Jump to content
Guest

MGI, ODL, Bitso and a theory about the price

Recommended Posts

9 hours ago, Frans said:

Hi @mouradski,

I was looking through the output you posted to see if EURUSD was 24/7 and I noticed this:

"dateAsString": "2020-05-22T11:52:28Z"
"dateAsString": "2020-05-23T12:01:28Z"
"dateAsString": "2020-05-23T12:11:29Z"

The date changes from may 22 to may 23, so I would expect the time to be 00:01:28 instead of 12:01:28

I looked if 12 would turn into 13, but found this:

"dateAsString": "2020-05-23T12:53:27Z"
"dateAsString": "2020-05-23T01:03:29Z"

Not sure if this could cause any problems with sorting as you probably use the timestamps for that, but might as well mention it.

 

TransactionID's:

"transactionHash": "0FFCHAIN_1590191548000_1590191537000",
"id": 143707
"dateAsString": "2020-05-22T11:52:28Z"

"transactionHash": "0FFCHAIN_1590192088000_1590192077000",
"id": 143708,
"dateAsString": "2020-05-23T12:01:28Z"

"transactionHash": "0FFCHAIN_1590192689000_1590192678000",
"id": 143709,
"dateAsString": "2020-05-23T12:11:29Z"


"transactionHash": "0FFCHAIN_1590195207000_1590195196000",
"id": 143715,
"dateAsString": "2020-05-23T12:53:27Z"

"transactionHash": "0FFCHAIN_1590195809000_1590195798000",
"id": 143716,
"dateAsString": "2020-05-23T01:03:29Z"

Regards, Frans
 

dateAsString field is not used but we need to fix the formatter, timestamp is used for processing 

Share this post


Link to post
Share on other sites
Posted (edited)
On 5/24/2020 at 3:27 AM, mouradski said:

monday I'm sure we'll got a new ATH

We only just fell short. Only logical, the 4 biggest corridors started trading at 7 am (Berlin/Amsterdam time) and the cut off point was 11 minutes ago, 2 am, 19 hours later, so we miss 5 hours, I forgot all about that  :D

Tomorrow we will break it.

EUR/USD alerady at 4.4 million, that's higher than I expected tbh.

Edited by Frans

Share this post


Link to post
Share on other sites
On 4/3/2020 at 9:59 PM, doopers said:

The spikes seem to always be caused by the same entity. Unlike moneygram, which is a steady drip throughout the day, there's another company that seems to send quick bursts $20k at a time. In the past for $300k before cooling down. This week they've been sending batches of $500k (as in, 25 transactions @ $20k each).

Destination tag is 47175592

happened again today, seen 2 bursts, the last less than an hour ago. Destination tag: 47175592

Since I think the $39.9k (prb $40k -/- some costs/fee) are Moneygram's and end up on tag: 25370527, it might be another client

here's one txid: 4F5028426B2EEF8844E98EF14B4737B8C44B44CB240C92C4E766E32D1C692621

Share this post


Link to post
Share on other sites
10 minutes ago, mouradski said:

You're welcome, glad the tool was useful :) 

Thanks for the awesome tool.  I check it daily.  Love your work.


Does the new ATH just reflect the fact that you are now adding in the Fiat transfers, or is it truly a new ATH even after adding in the Fiat retrospectively?

Share this post


Link to post
Share on other sites
26 minutes ago, Dogowner5 said:

Thanks for the awesome tool.  I check it daily.  Love your work.


Does the new ATH just reflect the fact that you are now adding in the Fiat transfers, or is it truly a new ATH even after adding in the Fiat retrospectively?

thanks

I think the new #ATH is due to the new corridor that was hidden for us

Share this post


Link to post
Share on other sites

One thing that was troubling me recently, is the disappearance of the AUD/PHP corridor as can be seen clearly in this plot - the brown bars go away near the end of April. This new off-ledger EUR/USD corridor makes me wonder if the AUD/PHP stuff has gone the same way. Are there exchanges that use both currencies that might be doing that?

I'm not able to extract off-ledger info using my current scripts and I don't want to duplicate the effort that @mouradski et al have put into utility scan - is there any way I can easily assimilate EUR/USD data into my plots? Question for @mouradski - how far back can I query the order books on bitstamp for example to get the data I'd need to add it to my charts? I could simply add my plots to the utility-scan code, but I haven't written java in 15+ years and have no desire to do so (all my plots are using python+pandas). Do you by any chance dump out any ODL data that I can scrape and include off ledger corridors? TIA

image.thumb.png.7a40a67ac02a630d7e62806d432a7546.png

Share this post


Link to post
Share on other sites
28 minutes ago, jbjnr said:

One thing that was troubling me recently, is the disappearance of the AUD/PHP corridor as can be seen clearly in this plot - the brown bars go away near the end of April. This new off-ledger EUR/USD corridor makes me wonder if the AUD/PHP stuff has gone the same way. Are there exchanges that use both currencies that might be doing that?

I'm not able to extract off-ledger info using my current scripts and I don't want to duplicate the effort that @mouradski et al have put into utility scan - is there any way I can easily assimilate EUR/USD data into my plots? Question for @mouradski - how far back can I query the order books on bitstamp for example to get the data I'd need to add it to my charts? I could simply add my plots to the utility-scan code, but I haven't written java in 15+ years and have no desire to do so (all my plots are using python+pandas). Do you by any chance dump out any ODL data that I can scrape and include off ledger corridors? TIA

image.thumb.png.7a40a67ac02a630d7e62806d432a7546.png

 

we are lucky that the implementation of the Bitstamp API brings together all the trades associated with the same order book on the same timestamp, so just group the trades by timestamp on one side as in the other, then compare the sums that are close

 

you can focus on this class : https://github.com/mouradski/xrapid_alert/blob/develop/src/main/java/space/xrapid/listener/OffchainCorridors.java

regarding AUD / PHP I also find it very intriguing

Share this post


Link to post
Share on other sites
10 hours ago, jbjnr said:

I could simply add my plots to the utility-scan code, but I haven't written java in 15+ years and have no desire to do so (all my plots are using python+pandas). Do you by any chance dump out any ODL data that I can scrape and include off ledger corridors? TIA

Or perhaps utility-scan could implement such a plot because it's beautiful and very effective imho.

Share this post


Link to post
Share on other sites
8 hours ago, Frisia said:

Or perhaps utility-scan could implement such a plot because it's beautiful and very effective imho.

I'd happily add my plots to utility-scan, but it would require a java-python bridge that is a waste of time really. That's why I was asking if their code dumps out any data about the corridors - it would be easy to run the plots as a post-processing or secondary step. I don't really want to reimplement the order book scanning as hopefully the number of exchanges and trades will grow significantly over time and it's a waste of resources having many people all doing the same scanning over exchange APIs.

Share this post


Link to post
Share on other sites

×
×
  • Create New...