"Their estimates did not include potential costs and investments required to deploy the technology."
well, of course not... because then there would be no dramatic teaser headline. Accenture money grab here - presumably they've positioned themselves accordingly already?
I agree with your sentiment. I do not see a scenario where financial institutions would agree to put their transactions into a distributed ledger with copies of the data (even encrypted) held by their competitors. If the financial institutions limit the ledger to computers within their control, what would be the benefit over a clustered database environment (I am assuming that most large financial institutions already have these).
The problem is without blockchain you don't have a meaningless buzzword to sell bullsh*t to banks. Blockchain is a vital technology for this.
Cryptocurrency is the future but I've never talked to more snake oil sales people than when talking to "blockchain" companies. I marvel at people's ability to sell nothingness
I worked in a consulting-type company for a bit and the most valuable lesson was that, generally speaking, nobody knows anything about anything. The most valuable institutions in the world have completely clueless people managing their tech and the ignorance only gets worse as you go up the chain.
It all boils down to "nobody got fired for buying IBM". Your blockchain initiative is only to get you noticed enough to get promoted or hired to a better position somewhere else. By the time anybody realizes how worthless it was you're long gone.
The discussion was around trustworthy data, so I think they want proof that a page served up and stored was in fact the actual page served up and stored, for later audit. HTTP doesn't give you that. Bittorrent doesn't give you that. But blockchain can, if you pay a fractional coin to store the hash of the page in the chain forever. With a reference to the page and to the transaction, you can then later determine whether the page is intact or has been modified. And for bonus points a signature could be stored as well.
This is just passing the overhead cost to whomever is mining the network though.
The trustworthiness of a blockchain comes from it's ability to repel attacks with computing power, with more power meaning more trustiness. This means tremendous overhead for a blockchain that could just as easily be performed with the centralized system we have now.
Banks spending billions on complex technical processes could save billions if they replace those systems with a new system and you ignore all costs associated with implementing and running that system. Well yes, but..
Blockchain is so badly needed in the startup world when allocating shares of a company. Eshares is promising but too pricey for bootstrapping entrepreneurs.
An open standard with distributed tracking of shares allocated for any given company would help institute much more trust into startup investing. Right now it's kind of tough to do due dilligence on a small private company.
> Blockchain is so badly needed in the startup world when allocating shares of a company. Eshares is promising but too pricey for bootstrapping entrepreneurs.
What, exactly, does a blockchain have to do with this? ISTM all that's needed is a spreadsheet with an audit log, and even the audit log is somewhat optional.
In theory, an open blockchain like Bitcoin eliminates the need for a trusted authority to be the system of record. A spreadsheet needs someone to manage it. That person, and that data, must be very trustworthy, else they can simply change the data for their benefit. The compliance procedures that assure trustworthiness cost money. Lots of money. And they still fail.
Blockchains, in theory, could also serve as a voting mechanism on splits, which could be manifested as blockchain hardforks. As Satoshi Nakamoto wrote:
> The proof-of-work also solves the problem of determining representation in majority decision making. If the majority were based on one-IP-address-one-vote, it could be subverted by anyone able to allocate many IPs. Proof-of-work is essentially one-CPU-one-vote. The majority decision is represented by the longest chain, which has the greatest proof-of-work effort invested in it.
Essentially a blockchain can codify the ownership of a stock share into a smart contract.
That's the theory, and only insofar as it applies to a Proof of Work blockchain like Bitcoin.
A private blockchain is more likely to be a Proof of Stake blockchain, in which shares (or tokens of other sorts) are issued in a "premine" which creates all of the tokens that will ever exist. These are then distributed to the initial group of trusted members of the private blockchain. From there, the blockchain can eliminate the need for a trusted authority to serve as the system of record of ownership.
Complete noob (in both FinTech and Blockchain), but in my limited experience in FinTech, if you could eliminate the need for the trusted nth party (Bank Of New York, for example, cleared all of the bond trades we made), then you could save a few 1/32's of a trade (my last client in FinTech was like 15 years ago so I don't even know if those fees are even valid anymore). In any case, as you indicate, this seems to be the canonical use-case for the block chain...
> A spreadsheet needs someone to manage it. That person, and that data, must be very trustworthy, else they can simply change the data for their benefit.
Google Docs has a change log could be a first step.
As an attorney, I can tell you I have spent countless hours simply verifying that cap tables/spreadsheets are accurately tied to the underlying capitalization documents in connection with VC financings and M&A deals. If this process was reliably automated, thousands of dollars in legal fees could be saved on a given transaction.
So ISTM what you want is a log of all underlying documents along with some machine-readable format for these documents so you can aggregate them into a cap table.
Using a blockchain for the log would work, but so would any other means of checkpointing the thing periodically and verifying that the audit log is really append-only.
IMO the real benefit of a blockchain is that an owner of some interest in a company could plausibly validate that their interest is actually logged in the source data for the cap table. But this would depend on having some way for the owner to know the current blockchain head, and this isn't going to happen by magic.
I understand this was prob a sort of joke, but I'd be interested in a data point on how much a free db could help save. Is $12B reasonable, would it be much lower/higher?
> I'd be interested in a data point on how much a free db could help save
Database companies price their products by how much they think they can extract from your business. That's why they're mostly priced at free or some nominal cost for the first few years, and then several million dollars per year afterwards.
free maybe. The real point being that it is one that works. A big part of the financial but also the entreprisey world in general have never left the 80s/90s. The idea of scale does not exist.
Blockchain here refers not to the distributed database technology loosely coupled with Bitcoin but rather "assert the existence of a database; simultaneous adoption of it across all parties that matter in the financial industry sure would be swell now wouldn't it."
Which is true. Databases are pretty useful things. It's a shame banks never thought of adopting one before.
"Another aspect of blockchain, as I often emphasize, is sociological: Blockchains are cool. If you announce that you are updating the database software used by a consortium of banks to track derivatives trades, the New York Times will not write an article about it. If you say that you are blockchaining the blockchain software used by a blockchain of blockchains to blockchain blockchain blockchains, the New York Times will blockchain a blockchain about it."
[p.s.] The "social" angle was the most ridiculous thing I've ever [read]. Yeah, let's pour in tons of money to build a tech so it gets printed in the New York Times. Is this guy for real?
This is tangental, but his analysis of Uber seems to be spot on:
>Eventually there will be self-driving cars, and there's no special reason to think that Uber will be the best at building them, or that its dominant position in the ride-hailing-app space will protect it from better self-driving-car businesses.
Indeed private blockchains never had a real relationship with the Bitcoin innovation. The Bitcoin innovation was a secure solution for performing transactions in a trustless/distributed/descentralized (e.g. the double spending problem in P2P transactions). The private blockchains are not trustless, so... this innovation doesn't applies.
Now, what we call private blockchains/distributed ledgers are basically a way to have a common workflow with [smart] contracts between different entities. This is a real alternative to just publishing APIs and orchestrating them.
But I think there is a "first mile problem" here: even if this technology solves some problems, how can you make different organizations to adopt a single technology? More when there are multiple competing projects.
Oh god, not this again. Sounds like Accenture are looking to bring in more 'technical integration' clients (read: marks). Government and old financial being the favourites, given their deep pockets and technical illiteracy.
If anyone is in a job that involves holding back some of this insanity, this decision-flow diagram might help: http://i.imgur.com/4nZ67Sq.png
I agree. There's no need for a blockchain if two parties can find a mutually trusted third party: use 2-of-3 multi-signature with one of the public keys being that of the trusted party.
I'd add do you need high level of transactions per second. If you need to reverse a hash every time you make a transaction, that's not going to be fast or efficient.
Also do you need transaction anonymity. A blockchain is also a public ledger. You may hide behind a secret key but everyone can see what that secret key is doing and has done since the begining. As soon as Morgan Stanley notices that this big trade has to be Goldman Sachs, then Morgan Stanley can observe all of the transactions that Goldman Sachs has done from the begining. That problem alone should close the debate. Markets need anonymity for all the other market players not to arbitrage your position.
Bank clearance only happens once a day, no need for high TPS levels. They do not transfer money bank-to-bank for every transaction, they just add it up and make one settlement transaction at the end of the day.
What I found especially amazing by that is traditional accounting is based on immutability already: you never delete a ledger entry, you add a correction. The same concept for a blockchain is trivial. Equally, if you _really_ do need to delete something, just insert a record into the blockchain instructing all nodes to delete that record. Sure, you won't be able to verify the chain 100% after you delete the data, but you'll have a auditable record of why that's the case.
Actually their view point is perfectly consistent. They criticized _bitcoin_ not the blockchain. While a blockchain could support some of the regulatory required undo stuff (by appending a record that it's been undone), bitcoin itself certainly doesn't won't let you roll back a bunch of transactions.
Read the PDF: they're working on a chameleon hash scheme so that their blockchain can be undetectably modified, not unlike making an accounting ledger with dry erase markers. That's going way beyond disagreeing with Bitcoin; what they're doing doesn't have the strong auditability that you'd expect from a "blockchain" solution. It's closer to a bog standard database really.
They talk about "reconciliation". Is this the practice of moving actual money following a bank-to-bank transfer of money? Or is this strictly an auditing function? Any bankers want to comment on this?
Reconciliation is the frequently manual process of making sure your records match mine, and vice versa. It's burdensome in finance because of the proliferation of systems used at various stages of a trade, and the number of parties involved as well.
Edit to expand on that a bit: it means making sure that every party involved in a trade has identical information. For a straightforward trade like an equity done on an exchange, it's more of an exception process. For more complex transactions where there are more data fields, it can be a very heavy process. Rough estimates put cost of reconciliation at tens of billions per year for banks, all of which has to be paid for somehow.
Can anyone name a single success story about the use of blockchain technology to solve a specific problem, and how this problem was solved using blockchain technology (besides Bitcoin)?
and much slower than Visa, never mind the sorts of transactions where there are algorithms earning big bucks profits from frontrunning trades with millisecond advantages....
Sure, my point is that where systems are optimised for speed Bitcoin is miles behind the curve. If a distributed ledger is simply updating to reflect the authoritative accounts of a centralised authority I'm not really seeing what trustlessness is adding to the reconciliation process either.
In a trustless decentralized world (ie the web) more speed means less security. There is definitely an equilibrium between centralized and decentralized systems.
I realize I'm starting to sound like a crackpot so I'll just recommend the writings of a more well-respected crackpot Dee Hock (founder of Visa Inc.)
Depends on which metric you look at. While Bitcoin has much faster settlement time than international wire transfers, it has slower throughput (~7 TPS max currently).
It'll be interesting to see what happens when(if?) Bitcoin upgrades to SegWit. This infrastructure is designed to support off-chain transactions while settling on the chain.
> IBM, which has been making a big push into blockchain technology, [...] aims to have it fully functioning by early next year.
I believe this is the timeless state of all big Blockchain projects: ready one year from now.
I specifically asked for success stories because I know everyone is throwing money after blockchain stuff. So I don't doubt that there are tons of planned projects. However, I haven't yet heard of blockchain technology being applied to solve an actual problem, except Bitcoin.
CME Group and the UK Royal Mint are planning to launch a blockchain-based product [1] -- also in 2017. But according to the article, it seems to be a simple digital promissory note, with the bearer owning the right to a certain amount of gold. Since we already have a central creditor, upon which the entire system depends (the Royal Mint), there is zero reason to use a distributed database to store anything.
Just put the database in the same safe as the gold.
Counterparty - create and register assets. Create secondary markets for those assets.
STORJ - (storage in a southern dialect combined with startup lingo) - A Counterparty asset that is used as fuel for a decentralized unlimited cloud storage platform. Farmers offer storage, consumers upload shards of their files to a variety of different farmers. Encrypted, unlimited, redundant. I believe the STORJ company now has VC funding as well.
OMNI - create and register assets. Create secondary markets for those assets. (Counterparty and OMNI compete for market share)
Tether - An OMNI asset which is used to create stable value cryptocurrencies that are representations of national currencies. The Tether company has a centralized account of USD and EUROs, but this enables people to hold stable value cryptocurrencies, or "cash out", but not have to worry about the long delay of getting national currency from a bank. Primarily used for trading.
All of the above metalayers create transactions on the bitcoin network, and settle on that blockchain, so if that disqualifies them all because they use bitcoin underneath, then Dogecoin has a Counterparty layer as well :)
> All of the above metalayers create transactions on the bitcoin network, and settle on that blockchain, so if that disqualifies them all because they use bitcoin underneath, then Dogecoin has a Counterparty layer as well :)
My point was precisely that you cannot, as far as I can see, separate blockchain technology from Bitcoin. My argument is that it's not blockchain technology that's useful, but rather a blockchain, in itself, that's useful. And it's useful because it's a distributed, proof-of-work timestamp server.
The value of Bitcoin comes not from the technology it uses, but from its incentive structure. An increase in the value of bitcoins feeds back into the system, resulting in an increase in security (hash rate) by way of an increase in the profitability of mining. Since Bitcoin's value comes from its time-stamping ability, and since this time-stamping is done using proof-of-work, it makes sense for an increase in the hash rate to also cause an increase in value, at least up to a certain point (provided no changes to the protocol are made to take advantage of the enormous hash rate of the current network).
"OCBC Bank said Monday (Nov 14) it became the first bank in Southeast Asia to use blockchain technology to make local and cross-border inter-bank fund transfers."
I've read the article, and many like it, but I can never find an actual explanation of exactly how a problem was solved using blockchain technology.
It's always marketing speak, with reference to all the wonderful properties of "blockchain technology", and diagrams like [1], which could have simply been summed up in three words: "no payment intermediary". Although... doesn't that diagram look like they just changed the intermediary's name from "Payment Intermediaries" to "BCSIS Blockchain platform", plus added some cool-looking server icons?
But the source code will almost certainly never see the light of day, so we will never really know how they use blockchain technology. At the end of the day, I think it's a lot more likely that this bank's software provider created a new solution and just said it has "blockchain technology" because it's a vague buzzword, and then the bank released a press release based on this, again because of the buzzword effect.
Headline - 4.22.23 -
The world banking market continues to struggle this spring. While blockchain financial services looked promising at the beginning of the decade, funneling as much as $12B per year from the human powered Financial Services sector, many viewed this as contributing to continuing wealth inequality. The "Decentralize and Replace" movement continues to grow. Many smaller banks have already folded, while larger banking institutions lobby for stronger regulations against decentralized blockchain technology...
propping up the existing banking system isn't the point of blockchain; sure it could save money but it could generate more revenue than 12billion a year if used for a new banking system.
its going to be fun when the banks and governments are issuing credit on blockchain assets, it is going to be a MESS when they go insolvent after the heist of their collateral.
2028? 2035? what year you guys think that'll happen
I worked for Accenture for a very short time, and I will say that they have no idea what they are talking about, regardless of what it is.
Somehow, I just found a charge for $3500 from Accenture on my credit report. It HAS to be from when I worked there. They either took back a bonus or thought that they paid me too much I guess. It's been years, I have never heard anything about it, and now my credit took a hit. Worst job/company ever would be an understatement. I quit in record time (<60 days).
The only reason your credit "took a hit", based on the way those formulas work, is because it went to collections or because you have barely any credit. And both of those reasons are completely your fault.
This is so terrible use case for blockchain. In a trusted environment blockchain might only introduce complexity and security problems. Hey, but isn't it what consulting is about?
82 comments
[ 3.0 ms ] story [ 157 ms ] threadwell, of course not... because then there would be no dramatic teaser headline. Accenture money grab here - presumably they've positioned themselves accordingly already?
With a blockchain, the database and the revision history would be part of the same artifact. Not sure how advantageous this is, however.
Cryptocurrency is the future but I've never talked to more snake oil sales people than when talking to "blockchain" companies. I marvel at people's ability to sell nothingness
When you combine the two it must be like a perfect storm of bullshit.
It all boils down to "nobody got fired for buying IBM". Your blockchain initiative is only to get you noticed enough to get promoted or hired to a better position somewhere else. By the time anybody realizes how worthless it was you're long gone.
How would storing this data in a blockchain have less overhead than an http server? Or even a torrent?
It seems like a hammer looking for a nail.
The trustworthiness of a blockchain comes from it's ability to repel attacks with computing power, with more power meaning more trustiness. This means tremendous overhead for a blockchain that could just as easily be performed with the centralized system we have now.
[1] https://www.accenture.com/us-en/service-blockchain-financial...
An open standard with distributed tracking of shares allocated for any given company would help institute much more trust into startup investing. Right now it's kind of tough to do due dilligence on a small private company.
What, exactly, does a blockchain have to do with this? ISTM all that's needed is a spreadsheet with an audit log, and even the audit log is somewhat optional.
Blockchains, in theory, could also serve as a voting mechanism on splits, which could be manifested as blockchain hardforks. As Satoshi Nakamoto wrote:
> The proof-of-work also solves the problem of determining representation in majority decision making. If the majority were based on one-IP-address-one-vote, it could be subverted by anyone able to allocate many IPs. Proof-of-work is essentially one-CPU-one-vote. The majority decision is represented by the longest chain, which has the greatest proof-of-work effort invested in it.
Essentially a blockchain can codify the ownership of a stock share into a smart contract.
That's the theory, and only insofar as it applies to a Proof of Work blockchain like Bitcoin.
A private blockchain is more likely to be a Proof of Stake blockchain, in which shares (or tokens of other sorts) are issued in a "premine" which creates all of the tokens that will ever exist. These are then distributed to the initial group of trusted members of the private blockchain. From there, the blockchain can eliminate the need for a trusted authority to serve as the system of record of ownership.
Google Docs has a change log could be a first step.
Using a blockchain for the log would work, but so would any other means of checkpointing the thing periodically and verifying that the audit log is really append-only.
IMO the real benefit of a blockchain is that an owner of some interest in a company could plausibly validate that their interest is actually logged in the source data for the cap table. But this would depend on having some way for the owner to know the current blockchain head, and this isn't going to happen by magic.
Database companies price their products by how much they think they can extract from your business. That's why they're mostly priced at free or some nominal cost for the first few years, and then several million dollars per year afterwards.
Which is true. Databases are pretty useful things. It's a shame banks never thought of adopting one before.
https://www.bloomberg.com/view/articles/2017-01-10/bank-bloc...
https://en.wikipedia.org/wiki/Dark_liquidity
[p.s.] The "social" angle was the most ridiculous thing I've ever [read]. Yeah, let's pour in tons of money to build a tech so it gets printed in the New York Times. Is this guy for real?
>Eventually there will be self-driving cars, and there's no special reason to think that Uber will be the best at building them, or that its dominant position in the ride-hailing-app space will protect it from better self-driving-car businesses.
Now, what we call private blockchains/distributed ledgers are basically a way to have a common workflow with [smart] contracts between different entities. This is a real alternative to just publishing APIs and orchestrating them.
But I think there is a "first mile problem" here: even if this technology solves some problems, how can you make different organizations to adopt a single technology? More when there are multiple competing projects.
If anyone is in a job that involves holding back some of this insanity, this decision-flow diagram might help: http://i.imgur.com/4nZ67Sq.png
Also do you need transaction anonymity. A blockchain is also a public ledger. You may hide behind a secret key but everyone can see what that secret key is doing and has done since the begining. As soon as Morgan Stanley notices that this big trade has to be Goldman Sachs, then Morgan Stanley can observe all of the transactions that Goldman Sachs has done from the begining. That problem alone should close the debate. Markets need anonymity for all the other market players not to arbitrage your position.
[1] Downside of Bitcoin: A Ledger That Can’t Be Corrected: http://www.nytimes.com/2016/09/10/business/dealbook/downside...
[2] EDITING THE UNEDITABLE BLOCKCHAIN (PDF) https://www.accenture.com/t20160927T033514__w__/us-en/_acnme...
[3] Accenture Launches A Way To Edit Blockchains: http://www.forbes.com/sites/tomgroenfeldt/2016/10/06/accentu...
Edit to expand on that a bit: it means making sure that every party involved in a trade has identical information. For a straightforward trade like an equity done on an exchange, it's more of an exception process. For more complex transactions where there are more data fields, it can be a very heavy process. Rough estimates put cost of reconciliation at tens of billions per year for banks, all of which has to be paid for somehow.
Investment banks have chosen to pursue 'blockchain technologies' rather than bitcoin because they don't like bitcoin's transparency.
Here's the problem. Bitcoin is secure and fast because of it's transparency.
Much faster than ACH payments, wire tranfers.
I realize I'm starting to sound like a crackpot so I'll just recommend the writings of a more well-respected crackpot Dee Hock (founder of Visa Inc.)
https://www.ratical.org/many_worlds/ChaordicOrg.pdf
These numbers will not change until the blocksize is increased, which still looks unlikely.
https://blockchain.info/charts/n-transactions?timespan=all
It'll be interesting to see what happens when(if?) Bitcoin upgrades to SegWit. This infrastructure is designed to support off-chain transactions while settling on the chain.
Hasn't come out yet, but they're clearly a boring company that isn't chasing trends. We'll know in a year or so whether this is all smoke.
I believe this is the timeless state of all big Blockchain projects: ready one year from now.
I specifically asked for success stories because I know everyone is throwing money after blockchain stuff. So I don't doubt that there are tons of planned projects. However, I haven't yet heard of blockchain technology being applied to solve an actual problem, except Bitcoin.
CME Group and the UK Royal Mint are planning to launch a blockchain-based product [1] -- also in 2017. But according to the article, it seems to be a simple digital promissory note, with the bearer owning the right to a certain amount of gold. Since we already have a central creditor, upon which the entire system depends (the Royal Mint), there is zero reason to use a distributed database to store anything.
Just put the database in the same safe as the gold.
[1] http://www.reuters.com/article/us-gold-blockchain-royal-mint...
It is possible there's just too much hype to find the existing use cases, that's one of the weaknesses of Google.
So you might have as well said "Signal".
STORJ - (storage in a southern dialect combined with startup lingo) - A Counterparty asset that is used as fuel for a decentralized unlimited cloud storage platform. Farmers offer storage, consumers upload shards of their files to a variety of different farmers. Encrypted, unlimited, redundant. I believe the STORJ company now has VC funding as well.
OMNI - create and register assets. Create secondary markets for those assets. (Counterparty and OMNI compete for market share)
Tether - An OMNI asset which is used to create stable value cryptocurrencies that are representations of national currencies. The Tether company has a centralized account of USD and EUROs, but this enables people to hold stable value cryptocurrencies, or "cash out", but not have to worry about the long delay of getting national currency from a bank. Primarily used for trading.
All of the above metalayers create transactions on the bitcoin network, and settle on that blockchain, so if that disqualifies them all because they use bitcoin underneath, then Dogecoin has a Counterparty layer as well :)
My point was precisely that you cannot, as far as I can see, separate blockchain technology from Bitcoin. My argument is that it's not blockchain technology that's useful, but rather a blockchain, in itself, that's useful. And it's useful because it's a distributed, proof-of-work timestamp server.
The value of Bitcoin comes not from the technology it uses, but from its incentive structure. An increase in the value of bitcoins feeds back into the system, resulting in an increase in security (hash rate) by way of an increase in the profitability of mining. Since Bitcoin's value comes from its time-stamping ability, and since this time-stamping is done using proof-of-work, it makes sense for an increase in the hash rate to also cause an increase in value, at least up to a certain point (provided no changes to the protocol are made to take advantage of the enormous hash rate of the current network).
Your question was vague.
"OCBC Bank said Monday (Nov 14) it became the first bank in Southeast Asia to use blockchain technology to make local and cross-border inter-bank fund transfers."
Textbook use case for blockchain IMO.
It's always marketing speak, with reference to all the wonderful properties of "blockchain technology", and diagrams like [1], which could have simply been summed up in three words: "no payment intermediary". Although... doesn't that diagram look like they just changed the intermediary's name from "Payment Intermediaries" to "BCSIS Blockchain platform", plus added some cool-looking server icons?
But the source code will almost certainly never see the light of day, so we will never really know how they use blockchain technology. At the end of the day, I think it's a lot more likely that this bank's software provider created a new solution and just said it has "blockchain technology" because it's a vague buzzword, and then the bank released a press release based on this, again because of the buzzword effect.
[1] http://www.straitstimes.com/sites/default/files/blockchain.p...
Not paying money to Accenture could also save billions.
the monopoly on credit has to end.
2028? 2035? what year you guys think that'll happen
Bank: We don't need another database
Accenture: Did I mention it's powered by blockchain
Bank: Shut up and take my money.
Somehow, I just found a charge for $3500 from Accenture on my credit report. It HAS to be from when I worked there. They either took back a bonus or thought that they paid me too much I guess. It's been years, I have never heard anything about it, and now my credit took a hit. Worst job/company ever would be an understatement. I quit in record time (<60 days).
[1] https://www.accenture.com/t20150523T115642__w__/us-en/_acnme...