In last week’s post, I promised, this week, to talk about Blockchain. I also asked for feedback if anyone thought my notions about Bitcoin were off target. I can tell you, that with nearly a thousand views, no one has taken me up on that offer. If they do, I will let you know. Now on to Blockchain.
Blockchain is probably best known as the technology that underlies Bitcoin. I am not going to review how Blockchain works, there are plenty of resources (one good high-level resource: https://en.wikipedia.org/wiki/Blockchain ) that can explain that. I am going to focus on what it can do, why it is appealing and why it isn’t. That said, what is it? It is:
- A public ledger (current transactions and complete history) that records transactions;
- A secure repository for transactions, i.e. once a transaction is posted to it, it is very difficult to corrupt the record;
- A mechanism through which two parties who do not trust each other can complete a transaction, have it recorded (and visible to all players) and ensure that neither party can renege on it.
These characteristics make it a very viable mechanism to implement a variety of systems (like digital currency, registering car titles or recording real estate transactions) that require a trusted system that is accessible, affordable and secure. It also makes it a good foundation for other processing, like smart contracts.
One of blockchain’s defining attributes is that it cuts out the middleman (the trusted third party) who normally acts as the intermediary and sometimes arbiter if transactions go wrong. Several examples of trusted third parties include:
- The Federal Reserve—who clears transactions between banks;
- Credit Card Companies (Visa, Mastercard, American Express)—who handle payments from customers and to vendors. By the way, credit card vendors are a good example of third parties who intervene if transactions go wrong. You know this if you have ever contested a charge (which you can for a variety of reasons) to your card;
- State Governments (US)—who manage the titling and title change process that records who owns a motor vehicle; and
- County Governments (US)—who manage titling of real estate and the transfer thereof.
Each of these intermediaries charge for the work they perform, and some folks argue that they charge too much. Blockchain is a way of disintermediating them. The blockchain technology replaces trusted third parties with a network of computers that record and secure a record of the transaction including all relevant (depending on the type of transaction) information. The mechanics of a blockchain transaction (if implemented properly) make it very unlikely that the details of the transaction will be wrong or that anyone will be able to change or corrupt the details of the transaction. But to be clear, there is a substantial infrastructure (and the associated costs) necessary to process, store and make completed transactions available to its participants, i.e. the mechanism is not a free good.
That said, it sounds like there are a lot of reasons that blockchain should be considered in a variety of applications and to an extent, it should. The question one must ask is: Why is blockchain better than the alternative(s)? Let me give you two examples, one where it (to me) makes (some) sense and one where it doesn’t. These examples, used as models of situations where blockchain makes sense and where it doesn’t might help you test and decide whether your potential application for blockchain does.
Recording Deeds and Real Estate Transactions
This example applies equally well to car titles. Here you use the blockchain technology to record who owns what and when a change of ownership happens who was involved and the relevant (like how much was paid) details of the transaction. The argument is that the blockchain will provide not only the record of transactions, but also a history of said transactions. It is also argued that the blockchain technology would record transactions in a way that would make unauthorized modifications or corruptions of the data (nearly) impossible. All of this is true, but is it really better than the alternative?
The trusted third party, in this case, is the government. And, one might argue that anything is better than the government (in terms of efficiency and security) managing something. On that point, I tend to agree. But the question here is not government vs. private, but trusted third party vs. distributed ledger. And, here the question of which is better is (as in most cases with blockchain) clear, but up for discussion.
If we are going to implement a new system to process title changes, the arguments for a distributed ledger are more straightforward. The mechanics of the blockchain would make the status and history of title changes available to all the users of the system and the security (ensuring the integrity of the records) would be ensured. I note that this does not mean that fraud could not be perpetrated, but that it would be unlikely that it would involve changes to the title records after they had been recorded to the system. So, on balance, one could make a reasonably (but not slam dunk if one considers a complete, including cost and the context, of the implementation) argument for blockchain.
Clearing Funds between Banks
Let’s say you are a bank and you use the Federal Reserve to process you interbank transfers. You might reason that the Fed is taking a lot of my money to process these transactions. And, other banks would probably agree with you. These banks might conclude that blockchain would be a good way to cut the Fed out of this process and thereby lower the cost of clearing funds. They could at least hope that the threat of an alternative (lower cost) service might induce the Fed to lower its price for clearing funds. There are two flaws with this thought process:
- That the cost (as opposed to what is charged) of a blockchain implementation would be lower than what the Fed (if they really put their mind to it) could possibly achieve. It isn’t clear to me that the cost of a blockchain implementation (distributed ledger) is any less than that of the trusted 3rd There is a lot of infrastructure necessary to support the clearing function, even with blockchain and the participating banks would likely be the parties responsible for setting up and operating that infrastructure.
That said, one could make the argument that: 1) the Fed might be capable of lowering the cost of processing the transactions, but is not likely to—I will explain this in the next bullet point; and 2) that it is easier (and cheaper) to replace the trusted third-party technology with blockchain than to replace the trusted third-party infrastructure. This might be true, but as I argued above, it might not;
- That what the Fed charges banks will go away if the blockchain is implemented and the Fed’s clearing service shut down. I would argue that the Fed would find some other way (charging for clearing is just a mechanism for collecting its operating expenses (other than the direct costs associated with clearing funds) and would find another way of levying its fees.
So, in this case there might be an argument that the cost to the banks of clearing funds might go down, but I would argue less than you might think it would, the Fed still needs to fund its operations. In any event, the threat of a blockchain implementation might (a big might) induce the Fed, my arguments notwithstanding, to lower its processing charges. So, in that context, threatening to implement blockchain might make some sense, but not otherwise.
So, is the blockchain technology all that much better than a regular database for storing records? I will concede that it does have some important security mechanisms built in from the start. This is a significant advantage and one that shouldn’t be discounted. So, I give blockchain a real edge on that one. At the same time, I would note that I have (with more than a little research under my belt) not been able to find a single instance where either a property or car title was compromised in situ using current technology. I would also note that I am skeptical that bank fraud that involved malevolent players electronically pulling funds out of bank accounts would have been prevented by the use of blockchain, i.e. the blockchain is good at preventing some kinds of fraud (i.e. those involving modifications to the record of a transaction), but not so good at others.
In thinking about this (and I know I am biased here), I have considered whether Textura’s payment application CPM might have been better implemented in blockchain than a proprietary database with Textura being the trusted third party. While I can imagine marketing benefits to promoting a blockchain implementation in sales materials, I can’t make a good business or technical case for using a distributed ledger over the open source (but traditional) database that we implemented. If we were doing it again, I would consider blockchain, but would only choose it if there was a clear business case (for both Textura and its customers) for the blockchain alternative.
Further, blockchain does require a significant infrastructure to make it work and it (generally) assumes that that infrastructure is in place and optimized for the type of transactions that are being processed. More work needs to be done to determine how well, under load, blockchain implementations hold up and whether a generalized implementation will serve different kinds of transactions (at volume) than any other general-purpose computer architecture.
Finally, much is made of blockchain’s “invulnerability” to hacking and that this means that fraud will go away. While it might be difficult if not impossible to change the records that have been committed to a blockchain (making the transaction well documented and trusted), I can imagine lots of fraud that could take place around fortress blockchain. For instance, bogus transactions (where no goods are delivered) could still be generated in the furtherance of a money laundering scheme. And, there has been plenty of fraud around digital currency. Most of this fraud has been around the edges (in the repositories where their private keys are being held) and lots of digital coin has been stolen from people who thought they were holding a secure product. So, dreams of auditors and other third-party guarantors being out of work is just that, dreams.
My purpose here (unlike my goal in my previous, Bitcoin, post) was not to say that blockchain isn’t fit for purpose, but to suggest that it isn’t the clear choice some commentators say it is and should be examined very carefully before making a decision to implement. And, I would be interested in feedback if anyone wants to pick a bone with the thoughts I have presented on blockchain.
Copyright 2017 Howard Niden
— you can find this (days earlier) and other posts at www.niden.com
and if you like this post: 1) please let me know; and 2) pass on your “find” to others.