Big Banks’ Blockchain: Boondoggle?

Suits join the hoodies with blockchain push, the FT informed us on Tuesday, adding, on Wednesday, in case we missed it, pretty much the same article again: “Big banks push forward with blockchain technology“, From the first one:

The latest example of big banks organising themselves to exploit the potential of blockchain technology came this week with the announcement that four big lenders have teamed up to develop a “utility settlement coin” — a new form of digital cash.

The four banks — UBS, Santander, Deutsche Bank and BNY Mellon, which are working with UK broker ICAP and developer Clearmatics Technologies — stress that they are not creating a new cryptocurrency.

The aim is to speed up clearing and settlement in financial markets by allowing institutions to pay for securities, such as bonds and equities, without waiting for traditional money transfers to be completed in the so-called delivery-versus-payment process.

By switching clearing and settlement of financial markets on to a distributed ledger, the banks hope to do away with much of their costly back office operations that process trades and keep records up to date. Quicker settlement should also free up capital that banks hold against trading risk.

Hyder Jaffrey, head of fintech innovation at UBS, says: “Every bank, exchange and clearing house, we all have our own sets of the same data, which get out of sync and have to be updated and reconciled.”

“The distributed ledger is the first technology which could implement a shared golden copy of that data,” says Mr Jaffrey. “If you have that breakthrough you can really see how that would be revolutionary in the financial world.”

Yet the banks don’t seem to be using blockchain’s key feature at all:

Some sceptics reckon the banks are missing the point. “This is banks talking to each other and the point of blockchain is to establish consensus in the presence of potentially untrusted actors, as with bitcoin, on the internet,” says Dave Birch, of payments consultancy Consult Hyperion. “It’s a sorry state of affairs, that technology is not going to fix, if the banks don’t trust each other.“

There are also doubts about whether blockchain technology can actually do the thing the banks want it to do:

The project, however, still faces many challenges. One is transaction speed. Bitcoin is often criticised for being unable to scale up because its blockchain can handle only about seven transactions a second, as opposed to, say, the 24,000 a second that Visa can. Any solution from the banks will have to be fast and capable of processing heavy loads.

So who’s made the massive software design breakthrough necessary to get the banks interested? A clue comes from a quick look at the web site of the consortium’s technology partner Clearmatics Technologies:

Clearmatics is based on the Ethereum Virtual Machine, specialized for financial and fiduciary computations, and it uses a new consensus protocol designed to achieve finality of settlement and eligibility as a Designated Settlement System.

Ethereum last hit the headlines just a couple of months ago, in connection with the DAO hack, in which $50Mn of the $150Mn raised by the DAO a month earlier suddenly went for a walk, with assistance from some hackers and a not-so-smart Ethereum smart contract. So that little misadventure is hardly Ethereum’s fault: Ethereum was just the infrastructure of the DAO, not the guiding mind behind the DAO dumb contracts. The real questions are still: is Ethereum any good for the banks? What’s the new technology?

This “new consensus protocol” seems to be Proof of Stake:

In this new algorithm, agreement within the blockchain would be measured not on the basis of how much computing power agrees with the current state, but instead on the basis of how much digital currency agrees with the current state. The owners of this digital currency hold a financial stake in the success of the blockchain that tracks it, which is where we get the name for the algorithm.

Proof of stake allows implementation of an idea called “sharding” (explained in brutal but still incomplete detail here by Ethereum’s founder, Vitalik Buterin, section “Solutions”, pp 21-30):

…the bulk of the [blockchain’s] state is in fact distributed between the different nodes in the network, with each node only holding a small number of ‘shards’ of the state.

Of course, the whole point of sharding is to move away from the “everyone processes everything” paradigm, and split up the validation responsibility among many nodes. With naïve proof of work, doing that securely is difficult: proof of work as implemented in Bitcoin is a completely anonymous…consensus algorithm, and so if any one shard is secured by only a small portion of the total hashpower, an attacker can direct all of their hashpower towards attacking that shard, thereby potentially disrupting the blockchain with less than 1% of the hashpower of the entire network.

This changes, however, with proof of stake…the participants in the consensus process do have some kind of identity, even if it’s just the pseudonymous cryptographic identity of an address, and so we can solve the “targeted attack” problem with random sampling schemes…making it impossible for attackers to specifically target any particular transaction or any particular shard.

With that, and one or two more tweaks, the scalability issues will be banished:

The long term goal for Ethereum 2.0 and 3.0 is for the protocol to quite literally be able to maintain a blockchain capable of processing VISA-scale transaction levels, or even several orders of magnitude higher, using a network consisting of nothing but a sufficiently large set of users running nodes on consumer laptops.

Well, maybe; as recently as this February, a paper summarising the state of the art in blockchain scaling had this to say about sharding:

Such schemes, however, can incur substantial overhead when cross-shard coordination is required in a Byzantine setting, so sharding protocols for blockchains are an open area of research.

It’s now August, so either the state of the art is advancing lickety-split, or there are still one or two loose ends in Mr Buterin’s scheme; the problem of duplicates, perhaps, not explicitly addressed in Mr Buterin’s paper, but just as relevant to Ethereum as it is to Bitcoin:

Bitcoin solves the “double spend” problem with double entry accounting, and it ensures the integrity of the accounting system by guaranteeing that all transactions are globally unique and partially ordered. A “new” transaction must not be a duplicate of a transaction in any prior block, nor may it be a duplicate of a transaction already in the current block. For this to work somehow we must to look at every pair of transactions to determine that all transactions are unique.

This point seems frequently misunderstood, as you can see in this quote from one of the many Bitcoin scalability white papers and which is representative of the general trend:

The problem of simultaneously achieving the best of both worlds: having only a small portion of consensus participants explicitly participate in validating each transaction, but have the entire weight of the network implicitly stand behind each one, has proven surprisingly hard.4)Vitalik Buterin (31 May 2015), Notes on Scalable Blockchain Protocols. Available at <https://github.com/vbuterin/scalability_paper/blob/master/scalability.pdf>

Well, no. If we want to provide a guarantee of each transaction being globally unique then we must inspect every transaction. There’s nothing “implicit” about this. And if we weaken this guarantee (such as only ensuring that transaction are locally unique) then we break Bitcoin.

The consortium promises to have a commercial-grade blockchain system by 2018. That’ll be a mighty impressive achievement, if it happens. If it doesn’t, there’s plenty of money riding on it, and on many other competing schemes too, so it’ll be a mighty impressive bust.

Over to our sceptic again (I particularly recommend the perpetual motion link):

A lot of smart people have been working on the Byzantine problem (which blockchain is a partial solution to) for over 20 years and it’s well understood. We know the shape of what is possible, and current focus is on incremental improvement and niche applications where we can relax a constraint.

If you want the guarantees of consistency provided by blockchain then ultimately everything must pass through a single consensus processes. This is not a programming thing, it’s the laws of physics. We’ve known this for nearly a couple of decades, but unfortunately folk keep proposing perpetual motion machines.

We can easily do a lot better than Bitcoin’s few transactions per second, but the only parameters we have to play with are dwell time, block size and the strength of the consistency guarantee. Assuming that some smart person is just going to walk in and solve this problem is hubris.

Blockchain performance might always suck, but that’s not a problem

The sweet spot for blockchain – and distributed ledgers – is low volume, high-value exchanges. There’s a lot of interesting problems to solve in this space, from tracking diamond provenance through contract attestation and so on. It’s just that the pie in the sky, blockchain-taking-over-the- entire-financial-system predictions are likely wrong.

The final word goes to Reuters, hinting at what this UBS blockchain initiative is really all about:

“The practical use and implementation possibilities of central bank digital currency is rightly becoming a hot topic in the financial service industry,” Deutsche Bank Global Transaction Banking Chief Digital Officer Edward Budd said.

“It raises questions, and possibilities, over a fundamental market structure principle: who can have access to central bank money and how,” he said.

More on that in another post, I hope.

Print Friendly, PDF & Email

24 comments

  1. Mr Sharter, Esq

    http://www.urbandictionary.com/define.php?term=shart

    shart

    a small, unintended defecation that occurs when one relaxes the anal sphincter to fart (blend of “shit” and “fart”)

    I sharted at the party last night and
    went home pronto to change my clothes.

    ++++

    Shard

    a small, unintended defecation that occurs when one relaxes the anal sphincter to fart and instead raises 3 billion in an IPO.

    I sharded on wall st and the bank boys threw me a few billion.

    1. TheCatSaid

      So life depends on practising advanced control of the anal sphincter to achieve the desired outcome. . .

  2. Barry

    I think Blockchain could be quite useful for local currencies, L.E.T.S systems, triple entry book keeping for Financial payments taxes.

    1. John

      But alas this subset of geeks are not interested in helping society out especially the ones most in need. Just with getting bought out or cashing in on an IPO and the big banks can help them out. Terrible state of affairs really.

    2. Steve H.

      Energy use for verification expanded over 1000% from 2013 to 2014. It’s unsustainable and biases for low volume, high-value exchanges. That cannot support a local economy.

  3. Wade Riddick

    Not one word about how you go to a judge to reverse an illegitimate transaction. So it’s your basic clerk of court’s office without any conception of property rights and all that entails. Wealth is so valuable they don’t think anybody can steal it. That would mean you need the government, God forbid! A judge, a jury, cops, etc…

    But Ayn Rand doesn’t do justice or infrastructure.

    These guys can fantasize all day long about how they’ll finally wrest the power of issuing currency away from government but it’s not going to happen. Bitcoin isn’t a currency; it’s an asset. You’re talking about recording asset transactions in the ledger – which is what governments are supposed to do too. This smacks of a new way to create dark trading pools.

  4. Synoia

    ultimately everything must pass through a single consensus processes

    Not quite

    ultimately everything must pass through a repeatable single consensus processes

    It has to be repeatable to be a proof. If not repeatable, then the system cannot be trusted. If not tested repeatedly, and results compared, than there is no proof.

    1. hunkerdown

      Software + Testing → Volkswagen Lügendiesel + worthless attestations. Such nationally important systems as this (and Class II/III medical devices) demand provable correctness, constrained capabilities, and a chain of custody for the entire stack, from code down to transistors, in order to serve the nation’s interests. “But that takes all the fun out of it!” cry the Ivy League man-children.

  5. john

    so all these smart geeks want to build software to further help the crooked banks. clearly the blockchain community’s heart is not in the right place.

  6. vlade

    The problem with low-volume, high value transactions is that exactly those are the ones that are most likely to be disputed, put for arbitration, target for fraud/theft/etc. (the real high value diamond can be re-cut. Yes, it loses value, but if it’s not, it’s value to the thief is nil) etc. etc. In other words, exactly those that anything like immutability, smart contracts (either Turing complete or not) etc. are worst suited for, and where human ingenuity is at its best (I’m not talking about the results of it, but the inventiveness itself).

    Finance/contracts etc. are human constructs, and thus malleable. Yes, the malleability means bad things can happen, but trying to make them law of nature can make even worse things happen – because ultimately things like justice, fairness good and evil are human constructs too, and laws of nature don’t recognize anything like that.

    1. hunkerdown

      Imagine the TV series Suits rewritten for a world where credentials are attested by blockchain. Synopsis adapted from Wikipedia:

      Mike Ross (Patrick J. Adams) is a genius college dropout, who, with his natural intelligence and eidetic memory, makes a living taking the LSAT for others. To keep his terminally ill grandmother in a private nursing home, he agrees to deliver a large case of marijuana for his best friend/drug dealer Trevor (Tom Lipinski). The deal happens to be a sting operation which Mike astutely discovers and narrowly avoids, only to stumble onto a Harvard Law interview with Harvey Specter (Gabriel Macht), a lawyer at Pearson Hardman and arguably the “best closer in New York City”. Recently made senior partner, Harvey must hire a Harvard Law–educated lawyer as an associate, per firm policy. Following an interview, Mike proves to Harvey he is undoubtedly the best candidate, but is unable to present his digitally signed diploma and therefore escorted from the building into the arms of waiting police. He is tried and convicted of credential fraud, jailed, and, after being caught unofficially advising other inmates, serves out his twenty-year sentence in solitary.

  7. Alex morfesis

    Blockhead chain…rear admiral “amazing grace” must be laughing her tail off at arlington…200 billion lines of code and the vast majority who can plain english code will retire in 10 years…but with less than 100 courses globally and over a billion lines of code per year still being added to cobol legacy systems…

    Pull the football lucy and let’s cut to commercial…

    1. Synoia

      200,000,000,000 loc – lines of code
      200,000,000 kloc
      2 kloc/year/programmer
      100,000,000 person years of code
      50 years
      2,000,000 programmers.

      1,000,000 in the US

      Hmmm seems very high.

      1,000,000,000 loc added per year
      1,000,000 kloc/year
      2,000,000 programmers

      0.5 kloc/year productivity.

      Nope, you numbers are BS. Please supply link to source of numbers.

  8. Alex morfesis

    Synoia…
    Sorry…some small little players…ibm and gartner…you would not be the first person to insist those two monsters are pitching product but although there are those who share your “concern” about an over state, my point was more perhaps on the notion the blockhead chain would sit and “advance/change” the legacy cobol systems no one wants to spend money on to clean up now…

    Many links but my eyes were at

    Hackerrank july 6 2015 blog.

    Where is the budget to go to blockhead chain ??

  9. blogging.plato

    I cannot be the only one who sees “Blockchain” and reads “MERS” ? For those of us who do not work in the securities industry, can anyone give the ten cent explanation of why you want banking & securities transactions to be “sticky” as in slow and verifiable, as opposed to “at the speed of light” ? Or should I just get over those qualms and accept that there is no alternative to the coming Blockchain

    1. JustAnObserver

      +1

      Blockchain = MERS for every security on the planet, owned by guess-who.

      Absolutely.

      That “perpetual motion” side-link is priceless & says in hysterically funny (for a geek) terms the kind of thing Izabella Kaminska’s been saying about Bitcoin over at Alphaville.

      http://ftalphaville.ft.com/tag/bitcoinmania

    2. flora

      my 10¢ explanation: The more human eyes that vet a transaction, the greater the likelihood of spotting a fraudulent transaction.

      blockchains will “do away with much of their costly back office operations that process trades and keep records up to date.”
      aka, do away with human eyes.

        1. TheCatSaid

          That’s also how electronic voting fraud has been detected–apparent insignificant anomalies on results tapes of one kind or another (e.g., tabulators whose results show rounding errors, thus revealing decimalization of the vote; or printouts of a voting machine tape with a tiny spelling error that can only have come about through a tampered memory card.)

          Human eyes are crucial.

  10. Patrick Morrison

    > The sweet spot for blockchain – and distributed ledgers – is low volume, high-value exchanges.

    Seems to me that the real estate chain of title is inside that sweet spot. I haven’t yet reached my conclusion about whether it’d be a really good idea, or MERS on steroids.

  11. ewmayer

    For you econ/financial-blog history buffs, over the timespan of roughly 2-4 years ago, popular technotopian blogger Mike Shedlock was completely gaga over how bitcoin/blockchain were gonna be the disruptively innovative solution to just about every financial-transaction problem faced by humanity. At the height of his personal BTC-mania he was literally posting on the subject nearly every day. That abruptly stopped a couple years ago, but he’s now gone gaga over the self-driving-car revolution™ in similarly over-the-top fashion, in which he has even taken to publicly belittling readers who dare question his optimistic take in the comments section of his blog articles.

    We can only hope his current self-driving kick ends in similar “…and then suddenly fell silent” fashion as the BTC one did. I wonder to what extent this kind of behavior has cost him readership. I still scan his headlines every day, read mre deeply into perhaps one article a day, occasionally forward one of his more-sensible pieces to Yves, but in the wake of the BTC stupidity I stopped sending him newslinks and drastically cut my time allocation for reading his stuff. (I was never a commenter on his site, though, so thankfully no feeling of having misspent lots of time in that respect.)

  12. flora

    “The aim is to speed up clearing and settlement in financial markets by allowing institutions to pay for securities, such as bonds and equities, without waiting for traditional money transfers to be completed in the so-called delivery-versus-payment process. “ (my emphasis)

    Can anyone tell me how this isn’t check kiting?
    (please, no references to technology woo.)

    Thanks for this post.

    1. flora

      adding:

      “It’s a sorry state of affairs, that technology is not going to fix, if the banks don’t trust each other.“
      Indeed.

      Once the banks have slipped the leash of law, how shall they trust each other to conform to the law? This is not a question of scalability.

      “agreeing with the current state” [of the blockchain] does nothing to stop data drift (current state) over time away from actual, real world legal requirements.

      Remind me to stay far, far way from UBS, Santander, Deutsche Bank and BNY Mellon.

Comments are closed.