Jump to content

JoelKatz

Ripple Employee
  • Content Count

    883
  • Joined

  • Last visited

  • Days Won

    82

Posts posted by JoelKatz

  1. There are probably a lot of factors at play. One of them is that there are a lot of transactions that seem, at least to me, to be basically junk. With transaction fees so low, there's no real disincentive to submit lots of junk transactions, and people do. The legitimate volume is somewhat lost in that junk.

    To get more useful numbers, you can limit yourself to just Payment transactions or OfferCreate transactions that crossed other offers. And then you can filter out Payment transactions that just seem to be moving assets for no particular reason (we used to see a lot of those, but not so much anymore). Our XRP Charts site and XRP site has metrics like that, and they're probably more useful than raw transaction volume.

  2. 1 minute ago, RippleHerToShreds said:

    "Weekly XRP Trade Volume is defined as the total amount of XRP traded on the Ripple consensus ledger during a given week."

    Off-ledger trading is the majority of trading today. I think the largest on-ledger XRP market is XRP to USD/Bitstamp and it's the 41th largest XRP market.

  3. On 11/24/2017 at 11:33 AM, galgitron said:

    I think this is a little flawed because it assumes an even distribution of those sales, but your selling strategy focuses on higher sales during periods of peak interest.  This subtle but critically-timed XRP selling strategy, artificially deflates the momentum that would normally gather around a speculative moonshot; the metaphorical beat of a butterfly's wings per se.  I've seen this in Monte Carlo sims where small cash amounts are ran through historical-data trade simulations, and the winning strategies work wonders in the real world for small amounts, but take those same strategies with ever-so-slightly more than inconsequential amounts, and now you have trades that influence the market and they completely eradicate the efficacy of the strategy.  It doesn't take much to make a huge difference.

    The point is, if we killed Kim-Jong Il when he was born, we wouldn't have his nuclear arsenal to deal with today.

    To your point about using that cash to cement in the foundation for XRP, by all means, go for it; I couldn't agree with that strategy more.  I just don't think denying price manipulation, however subtle, indirect, or tangential, is completely forthcoming.  If you truly want to be non-disruptive, then your volume of sales should be consistent irrespective of volume peaks, or at least adjust on a running 30-day median.

    I agree with you. Unfortunately, I don't know the exact details of how the transactions are timed or how the spreads are computed. The third party market makers we've employed understand that we don't want to kill rallies or engineer the price and that our desire is to choose sale targets and approximately reach them having as little effect on the price as possible. To prevent me from gaming the system either for personal benefit or to benefit Ripple, or inadvertently giving others the ability to do so, I am not given access to any details. Every suggestion I've made to them got back "we're already doing that", so I think they know what they're doing.

    They might adjust it to a long term median. They might back off when volume is above normal. Those things I don't know.

  4. 20 minutes ago, Sukrim said:

    Nearly nobody seems to use custom tools anyways, most don't even seem to run their own rippled server...

    Would removal be permanent? Why would a payment of 1 XRP to a removed light account fail instead of just re-opening it (is there a minimal initial balance to create one?)?

    I guess it depends on the rules we decide to go with. If the rule is that sending less than 20 XRP automatically creates a light account, then it would succeed. That's probably what we'll do, so it won't fail. But if the account was removed and 20 or more XRP is sent, then a regular account will be created and 20 XRP will be locked as the reserve. That might annoy people.

    We could change it so that the Payment transaction always creates a light account by default. The account owner can promote it to heavy account if they wish. But we'd need a way to create a regular account with a Payment transaction because otherwise, existing flows for offline account creation will break. (Because you don't know the new account's starting sequence number and so can't prepare transactions offline.)

  5. 4 hours ago, Sukrim said:

    "be there" in the sense of "being present in the current ledger's AccountState", right? Because it would just get re-funded once I send some XRP to it. Maybe a different prefix would be useful for these accounts, but then again it might make the distinction between account types sharper than it maybe has to be and would make upgrades painful.

    Yes, that's right. However, there are still issues with funding the account after it's been removed.

    For example, suppose you know that an account once existed and you try to send 1 XRP to it. If it was a light account that was removed, that will fail.

    Say you know that an account once existed and you send 20 XRP to it. If it was a light account that was removed, you'll create a regular account.

    So it's not perfect. Our recommendation will be not to remove a light account unless you're relatively certain no more funds will be sent to it.

    We thought about making a distinction between account types, but the problem with that is that all tools would have to change for light accounts to work properly. Everyone would have to understand the new account format.

  6. 22 minutes ago, WrathofKahneman said:
    • JK says that's great, but makes a quip that if Mexicans really do love XRP that much - if they are buying it sooo much, then  perhaps any raise we see on the Mexican exchange might not be due to Cuallix using xRapid afterall. 

    Yes, that's my point. Of course it's great that Mexicans like XRP. But to the extent that's true, that makes the price/volume anomaly of XRP on Bitso less likely to be due to xRapid and/or Cuallix. This is the problem with making an inference from only one data point.

  7. 7 hours ago, Sukrim said:

    But they can never be demoted back to a light one, right? Or could they (once their sequence number is back to the rules of a light one and their Owner Directory is empty etc.) be set back too?

    Not without an awful lot of work. We'd need some way to reliably ensure that they weren't referenced anywhere in the ledger and many complex code paths would have to be audited. It's not impossible, but adding it to the initial design would make it much less likely that the the change would actually happen.

    If the light account proposal is adopted, it could happen that over time it gets expanded. Already, with the proposal as is, all ledger users will have to be careful that their code doesn't make the assumption that an account they saw on the ledger, sent XRP to, and received XRP from will necessarily always be there.

  8. 5 hours ago, Sukrim said:

    I don't really see the advantage in that proposal - how would it reduce spam or ledger bloat?

     

    A light account is guaranteed not to have any owned objects and hence no owner directory. As a result, it can have a lower reserve. A regular account can't be completely deleted because other people might have created trust lines to it and because that's needed to protect against transaction replay. Light accounts can have any trust lines created to them and they use a different mechanism to prevent transaction replay.

     

    5 hours ago, Sukrim said:

    Also it would give a LOT of power to V2 account holders for no real benefit imho. Also why make trustlines so extremely expensive? If anything, they should become cheaper so people are incentivized to use them instead of funding some XRP only accounts.

     

    Trust lines just increase the account's reserve by 5 XRP. They only charge the side that puts the line into a non-default state (unless both do). A gateway can easily have an owner count of zero because it doesn't have to put any trust lines into a non-default state! The 20 XRP base reserve covers things like the owner directory and costs associated with attaching offers. A light account can be promoted to a regular account at any time.

    The idea of the light account proposal is to allow people to have their own accounts on the ledger and send and receive XRP without losing 20 XRP they can never get back.

  9. 1 hour ago, John_Buh said:

    I have tried and tried to embrace TA but the one nagging problem I have with it is as time's arrow moves to the right the overall context that led to the curves and whatnot behind it have necessarily changed.  A butterfly flaps its wings in Australia, a Korean Finance Minister says the wrong thing, etc.  How then any sort of meaningful prediction can be made is a deep mystery to me.

    I find that my TA is nearly 100% accurate when I can see both to the left and to the right of the forecast time. But when you try to tell whether or not there's consolidation without knowing whether it was followed by a rise or a fall ... I never seem to be able to do it.

  10. We've thought about offering such hedges ourselves. We already have effectively infinite exposure, so a little more wouldn't hurt us and it could even be a small source of revenue.

    We've also talked to partners about forming their own entity with the same ownership but not technically a bank. That entity can hold their XRP or hedge it. This is especially helpful for FIs whose transaction banking portions can't own speculative assets. It can always be structured so the entity that holds the XRP can be expected to make a profit. However, if XRP is too volatile, that can result in the transaction banking entity not finding XRP competitive. So low holding cost is important to targeting higher volume corridors.

×
×
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue.