Q3 report, ripple state that their programmatic sales are 0.17% of market total (and total sales 0.43%). This is quite a significant difference from the 3% sales I have calculated. Either Ripple are using a different source of volume figures for their data, or I am not looking at the right numbers/wallets. Perhaps if I take only a subset of the 3% sales, they will match the numbers reported by ripple. In the previous analysis thread, I did not manage to find a good match between ripple's declared figures and the ones I presented - to try to improve this, I looked at the wallets that supply RP2 with the xrp that is sold daily. There are 9 of them, and I have labelled them as RD1, RD2, ... RD9 (for Ripple Daily 1-9). The payments from April until now are It's clear that they follow a similar pattern to the overall sales in the earlier plot since these are just the payments into the RP2 account that are then distributed to exchanges. Perhaps the 9 accounts represent different sources of xrp that correspond to warchest/founders/charity/other wallets that contribute to daily sales. This is a breakdown of the payments from the 9 into RP2 for recent quarters And let us remind ourselves of the ripple sales as published in the quarterly reports We are interested in 2018 Q2/Q3 programmatic sales and if any of the accounts RD1-9 match the figures quoted. The answer is "no", but perhaps several of the accounts together combined match the figures we are looking for. Summing RD1 RD2 RD3 RD9 gives a quite close result for Q2/Q3, here are the numbers The totals are 57.56 and 66.71 compared to 56.55 and 65.27. Those numbers are just 1.7% and 2.2% out for Q2 and Q3 respectively. Perhaps those are the accounts that are feeding the programmatic sales. If that is a correct assumption, then 2018-Q4 programmatic sales should be $83.47m. Allowing for +/-2.5% we would have a range of $81.4m to $85.6m. Actually, I suspect the number will be a little less because RD1,RD2,R3 look like good candidates, but why add RD9 in Q3 which had zero sales prior to then. Removing RD9 from the sum increases the error and reduces the total so I would not be surprised if the figure was a little lower. But I'll stick to $83.5m as a prediction for Q4 sales. One thing that troubles me is that the figures I get for programmatic sales using RD1,2,3,9 amounts to 1/57.33 = 0.1744 or 1.7% of the volume (which is about right, since we are only including half the accounts in our programmatic sales estimate that was ~3%). But in the Ripple Q3 report, they state that they sold only 0.17% of volume programmatically. And based on the figures presented in the report, it looks as though the volume data they are using is much higher than that sourced by myself, so I shall revisit these figures with better volume data. The average daily volume from their data is >$400M daily, which seems about right. It may be that the /v2/network/external_markets API can provide figures that improve the correlations with sales and give a better match to the %volume figures. Part II I said earlier that two further pieces of information were useful. One was the sales % numbers in the markets report, the second is that we do have extra data that might help us identify direct sales as well as programmatic ones. The extra data is the xrp distribution figures. Ripple provide data on how much xrp is in existence, how much is distributed/undistributed and escrowed. If we know how much xrp there is in April 2018 and we know how much is sold, escrowed and destroyed etc, then whatever remains in the difference between distributed and undistributed must be xrp that ripple has either loaned out or sold OTC. If it is undistributed, then it is part of the monthly escrow release but still sitting in a ripple wallet - if it is distributed, then it has been 'used' in some capacity. Lets have a look at the raw distribution data, here I've added 3 columns, 'remains' is just a sanity check to make sure that the totals are consistent (total-(escrowed+dist+undist)) and is zero or occasionally 1 due to roundoff errors. burned is the change in total each month and diff is the change in distributed. The other columns are as read from https://data.ripple.com/v2/network/xrp_distribution date distributed escrowed total undistributed remains burned diff 2018-01-07 39029058672 54000000024 99992855589 6963796893 0 NaN NaN 2018-01-14 39029001738 54000000024 99992777885 6963776123 0 77704.0 -56934.0 2018-01-21 39029011222 54000000024 99992725510 6963714263 1 52375.0 9484.0 2018-01-28 39032356092 54000000024 99992664799 6960308683 0 60711.0 3344870.0 2018-02-04 39094802192 53900000024 99992622540 6997820324 0 42259.0 62446100.0 etc etc ... when plotted the data for 2018 looks like the following graph. Note that we see drops in distributed XRP when it is placed in escrow (or potentially when large amounts are burned, but this is very small since the cap on transaction fees was introduced after a user lost 100k+ xrp in accidental fees). If we resample the distribution data to end of month totals, subtract the total distributed at the start of the period (giving zero initial distribution at the start of April 1st 2018, = start of Q2), then subtracting what we believe has been sold (using the figures we know from part I above) for each month, then the left over should be the amount that has been distributed by ripple, but not declared as part of programmatic sales. It should be 'direct sales' + xrp distribution from 'any other business' (sch as incentivisation of market makers and loans of xrp etc). The next graph shows the distributed xrp starting at zero on 1st April, with monthly and quarterly final amounts alongside. Note that as the distribution data is only published weekly, there can be big differences between the month end resampled and original data. For example, the data for the first week of Dec 2018 is 500m higher than the previous entry during the last week of Nov, and so the monthly Nov fig is much lower than the true data. Fortunately no big differences exist at quarterly boundaries so we don't need to make any adjustments to the data or our sampling. Previously we showed the quarterly sales from RD1,2,3,9 in $$$, the sales in terms of xrp tokens are so we should subtract those from the quarterly distribution numbers which are -------------------- Quarterly distribution -------------------- date distributed(since start Q2) 2018-06-30 146046852 2018-09-30 881583920 2018-12-31 1924354618 which gives Q2 65m, Q3 583.7m, Q4 1042.7m as the unaccounted for XRP for the 3 quarters we are interested in. Ripple have told us that in Q2 their direct sales were $16.87m - but we do not know what price they were sold at. The best we can do is use a flat rate based on the sales we do know about. In Q2 we found 81m xrp sold for $57.56m so we estimate that 23.7m xrp would have sold for $16.87m. This gives us (if we repeat the process for Q3) Q2 programmatic = 81m, direct = 23.7m, mystery remainder = 41.3m xrp Q3 programmatic = 151.8m, direct = 221.1m, mystery remainder = 583.7-221.1 = 362m xrp for Q4 we have Q4 programmatic = 196.3m, unaccounted for 866.5m of which some is going to be direct sales and some is unaccounted for. I had hoped that by conducting this little experiment, I would be able to recover the direct sales numbers from ripple, but unfortunately the numbers don't add up. There is still xrp being released that is unaccounted for. In Q2, it's 41m xrp, in Q3 it jumps to 362m and in Q4 we do not know the direct sales figures yet, but it could be anywhere from 0-866m xrp which would translate into a very large figure in $$$. Probably some of the xrp bound to the R3 settlement is included in this number, and probably the direct sales for Q4 will be very high (I'd guess over $100m). We will know soon and I will update my calculations when the Q4 report is released. Conclusion. TL;DR : My best guess for Q4 figures is between $81m and $86m for programmatic sales, over $100m for direct sales and a ton of xrp being loaned out, or distributed as part of other agreements. As usual, all my numbers are guesswork and as soon as I click send, I'll find loads of mistakes, please consider this analysis as a simple diversion from other mundane aspects of life. edit : deleted an image pasted by accident and fixed some typos
  14. tid amount book created_at maker_side pesos price 9742548 12821.933253 xrp_mxn 2019-01-10T07:43:46+0000 buy 8.667627e+04 6.76 I was also looking at this transaction last night, but ruled it out because the sell order on bitso is larger than the xrp transfer on the ledger. I would expect that the sell order would be the same as the transfer, or possibly slightly smaller (if they took a cut in xrp). The conversion to pesos for this transfer gives 86,676.27 which is slightly larger than the final recipient gets (if we assume that the recipient got 86,633.00 as indicated in the tweet), so the transfer to final bank might take the 43.27 pesos difference, but why sell more xrp than needed? 12821.933253 is an unusual amount to sell if you're matching a limit order. I can't find any other transactions that look like good candidates in recent days, so this looks like it must/might be it. I'll do a bit more checking tonight nd see if any more transactions look promising. edit: deleted comment that didn't make sense edit again: Actually, just noticed that that bitso transaction is marked as a buy order and not a sell.
  15. There's no need for ripple to 'repel' such an attack. The idea behind valor is fatuous and flawed. It relies on the principal of convincing a majority of xrp holders to swap their xrp for valor and thereby replace one token with another. Ripple has spent the last several years building a marketplace for xrp, incentivising market makers, bringing FIs on board and integrating xrp via xRapid into their tools. The reason xrp has any value is because it can be used as a payment token carrying its value across markets and there is simply no need or point in replacing it with a cloned version. No Fis would be interested in adopting it and ripple would not modify xRapid to work with the new token as it would have no advantage over the old one. No Fi would be interested in it for the same reason. A few delusional investors might be convinced to swap their tokens, but they would simply lose their investment and hold worthless tokens whilst the valor people would hold their xrp. It is a laughable idea. You are not wrong. Though I cannot imagine who would pay for such crap.
  • Create New...