back to article Australian exchange pauses project to move stocks to blockchain

The Australian Securities Exchange (ASX) has paused its multi-year effort to replace its core trading systems with a blockchain-powered platform, written off up to AU$255 million ($171m) of work, and been advised to reconsider whether distributed ledger technology has a role in the project. "We have concluded that the path we …

  1. Pascal Monett Silver badge

    "Blockchain itself has not failed at the project"

    No, of course not.

    No one is authorized to criticize blockchain, the new darling of tech.

    It's the implementation that screwed things up, obviously. Blockchain is flawless.

    Except that it slows down transactions, can't manage concurrent access, needs to have endless workarounds implemented, etc, etc. In short, if they do insist on making it work, the code will be such a festering pile of crap that nobody will want to touch it with a bargepole.

    $171 million blown on this failure, and a lot more work to come. I think it's time to bury this turkey.

    1. LionelB Silver badge
      Meh

      Re: "Blockchain itself has not failed at the project"

      Sure you can criticise an application of blockchain technology, but there's really no issue with the technology per se. In this case it was quite clearly the wrong technology for the problem at hand - where "wrong" includes (but is not limited to) the fact that current implementations of blockchain had no chance of being able to satisfy the project requirements (this, you'd think, should surely have been spotted earlier). Plus you might well question (seems no-one did) whether decentralisation is actually a Good Thing for an exchange in the first place.

      But sure, whether blockchain turns out to be a useful technology for anything (and I'd include cryptocurrency in that anything) remains moot. As things stand, it still seems to be a solution awaiting a problem. Then again, you could have said the same about deep ANN technology a mere few decades back (and to anticipate snark, yes, that technology is now useful!)

      1. I ain't Spartacus Gold badge

        Re: "Blockchain itself has not failed at the project"

        Decentralisation seems a very odd decision for a stock market to pursue. Your very existence, your income and your customer lock-in all stem from the fact that they have to come to you. Push yourself out of the process, and eventually somebody might start asking awkward questions like - seeing as we're running all our transactions on the blockchain, what do we need an exchange for?

        Plus there's the obvious security problem. If people can find ways to muddy the ownership of shares, then they might be able to steal them.

        To be fair, there is a killer application that blockchain will be great for in financial services. If it ever works. And that's assets pledged as security. This can be for loans, short-selling, whatever.

        Let's say I have a government bond that earns a small amount of interest. I want to keep earning that interest and also make some more. So I borrow against it to get money to trade with. I still own the bonds, and I've got money to try and make profits with, and at a set time I've got to give the people I borrowed from that money back (plus interest) or give them my bond. Maybe my bets go bad, or just my company is in financial trouble. Perhaps I got to a second bank and borrow some more money, secured on that same government bond. Now we owe money to two different people, secured on the same bond. Naughty. But the incentive is there for companies in trouble to double down, break the rules and take some extra risk in the hopes of salvation. Their collapse then risk huge ripple effects as those debts that should have been secured, turn out not to be. This is why the inter-bank lending system broke down in 2008 - which could have killed a lot of banks if the Central Banks hadn't stepped in. Trust broke down, and nobody was sure if they were the only ones who'd been promised particular assets as collaterol. Made much worse by the fact that people no longer trusted each others' valuation of mortgage bases assets - so only really high quality bonds were acceptable, and everybody only had a limited amount of that. Central banks lowered their standards on what they'd accept as collaterol to cover this problem too.

        If you could set up an asset register on a blockchain, you could stop this happening. Which means if an institution collapses, there's much less risk of a cascade of failures taking out other institutions, because money they've borrowed will get paid off with the promised assets. Secured creditors get first dibs on the assets.

        The amount of work required to make this work though would be immense. As you'd need a trustworthy asset register in order to make the blockchain. So would that mean government detb, shares, even blocks of mortgages, would also have to be on a blockchain? I know the Bank of England are doing research on this, but I really can't see it happening without decades more work and large changes in how financial markets are run.

        1. Schultz
          Stop

          The problem of Blockchain in finance...

          The argument for blockchain technologies in finance (including exchanges, cryptocurrency, and the bunch) is that it pretends to solve the problem of trust -- but this problem cannot really be solved.

          Do you trust your bank, the stock exchange, your central bank to keep your money safe from fraud, theft, and inflation? They have a good track record, that's why they exist in their current form. If you replace these institutions with blockchain software, then you have to trust the programmers and bookkeepers of the newly created systems. Any blockchain needs to be maintained and stored and your 'safe' transactions still need to be properly recorded. So, really, you now just trust another set of people (unless you are inside the system, be it a bank or a crypto outfit, and roll your own). Be careful whom you entrust with your money <cough>FTX</cough>.

          1. LionelB Silver badge

            Re: The problem of Blockchain in finance...

            The argument for blockchain technologies in finance (including exchanges, cryptocurrency, and the bunch) is that it pretends to solve the problem of trust -- but this problem cannot really be solved.

            My understanding is that the argument for blockchain is more about decentralisation of escrow and control — and that this in itself gives rise to a problem with trust. If there is no 3rd party to trust with escrow and control, what can be the basis for trust? I think the prevalent view among cryptocurrency aficionados has been that mutual trust will evolve along with scale. We now know that, at least so far, to be optimistic. The norm, even at scale, seems to be huge volatility. And then, as you say, there is still the issue of trust in the custodianship and control of the technology itself.

  2. DS999 Silver badge
    Facepalm

    What a shock

    Replacing decades old tried and true database technology with the latest buzzword, which has as its main advantage "decentralization" - the last thing a stock exchange needs or wants - has proven to be a waste of time and money.

    Did I mention I was shocked?

  3. Richard 12 Silver badge
    Holmes

    So they finally noticed?

    Multiparty consensus is slow by design, because you have to poll all the parties.

    There's also the possibility that consensus is not reached, and then what?

    It's not possible to avoid these two things, they're fundamental to the entire concept of a consensus-based system, no matter the technology- or even none, simply discussion and voting amongst humans.

  4. ChoHag Silver badge

    > the application design leaves the ASX as "the central source of truth and final arbiter of outcomes, minimizing many of the benefits of a [distributed ledger technology] architecture."

    You can be distributed, or you can be centralised. Pick one. Blockchains don't let you avoid the implications of the CAP theorem, they're a way to relax the constraints/benefits of C in favour of A and P. Centralised systems sacrifice A in favour of C and P; things financial organisations tend to be keen on.

    But then again, it's not the rules of mathematics which apply in Australia.

    1. Claptrap314 Silver badge

      Huh. I've had my eye on this for quite a few years, and had always missed the application of the CAP theorem to the blockchain.

      But I'm going to disagree with your characterization. I would argue that blockchain actually sacrifices A, and that traditional (centralized) systems, not being distributed, drop P. For Bitcoin and Etherium, at least, a transaction is not added to the chain unless a majority agree to use it for the next block (thus preserving C), however, proof of that (the transaction settling) waits until some number of blocks are added (thus sacrificing A). I expect that other systems work in a similar fashion. There is really no way to talk about a "chain" unless there is only one.

    2. FrogsAndChips Silver badge

      Ah, the infamous "The laws of mathematics are very commendable, but the only law that applies in Australia is the law of Australia" by ex-PM Turnbull, thanks for reminding us of that!

  5. yoganmahew

    Hahahahahaha

    Hahahahahahaha, excuse me, I'm short of breath, hahahahahaha.

    As someone who spent many hours arguing against blockchain and the usefulness of a distributed ledger, let me just say

    hahahahaashasjdasoidhasdasdkjald

    Oh no, it's started again...

  6. Potemkine! Silver badge

    And another tech hype goes to the bin...

    == Bring us Dabbsy back! ==

  7. trevorde Silver badge

    Absolutely Incredible

    Accenture criticising blockchain. Must be sour grapes because they didn't get to do the implementation.

    1. LybsterRoy Silver badge

      Re: Absolutely Incredible

      I was also wondering which consultancy advised them to use blockchain to start with.

  8. Rich 2 Silver badge

    But it doesn’t scale

    One of the many problems with blockchain is that it doesn’t scale. The dataset just gets bigger and bigger and bigger. By design

    And the more transactions you have the faster it grows

    It seems to me a fatal flaw that nobody pushing it seems to acknowledge

    1. This post has been deleted by its author

      1. I ain't Spartacus Gold badge

        Re: But it doesn’t scale

        Def,

        But isn't the point of a blockchain that it's a permanent record of all transactions? And an ever growing one too.

        Whereas the point of a database of share or bond transactions, is to record current ownership and then the price/turnover history. Which is a lot less data.

        1. ChoHag Silver badge

          Re: But it doesn’t scale

          You *could* run a bank without any more than a recent snapshot of the transaction history. I wouldn't recommend it.

          It would be wiser to assume no financial institution has a delete function. Storage is cheap.

          1. This post has been deleted by its author

            1. LybsterRoy Silver badge

              Re: But it doesn’t scale

              I can only comment about the UK - businesses are by law required to keep detailed financial records for 7 years (at least that was true when I retired) and I suspect that banks are businesses so ....

          2. I ain't Spartacus Gold badge

            Re: But it doesn’t scale

            Well obviously the exchange has to keep data. But it doesn’t have to be in the live database, after a certain period of time. A blockchain is keeping all data ever. An exchange could dump data after ten years if it wanted to. But, as you say, storage is cheap. Data can be sold for research.

    2. DS999 Silver badge

      The difference is

      While a database also has a running log of all changes, you can archive older information that isn't current (i.e. that I owned a share of stock two years ago that you own today) so the amount of information being dealt with does not grow all that quickly. And some information can be deleted entirely, subject to laws about data retention for e.g. tax reasons.

      With blockchain it is all in there and each update depends on what came before so there's no way to archive off the older information and preserve its integrity. The fact I owned a share of stock you own today would still be carried in the blockchain 50 years from now, after I'm dead and that fact has no reason to be preserved - let alone be preserved in an active and immediately accessible form. It would be preserved even if the company in question was long gone!

  9. Pseu Donyme

    re: slowing down transactions

    This could actually have benefits if done right. Consider a stock market where buy and sell bids are paired at the end of the day so that the highest buy bid would be paired for the lowest sell bid at (buy+sell)/2, the pair is removed from the pool and this goes on until either buy or sell bids are exhausted. This would eliminate short term (<= 1 day) trading which destabilizes the market* and leaches money to the short term traders from the rest of the market for no discernible general benefit**. There is no real reason to run the market on a shorter timescale than a day (never mind a microsecond one): it is, after all, closed down during nights and weekends***.

    * potentially resulting the rest of the economy going pear-shaped for no good reason (screwing up even those who don't participate in the market)

    ** except, allegedly, providing liquidity, which this scheme would do as well if not better

    *** actually 1-3 times / week seems enough and would result in more stability and effective liquidity (for a stock or bond market, forex might need a shorter timescale (a couple of times / day, maybe))

    1. I ain't Spartacus Gold badge

      Re: re: slowing down transactions

      Pseu Donyme,

      How would your scheme provide more liquidity?

      The point of allowing people to make small amounts of money off short-term trades is that this profit incentivises them to do that very thing. If they don't exist, then the only people interested in buying are going to be holding for longer periods of time. There are going to be fewer of those, and if supply doesn't meet demand in the short term, you will get much larger fluctuations in price because of it. Just based on whether someone wants to be holiding a few £ million of say Apple stock this week, or doesn't want it for another week or so.

      Sometimes you may want to hold stock for a long time, but need the money for something else urgently, and change your mind. If you can't do that, then you will be less willing to buy. This is going to be particularly true for less traded stocks.

      HFT may be a problem that needs solving. Maybe we should be going back to the old days of market makers, or something new entirely.

      Also, trading is price discovery. The less frequently trades happen, the more you risk price volatility.

      Basically giving profits to those willing to risk money in very short term trades is the price you pay for liquidity.

      It's like people complaining about futures markets. That nasty old finance companies use them as a venue for gambling on the future price of wheat and this does no good for society and so should be stopped. The point is that there are people with money who want to buy risk - because they hope to turn that risk into profit. And there are say farmers growing food who don't like risk, they'd like to sleep at night. So are willing to forgo the possibility of future potential profit in order to have certainty of the price they're going to get for their crops. Speculators can be put to good use by other people, and that's one of the things financial markets can do.

      1. NeilPost Silver badge

        Re: re: slowing down transactions

        Buy risk, yes and good.

        War-Profiteering, no and BAD !

      2. LybsterRoy Silver badge

        Re: re: slowing down transactions

        Reading the above two posts I'm reminded of a number of others that I've read. When you manage to beat your way through the obfuscated jargon it seems that the whole point is to allow the traders to make more money.

        1. I ain't Spartacus Gold badge

          Re: re: slowing down transactions

          Reading the above two posts I'm reminded of a number of others that I've read. When you manage to beat your way through the obfuscated jargon it seems that the whole point is to allow the traders to make more money.

          I didn't think I'd used any obfuscated jargon. Please tell me where I have and I'll de-obfuscate it.

          The purpose of a financial market is for people to make money.

          And I mean this literally. The whole purpose of a financial market is that some people want money, in order to do stuff. That stuff might be starting a new company, buildling a factory, investing in some new piece of R&D, expanding into a new market.

          So they can ask people who've got money to lend them some. This is done in various ways. But that's what it boils down to. And those people lending money are taking a risk of losing it, so want a reward. You take a small risk by putting money in a bank's savings account, so you get a small amount of interest.

          Or for a little bit more you can buy government debt, or a corporate bond off Apple.

          Or you might take more risk and buy shares. If this is an IPO (a company issuing new shares) then you're investing directly into a new-ish company (or one that wants to expand) and probably taking more risk, for the expectation of higher rewards. Or you might buy shares in a safer bet, again say Apple. Who've had their explosive growth phase, but make nice juicy profits you can share in. In that case you're not funding investment - but nobody would buy shares in new companies if they didn't eventually have a hope of selling them. So there has to be a market in shares willing to buy (liquidity) or nobody would invest in new companies. Venture Capital takes a stupidly large risk by early investing into a tech startup - you get to own a percentage of the company. Most of the time it goes bust and you lose everything, but maybe 1 in 20 times your bet will pay off and you'll own a small percentage of Google or probably something a bit smaller but still profitable.

          This kind of investment wouldn't happen if there weren't people wanting to take stupid risks to make stupid profits, like say Softbank. But also then share markets for them to sell their few successes to, so they can cash out buy yachts and fast cars. And nobody would invest in those IPOs unless they had a chance to sell their shares to ordinary low-medium risk investors (like pension funds). It is complex, and there's no way to avoid that. But it also works. Most of the time.

          A lot of the people who manage the money don't actually own it. They're doing it for other people. That's our pension funds, for example. Or insurance companies. They need to make a certain amount of return on the money they're investing. So they'll invest some in boring, but safe, government bonds. And try to make higher returns with some - because that's their job. To try and grow our pension pots.

          There's a concept in economics called the agent problem. Which is that the people you hire to do stuff for you have their own interests, which may differ from yours. So as a shareholder you want regular dividends and a company growing in a nice predictable manner. But the CEO might want to get rich quick at the expense of long-term steady growth, or show what a genius he is by making some flashy deals and getting on the front pages all the time. Or a pension investment manager might be more interested in his bonus than he's worried about fulfilling your cat food needs in 40 years time. This causes problems, and it's almost impossible to devise pay and bonus structures to keep these people on the straight and narrow. So regular scrutiny is required, from investors, shareholders and government regulators.

          The problem being we want to incentivise them to take controlled amounts of risk in order to grow our money (if we're lucky enough to have pensions or life insurance) without going mad and risking everything in order to make their next bonus stupendously huge. A balance that is impossible to strike perfectly.

          If we take away the markets, then our only route to the investment that leads to economic growth is the banks or the government. The banks are no better at it than the markets, but more risk averse, so we'd certainly lose out on companies like Google and any kind of new business that's highly innovative that the bank manager can't understand. Government has a mixed record in business investment. And again, tends to be risk averse. So government investmtent banks can work brilliantly, in both Japan and Germany after the war for example. Because they were investing in known industries to build up their national economies. But then all but one of the Landesbanks in Germany (controlled by the State governments to invest in local industry) went bust in the 2008 crash and had to be bailed out. Only Bavarias was still being well run by that point. Similarly almost all the Cajas in Spain, regional semi-publicly owned savings banks, needed bail-outs in 20010-12. And the UK has a sad history of nationalising companies and then the government running them incredibly badly - with some exceptions.

          Financial markets do some incredibly useful things, with annoying side-effects. And so far, nobody has found a reliably better way of doing it - and we've had modern (ish) style banking for 600-odd years and modern (ish) financial markets for 300.

          1. Anonymous Coward
            Anonymous Coward

            Re: re: slowing down transactions

            There should be no problem with mitigating the side effects when it is easy. Like removing the incentive for high-frequency trading, which itself becomes a tax on transactions for non HFT transactions, and short term price manipulation with a Tobin Tax - a small tax an every transaction.

            1. I ain't Spartacus Gold badge

              Re: re: slowing down transactions

              Like removing the incentive for high-frequency trading, which itself becomes a tax on transactions for non HFT transactions, and short term price manipulation

              Anon,

              The problem with HFT isn't that it's become a tax on transactions. HFT has been beating out the old traditional market makers because it's cheaper that the service they used to provide. In fact, that's one of the few good points of HFT, that it tends to lower transaction costs.

              If we come to the conclusion that we don't want HFT to happen, which is something I do think (but I don't know enough about the markets to feel certain), then it's easy. We just ban HFT. Job done.

              Introducing a Tobin Tax creates other problems. You'll notice, for example, that the Swedes introduced on in the 80s, it didn't raise anything like the amount expected and vastly reduced trading, and they scrapped it in the early 90s. Also the EU looked at it after the 2007 crash. 11 (then 10) countries were going to try one, under the "enhanced cooperation" system, as the whole EU wouldn't agree. The Commission went off and researched it, and from memory the figures were that the tax would raise €30-35billion a year, but cost 0.2% of GDP to the economy overall. From memory, because I used to have a link to the EU Commisison reasearch on this, but they deleted it from their website. The costs came out at several times the amount raised in tax.

              But the other problem was that this was for having a tax on both trades in shares and bonds. And they were deeply worried about damaging the market in government debt, right during the middle of the Euro crisis.

  10. Berny Stapleton

    Testing done in Australia

    If they managed to pull it off, they could have sold the implementation. Win.

    With the lower latency of LSE and NYSE, there's no way they could have been a testing ground. I guess if they mothball the project now, they've got enough spare hardware that they won't need a server infrastructure budget for the next couple of years at least...

    1. NeilPost Silver badge

      Re: Testing done in Australia

      It’s probably in AWS.

      Shirley they wouldn’t contaminate the one true Block-chain with being on-Prem.

  11. Claptrap314 Silver badge

    Disappointing, but not unexpected...

    Transaction speed has always been a weakness of these systems. Now that ChoHag has brought up CAP, it really looks to be core. In order for these things to work at all, they have to be CP. That means A is flying in the wind.

  12. M.V. Lipvig Silver badge
    Pirate

    I can see the real goal here

    Of course, I'm a cynic with decades of experience.

    Stock market on blockchain - trade go poorly for a big dollar customer? Unwind it for a do-over. Big customer happy, little guy wonders what happened to the stock he bought. Think of the money that could be made while a court case to keep trades from being unwound slowly winds its way through the courts.

    I think this could be possible too, based on how scammers have already been able to screw up blockchains to steal millions with no real way to get it back.

POST COMMENT House rules

Not a member of The Register? Create a new account here.

  • Enter your comment

  • Add an icon

Anonymous cowards cannot choose their icon

Other stories you might like