Decentralised Digital Asset Registers – Concepts

I am hugely optimistic about the role cryptocurrencies (such as bitcoin) will play in the future – and one of the reasons is that they enable us to build decentralised digital asset registers.  I’ve written about this concept here.

In this post, I’ll explore some of the current thinking on how to build such a system.

The simplest way to think about this subject is to imagine you own one hundred twitter shares that you would like to sell and, because you’re one of these early-adopting trailblazers, you want to sell the shares for Bitcoins and want to do so using only the bitcoin system.  Here’s how it could work:

As I write, Twitter shares trade for just over $40, so your one hundred shares would be worth about $4000.  So you could announce to the world that a particular Bitcoin you own (strictly, a transaction output) is for sale and that you will give whoever buys it all the rights associated with the twitter shares… e.g. dividends, votes, etc.  You don’t plan to transfer the share through the regular equity settlement systems in your country, though; you’ll remain the registered owner there… but provided you are trustworthy, the recipient will trust you to pass on the benefits you receive to them.

A Bitcoin today trades for about $300.  So if you could find somebody who trusts you, they might be willing to pay you about $4300 for your special bitcoin ($300 for the Bitcoin plus $4000 for the rights to the shares).  Perhaps they’d demand a discount to account for the ongoing counterparty risk they have to you.  So let’s imagine they offer to pay you $4000 to keep the maths simple.

To make it interesting, let’s imagine that Alice is willing to buy 25% of your holding ($1000, or 3.33 XBT) and Bob wants 75% ($3000 or 10 XBT).  What we want is for them to transfer these coins to you and for you to transfer your “special” (or, colored) coin to them so that, once you’re done, you have 13.33 XBT and they have a share of the colored coin.

Graphically, this is what it might look like:

Colored Coin Diagram

In this picture, we see that a Bitcoin transaction has been constructed that has the following interesting properties (in reality, it may be done as a sequence of transactions and I’m not 100% sure you could actually do it this way for real on the current Bitcoin network but the concepts remain the same so we’ll stick with this for the purposes of this post)

  1. Alice and Bob pay their agreed amounts of Bitcoins into the transaction (i.e. their 25% and 75% share of the costs) – colored green in the diagram
  2. I pay in the coin I have previously asserted to be “equivalent” to 100 twitter shares – colored orange in the diagram
  3. Alice and Bob receive 25% and 75%, respectively, of the special (colored) coin so that everybody can now see that they are the owner of the coin, and hence entitled to their shares of any benefits associated with the shares – shown in orange on the right-hand side
  4. I receive Alice and Bob’s payments – shown in green on the right.

Easy, right? Well…. not quite.

The problem is step 3.  If you are an independent third party arbitrating a future dispute, how do you know that the 0.25 XBT and 0.75 XBT received by Alice and Bob relate to the ‘colored’ 1XBT I paid in?  What would have happened if I had also paid somebody else 0.75 XBT in that transaction? How would we know one of these payments was the special colored coin and one was just a regular bitcoin?

Worse, what would happen if Alice or Bob somehow temporarily forgot their coins were special and spent them as if they were normal coins? They could lose a fortune! It would be helpful if their wallet software warned them. Which means the wallet would need to be able to tell automatically. Worse, what if Bob used his colored coin as part-payment for a larger expense, with regular coins making up the difference?  How on earth would the recipient interpret their receipt of a transaction output that was formed from combining colored and non-colored coins?!

The answer is that there is nothing in the core bitcoin system that allows you to tell.  So various conventions have been proposed.  The simplest is one that just relies on ordering, but there are others, none of them particularly satisfactory.  But it’s being worked on.   I think one piece of work in particular has helped in this space by generalising this problem by introducing the idea of a “color kernel“, whose job it is to decide which outputs are related to which types of coin.  Projects such as bitcoinx are working on implementing a system based on these concepts.

In a future post, I’ll discuss a very different model, that of mastercoin.


24 thoughts on “Decentralised Digital Asset Registers – Concepts

  1. Pingback: Decentralised Digital Asset Registers – Mastercoin | Richard Gendal Brown
  2. Excellent post! The lightness of your gray font made it somewhat more difficult to read until I zoomed in.

  3. Pingback: Kiến trúc sư trưởng của IBM lạc quan về Bitcoin | Bitcoin Việt Nam
  4. Pingback: Bitcoin、未來貨幣與金融業,IBM「可能」是這樣看的 | TechOrange《 爆橘 |
  5. Pingback: A Simple Explanation of How Shares Move Around the Securities Settlement System | Richard Gendal Brown
  6. I came here through I could not understand the concept of colored coins. But your example clarifies a lot. But i still have a few questions. I understand that the formal registration of the shares stays with richard.
    1. How do alice en bob receive the dividends? Does richard have to transfer these to alice and bob. Is this not a life long bureaucratic burden?
    2. What happens if Richard dies? Are the colored coins worthless then?

  7. @erkin: great questions.

    1) Yes – the issuer of the coloured coin would be responsible for securities servicing — paying dividends, informing coin holders about forthcoming votes, etc. i.e. all the things a custodian bank does today.

    2) Yes – generalising your example, what happens if the issuer disappears/goes bust? My take is that your coloured coins would very quickly lose their value. I see three ways round this: 1) have the issuer be a custodian bank… this is their business, after all. 2) have the issuer be the company itself (if it goes bust, the shares would be worthless anyway!, 3) accept the risk and model/price the coins as if they were CFDs/Equity Swaps, not direct holdings. I’ve expanded on this thought here:

  8. Biggest issue is managing of physical asset in relation to virtual one, in this case a color coin (or some derivative of it). This still doesn’t eliminate custodians and intermediary so I don’t see how this can reduce transaction cost like what bitcoin and other variety has done for money transfer and payment?

  9. Hi Joe,

    Fair points. I agree that there is an ongoing requirement to service the security (corporate actions, etc). However, representing it on the Blockchain might allow transaction cost associated with settling trades to be reduced. However, if one thinks this through further, it becomes possible that the public nature of the blockchain allows the roles of custodian and registrar/stock transfer agent to be coalesced. My thoughts are still at an early stage but I’ve tried to elaborate on them further, here:!

  10. I am still finding colored coin very confusing. First I need to buy a whole bit coin ? then I color it ? then I can trade the new colored coin for an asset like a house ?

  12. Pingback: 74K Jobs Added in Dec & QE5 / Iran, Russia: oil-for-goods / Terrorism as an Instrument of Imperial Conquest | paulthepaperbear
  13. @Brent – not quite. In your example, you talk about buying a house so let’s go with that…. The house you want to buy is owned by the current owner and his/her title will be recorded in a land registry somewhere. If they sold it to you, it would be necessary to update the registry, which would take time and incur fees, etc. (In what follows, let’s assume the house is worth $200k and that one Bitcoin is worth $1000, to keep the math simple.)

    The idea with colored coins is that they could do the following: they could announce to the world that a particular Bitcoin they own (or a fraction of a Bitcoin) will represent their house from now on… and that whoever owns that coin owns the house — i.e. has the right to live there, will be responsible for property taxes and has the right to sell it on in the future. So if they were to transfer that particular Bitcoin to you, then you would now be considered to own the house — and because they had announced to the world that this coin represented ownership of the house, everybody else would now recognise you as the owner.

    In such a scenario, you might be willing to buy that “special” Bitcoin from them for more than the value of the Bitcoin – you might consider that particular Bitcoin to be worth $201k… since ownership of that particular Bitcoin is equivalent to owning the house ($200k for the house plus $1k for the underlying Bitcoin itself). In this way, “ownership” could be transferred through the transfer of Bitcoins, without any updates to the land registry. You might pay them in dollars… or you might pay them 201 “regular” Bitcoins — since that single colored coin (which represents ownership of the house).

    Now, there are several problems with this idea, not least the fact that you are completely reliant on the ongoing honesty and performance of the “seller” – since they are still recorded as the owner on the land registry. But hopefully this example motivates the idea? i.e. it’s not a stretch to consider what happens if the land registry itself acts as the issuer of the “property colored coin” and so on.

    I accept that this space needs far more thoughts and development before it could work in reality – but hopefully the potential value of “coin coloration” is clearer now?

  14. Pingback: Bitcoin 2.0: Distributed Corporations, Derivatives, and Information Markets | The Ümlaut
  15. Pingback: Kiến trúc sư trưởng của IBM lạc quan về Bitcoin | Nguyễn Hoàng Huy
  16. Pingback: קריפטוכסף | אסכולת הכורסא
  17. Pingback: Microsoft ermöglicht Bezahlen per Bitcoin |
  18. Pingback: Microsoft ermöglicht Bezahlen per Bitcoin | Neues Münchener Tagblatt
  19. Thank you for taking the time to explain this. All the posts on your site have been a tremendous help in clarifying developments in the bitcoin/blockchain space.

    A few questions – though I see this post is outdated (if you’ve addressed these questions in a more recent post, would appreciate the link):

    I read about these colored coins and construct a thought experiment: imagine I were to write the terms of a binding mortgage onto the face of a physical U.S. dollar bill. From a purely monetary perspective, do you think this comparison is 1) legit and/or b) worthwhile? I ask because, while I appreciate that media too often focus on bitcoin price, I wonder how colored coins might affect the money supply or price discovery of bitcoin as an asset. I’m also curious about how a market might be lured one way or another to select a specific amount of bitcoin to represent another asset. Am I just going down a rabbit hole here?

    To carry on with that mortgage-on-a-dollar, who ‘governs’ the ‘ink’ that I used to write the contract on that dollar? If it represents a stock, and the company tanks, and we want to erase the ink, who erases that ink? Is this amount of money forever stricken from participating as a currency? Is this the business of some 3rd party securities service (i.e. a bank)? Should we even care?

    Lastly, when it comes to colored coins representing physical assets that can be theoretically governed by smart contracts to mitigate (if not eliminate) counterparty risk (to continue with the dystopian faddish example: that mortgage contract locks my doors if I don’t submit this month’s payment, or my car won’t start if I don’t pay my car loan), is that possible with colored coins? Any reason we wouldn’t avoid coloring and just implement a sidechain here?

    Again – I understand that the conversation has come a long way since the original post. I appreciate any direction to more recent posts or an update here!

  20. @jeff – great questions – thanks. I’m not aware of a fuller treatment anywhere else (please do share links if you find one) so some quick thoughts here:

    1) The “writing onto a dollar bill” analogy is one I use when presenting and I think it’s a good one… as your example shows, it provides a visual image for the concept and immediately highlights some of the conceptual issues. With respect to the “ink” question, my take is that it’s still a dollar bill…. so if somebody wants to accept it as payment for $1 worth of services, they’re welcome to. So you could, perhaps, model the value of a colored coin as “value of the asset it represents PLUS the value of the bitcoin (fragment)”

    2) I guess a system like this might drive up the demand for the underlying token to some modest extent (since you need an underlying token in order to colour it) but I don’t imagine it would be a material factor in the price (I could be wrong though)

    3) The interaction of colored coins and smart contracts is an interesting one. I’m working on a post about this as it happens…. should be ready in the next few days. (In short: I wonder if the right model is a standalone smart contract that governs the behaviour of an issue and a collection of tokens (colored coins if you like) that represent ownership… so the smart contract and the ownership tokens would be distinct and the former would track the latter)

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s