Why blockchain won’t fix payments

Why blockchain won’t fix payments

There is a misconception about blockchain in the industry surrounding the belief that it is a solution to making faster and securer payments. There are some issues around the blockchain that explain why, in its present form, it isn’t an ideal replacement.

For example, the Bank of Canada, the country’s central bank, has discovered that blockchain is not presently an option for underpinning payments.

A July 2017 article by Out-Law quotes the Bank of Canada: “Underpinning wholesale payment systems with distributed ledger technology (DLT) would introduce greater costs and risks for institutions than those which apply under existing wholesale payment systems.”

The article continues: “A study carried out by the Bank of Canada, into the feasibility of using DLT to create new distributed wholesale payment systems, identified potential for DLT-based wholesale payment systems to deliver benefits if they could be linked in to other financial market infrastructure.”

“Such benefits may be obtained by integrating other assets on the same ledger as payments – which could greatly simplify collateral pledging and asset sales – reaping economies of scope and reducing costs to participants by integrating back-office systems,” the Bank of Canada reported in response to its study findings.

Blockchain immaturity

A September 2017 article by Reuters, comments that DLT/blockchain won’t consequently be able to power the world’s biggest payments systems, though both the European Central Bank (ECB) and the Bank of Japan (BOJ) were quoted to state that “it has the potential to improve system resilience”.

The reality of blockchain supports the findings of the US Federal Reserve and the Bank of Canada. In one speech on DLT, the Fed says there are concerns about smart contracts:

“Although the idea of automating certain aspects of contracts is not new, and banks do some of this today, the potential introduction of smart contracts does raise several issues for consideration. For example, what is the legal status of a smart contract, which is written in code? Would consumers and businesses rely on smart contracts to perform certain services traditionally done by their banks or other intermediaries? Could the widespread automated interaction of multiple counterparties lead to any unwanted dynamics for financial markets? These and other considerations will be important factors in determining the extent of the application of smart contracts.”

The problem, according to CEO of Kalypton, Lars Davies, is that many people are trying to separate Bitcoin from blockchain. Bitcoin is often perceived as being bad, while Blockchain is perceived as good.

“Bitcoin and blockchain are intrinsically linked, they are symbiotic,” says Davies. “People assume that it is a valid solution, but there is very little analysis. What problem does it solve? They are often spoken about in separate ways and there is an assumption that blockchain is useful.”

Rather than being useful, “it’s a self-feeding frenzy, which economists refer to as an information cascade”, he says. “As so much has been written about blockchain, surely it must be a solution? But what is it a solution to? There is far too much hype about blockchain, and next to no real analysis about what it can actually solve.”

R3 and Ripple

“Look at the dispute between R3 and Ripple. Here we have two companies that aspire to provide an enterprise grade solution to financial services infrastructure but that means burning through cash like crazy trying to make blockchain work.”

Davies adds: “If it weren’t for the vast amounts of supposed gains from cryptocurrency speculation, the blockchain probably wouldn’t survive. The heart of the dispute is that Ripple is trying to stop R3 from profiting from options ‘worth’ $1 billion it holds on Ripple’s in-house currency XRP. This is a classic ‘misdirect’ trick. Nothing up our sleeves.”

So why do people believe that blockchain is good and Bitcoin is bad? Indeed, why do people support blockchain and think that it will work? Davies believes this is because everybody says blockchain is good, and so they don’t question or analyse it appropriately to ensure that it is the right solution for payments. In essence, people believe what they want to believe – even in an industry such as banking and financial services, where due diligence and regulatory compliance are crucial.

“It’s similar to what sometimes happens when there are multiple witnesses to an incident. For example, each witness assumes that somebody else has reported the incident and so they don’t call the police,” Davies explains. He adds that the same bandwagon applies to blockchain. “There’s so much noise around blockchain that people jump on it, assuming that someone must have validated the technology. The information cascade becomes the conventional mode of thought.”

Irrational hype

In his opinion, the hype around blockchain is irrational. “Nobody has analysed what it can or can’t do, but people have rationalised utilising it as being safety in numbers”, he says. The reasoning behind his thoughts are founded on a citation of economist John Maynard Keynes, who once said: “Worldly wisdom teaches that it is better for reputation to fail conventionally than to succeed unconventionally.” So, by being part of the blockchain hype, if it fails you’ve failed conventionally and so your job is safe.

Davies elaborates: “A similar thing happened with the CDS market, and that happens with every bubble. Central banks are now saying that blockchain is not the solution, but so much cash is being burnt supporting it because to say otherwise is to be unconventional. In the minds of the participants, so long as they attach themselves to the convention they cannot be fired if they are wrong; everyone else is doing the same thing.”

He adds that the dispute between R3 and Ripple is insignificant: If two companies are quibbling over options about an asset that has no intrinsic value, then leave them to it. Bitcoin and cyber-currencies are no more than programming code. Bitcoin has no value except what someone is willing to ascribe to it. Bitcoin’s nominal value is zero. It has no collateral behind it, and no intrinsic value. At least you could plant a tulip bulb and have a reasonable chance of growing a tulip.”

Gold Standard

He then explains that bitcoin books offer support for the cryptocurrency by citing the Gold Standard without understanding how it works. “The Gold Standard measured currencies against a defined weight of gold. Gold has a value, and everybody agreed that it was the commodity to use. That’s because gold, or some other commodity, has value as a commodity you can ascribe value to it.”

Compared to cryptocurrencies like Bitcoin, fiat currencies such as the dollar, pound sterling, and the euro “have a value ascribed to them based partly on how well the economy behind the currency is performing, and the record of the government”, Davies comments. Today, the fiat currencies are no longer pegged against gold. He says this is because the Gold Standard became hard to manage.

“During times of war the UK government suspended the Gold Standard because that allowed it to deflate the pound and to introduce inflation to pay down its debt. This is why governments like inflation, which is a result of policy and little else”, he says, suggesting that this leads to a debate about monetary policy.

He elaborates on the reason this arises: “Libertarians believe that cryptocurrencies prevent governments from interfering with the value of the currency. The idea was that the currency should be free of government control. This is a misreading of why governments control currency. Governments are best positioned to protect the value of currency by preventing counterfeiting and by making sure that supply doesn’t grow out of control.”

“The supply of money has to grow to allow an economy to grow, but bitcoin has a limited, finite supply”, he explains before adding: “When the supply of money grows, the economy grows. It is true that if you have a fiat currency, then there can a temptation to print currency. This can decimate the value of your currency. It depends on how the government or central bank manages monetary supply. Governments can choose to have a hard fiat currency; it all comes down to how the government and its central bank chooses to manage that currency.”

Market fluctuations

In Davies’ view, the fluctuations in the market show that cryptocurrencies aren’t suitable for payments systems. “They don’t have the stability; they are not a means of exchange”, he believes. So, to gain value from any cryptocurrency, there is a need to cash in immediately. Unlike the pound, for example, which is a means of exchange, cryptocurrencies won’t hold their value. In contrast he suggests that “£100 today will still be worth £100 tomorrow.”

“Bitcoin does not have this feature, and that’s one of the reasons why the courts are saying that bitcoin is not money because it is not a means of exchange, and that’s why even the Bitcoin Foundation itself used this argument in its recent letter to the US Senate Judiciary Committee.” Davies elaborates: “There is no understanding of what money really is and what its function is. There are, quite simply, too many issues in the design of bitcoin and other cryptocurrencies that cannot be solved.”

Security questions

There are questions about the security of cryptocurrencies. The misconception is that blockchain and subsequently cryptocurrencies are fundamentally secure when they aren’t. As programming code forms their basis, they can be hacked. They are therefore inherently insecure. So, can they be developed to make them secure and free from hackers? According to Davies, they can’t.

He gives his reasons why: “The very nature of DLT means it can’t be made more secure. It’s design assumptions prevent you from changing the way it operates. If you do, it breaks. Ideally, it’s programmed to have as many nodes as possible with the assumption that the more nodes you have, the more secure the blockchain or distributed ledger will be. But investigators and others can follow it. If you look at GDPR, pseudo-anonymity is not by itself sufficient.”

In contrast to the many pundits promoting blockchain and cryptocurrencies, Davies finds that central banks understand what money is. So why are there some reports that they are embracing blockchain and cryptocurrencies? “They were compelled to analyse DLTs such as blockchain, but found that it can’t be used for payment systems”, claims Davies.

“Some argue that you might be able to use it across borders if you follow the existing method of having multiple parties within the transactional process; this would lengthen the time it would take to process a transaction to get around the fact that DLTs simply cannot process transactions in real-time”, he explains.

The problem is that he finds that even this use case doesn’t stand up to scrutiny. “A digital payments system, even one that managed cross-border and cross-currency payments, would avoid the need for such intermediate steps if it was designed properly, and so settle in near real-time: This is something that DLTs cannot support”, he comments.

Monolithic architectures

In his experience, “Blockchain and monolithic architectures do not scale well and this is the reason central banks are abandoning it. You just can’t scale DLT and blockchains. Another issue is causality; DLT and blockchain can’t prove causality. For payments, you need causality. So, they are unsuited for payments.”

Davies then points out that one of the problems is that authentication takes place after the event. This means that blockchain can only show what was submitted to the system.

Authentication issues

“One of the design assumptions is that you can trust both parties to issue a true record of the transaction, but there is no guarantee that record is a true record of the transaction”, he comments. Cryptocurrency miners must do the proof of work, which is used to limit the number of forks that can occur when miners try to create a block. “But that does not prove the truth of the records themselves. The proof of work is simply that, it proves that they performed a calculation or solved a problem, nothing more,” he says.

“Each miner will select transactions form the pool waiting to be added to the blockchain (they won’t necessarily all choose the same transactions – more on this later). The miners then each solve the proof of work problem and the first to do so will have a block and add that block to the block chain. The other miners will then stop attempting to generate a block, and add any transactions not in the successful block back to the pool (this is one of the reasons why the blockchain cannot show causality).

“If two or more miners solve the proof of work then each will add their block to the blockchain, and the blockchain will fork (the number of forks will depend on the number of simultaneous blocks; usually no more than two but there can be more).”

The process then requires the miners to generate new blocks as before, and then they add them to their forks. This is then repeated until one fork becomes longer than the others.

“This longer blockchain becomes the blockchain, and the other forks are discarded (any transactions that were processed in these forks that are not in the accepted blockchain are added back to the pool of unprocessed transactions)”, he explains.

The key issue is that all the proof of work shows that the miners have exerted their effort to solve the problem, but it fails to prove that the transactions are valid. “That latter issue comes down to the assumption that the parties will submit a true record of the transactions,” Davies says. “If a transaction is recorded in a block more than five or six from the top of the blockchain, then users can regard that transaction as irrevocable, as it is very unlikely to be discarded later; forks rarely last for more than two or three blocks. However, that still does not prove the truth of the record.”

Not for payments

He therefore argues: “Bitcoin and blockchain were never designed to fix payments. They simply will not work in any meaningful way. The security of blockchain depends on having multiple ledgers. Blockchain can only manage seven to 14 transactions per second. The assumption is that if you make it harder to change the blocks but increasing the number of copies of the blockchain or ledger, it becomes harder to hack. However, there are fewer and fewer miners, as the cost of mining a block continues to increase, meaning that this assumption has become a security flaw.

“If you want to increase this throughput, and you can do so, then you have to sacrifice other aspects of the design and so accept that you will have less security, less distribution, etc. You can get a high throughput with two or three nodes or ledgers, but then why bother? You no longer have the presumed security of presenting multiple nodes that an attacker would need to compromise.

“However, even that security model is suspect as various security researchers have shown. Others assume that you can do away with the proof of work and use a consensus protocol to synchronise ledgers. That is not a solution as the performance starts to degrade as you increase the number of nodes.”

Regulatory compliance means that banks need to comply with know your customer (KYC) regulations. To achieve this, they need to have access and to include all the contextual information in the record of a transfer or a payment. The trouble is, as Davies argues, blockchain can’t store all the required information because the blockchain is not designed to do so. It is a log and not a database.

So, with regards to GDPR as an example, if you have consumers involved with the transactions, Davies believes there may be a need to “expunge the record of the false transaction, but you can’t do this with blockchain without breaking the audit trail. This is why blockchain is completely unsuited.”

He concludes that the hype around blockchain is about fear of not being part of the convention. This pursuit could be disastrous. Both bitcoin and blockchain therefore deserve more scrutiny – particularly with regards to payments.

By Graham Jarvis, freelance business and technology journalist 

Comments
  • CK 7 November, 2017 at 1148

    Just read the part about R3 and Ripple, very uninformed. The author obviously has no clue what’s going on. We’re not just talking greed here, we’re talking fraudulent maneuvers by R3 by promising to bring in clients for Ripple under the XRP options contract while the moment the agreement was signed major clients left R3 and no benefits for Ripple substantiated.

    Research first, then write.

  • baker 7 November, 2017 at 1316

    Monolithic architectures:
    “You just can’t scale DLT and blockchains.”

    huh, why not? just because the problem is hard does not mean you can’t do it. Sharding + POS will help scale ethereum. For Ripple this has not been a priority because they can already handle 1000 tps.

    “Another issue is causality; DLT and blockchain can’t prove causality. For payments, you need causality. So, they are unsuited for payments.”

    -What does this mean? If the transaction is being settled through FX markets I can guarantee it can be tracked and documented all in 3 seconds. Its not some hodepodge black hole of mystery.

    “Davies then points out that one of the problems is that authentication takes place after the event. This means that blockchain can only show what was submitted to the system.”
    Is there some tech that allows you to see something before it enters the system, besides SKYNET.

  • Garbage 7 November, 2017 at 2043

    LOL – Ripple has already changed the way banks transact for over 100 financial institutions. This article may have been okay five years ago, but today, FIntech, specifically, blockchain, has already rewritten the book!

  • Nancy Sakamoto 8 November, 2017 at 2144

    The fact that many,many banks have done proof of concpets with Ripple does not mean it has “re-written the book”. In fact they haven’t really proved any concepts. What happens in the typical Ripple PoC, is that one or more banks set up a private version of the Ripple blockchain. One of the banks (or one of the bank’s branches, if an individual bank) acts as the nostro agent for the funds. The banks or bank branches then use Ripple as an overcomplicated messaging system to instruct the transfer of cash between accounts. You could do the same using SWIFT, fax, telephone, email or messenger pigeon. It proves nothing.

  • Post a comment

    Threaded commenting powered by interconnect/it code.

@banking
techno