Jamie Dimon - Your Big Bank can be the Big Winner in Crypto-Currency

2/1/2018

Jamie Dimon, Chairman CEO and President
JPMorgan Chase and Co.
270 Park Ave.
New York, NY 1007

Hi Jamie,

I am reaching out to tell you about an idea that I just can't keep to myself any longer.  This idea does me no good unless I share it, so I am passing it along to you, free of charge.  Most people would think it is pointless for me to bring this idea to you.  You famously called Bitcoin a fraud, but I'm not buying it.  I refuse to take this statement at face value and assume you, and your company, are not believers in crypto currency.  You have one of the brightest minds in the financial sector working for you and leading your very own blockchain program, Amber Baldet. Quorum, JPMorgan's blockchain project, is interesting, but limited and flawed like all blockchain based ledgers.  Blockchain is already obsolete.  It's time you move on to the next generation of distributed ledger technology, Hashgraph.

Before we go any farther lets clarify the difference between a crypto currency and a distributed ledger technology (DLT).  A DLT is an underlying platform or layer that provides consensus in information sharing.  Blockchain is a form of DLT.  A crypto currency is simply an app or program that runs on this technology. Crypto currencies are not the only types of apps that can run on the technology.  DLTs can be used in more ways than anyone fully understands.  Theoretical uses include everything from registering titles, to online gaming in massive multiplayer worlds, to replacements for existing payment systems.  The DLT creates a layer of trust and consensus that everyone using the program relies on to keep the system/game/app fair.  This trust is what has been missing from the internet since its inception.  Adding this layer will revolutionize the way we communicate and will certainly change the world.

This trust layer requires decentralization.  You give away control of information.  Information in the crypto currency context is a database of transactions. Giving up control of the database and making the community validate the authenticity of the database is the core belief behind a decentralized ledger.  This is meant to guarantee fairness by sharing data.  In this way, a malicious attacker cannot attack a keeper of information.  The lack of control by one party gives the community an opportunity to protect the shared information by validating truth and making an attacker visible as an error in the information.  You cannot hack the information if it is everywhere.  The third party that keeps the information is the community itself and the security of every individual is impossible to break.  In theory.

What follows is an extremely oversimplified explanation of blockchain.  In a blockchain DLT, transactions are recorded in blocks.  These blocks are created and validated by "miners" or validators.  Miners all run a common program that includes a consensus algorithm. The algorithm is used to define a winner in the mining game.  The miner takes a group of transactions and puts them into a block.  They run the algorithm at the same time as all the other miners in the system who have also constructed blocks of transactions.  This algorithm becomes more and more difficult to solve over time and requires ever increasing amounts of computing power (and electricity) to solve the problem.  Once the problem is solved, that miner's block of transactions are added to the chain of previous blocks and the entire network is given a copy of the updated chain of validated blocks.  In a crypto currency blockchain, the miner may be rewarded a number of units of the crypto currency as a reward for solving the problem.  This is the incentive the miners have to be in the business of validating transactions and expending the computing power required to solve the problem in the algorithm.

We could go much deeper into this explanation.  Miners are only one type of node, or full user who contributes to the blockchain.  The example I gave is a proof of work, blockchain, distributed ledger technology.  There are also proof of stake, blockchain, DLTs.  Each have their own risks, rewards, pros, cons, etc.  The merits of these two competitors, proof of work and proof of stake, are hotly debated in the crypto currency and DLT communities, but they share a few common flaws:

    1. Computing power and electricity consumption are growing at an unsustainable rate.  We don't have to get into the details of this, but one Bitcoin transaction consumes more than 5000 times as much energy as a traditional electronic transaction and this number increases every day.  Proof of stake systems are being designed to control this cost, but it is still much greater than traditional transactions.

    2. Not scalable. The current blockchain technologies only allow for a few transactions per second (TPS).  A "few" is roughly 5 for Bitcoin and 20ish for Ethereum.  Visa is said to be able to handle 56,000 TPS but routinely runs around 2,000.  Blockchain technology is forced to run slow to reduce the risk of errors or duplicates showing up in the system.  Sometimes the miners "tie" when solving a problem and the system forks into two separate chains.  This error is corrected by additional mining, but this wastes time and power and limits the scalability of the system.

    3. Asynchronous Byzantine Fault Tolerance (ABFT). This is a fancy way of saying, "How certain are we that we all have the right information and are in consensus?"  Blockchains are not truly ABFT. They get very close, and that's good enough for most advocates of blockchain, but they don't quite get there.  ABFT is the gold standard.  Any system that is ABFT would have to be considered a security and reliability improvement over a non-ABFT system.

    4. Incentives and fairness.  A system could be considered fair if all users receive equitable rights, privileges, and service regardless of location or status.  Blockchains are already confronting issues that are leading many of these technologies to use less fair methods of calculation and inclusion to overcome other limitations of the system.

    5. Cartels. If mining is disproportionally weighted and controlled by a few major players, as is the case currently for both Bitcoin and Ethereum, then these dominant players can get together and control the system.  This kills the idea of decentralization and creates a very dangerous situation for all the users of the system.

    6. Quantum computing defense.  The existing cryptography of every blockchain solution is hopelessly defenseless against threats that will arise from quantum computing.  

Advocates of blockchain will tell you that all of these issues are just technical puzzles that will be solved.  They will throw out phrases like sharding, sidechains, benevolent dictators, validators, state channels, off chain computations, cartel monitoring, et al.  The solution I propose is to get away from blocks and chains and instead use a different type of DLT, a directed acrylic graph or DAG.

DAGs are a way for all the users of a system to communicate with each other and verify work as they go along.  I have heard this described using a classroom as an example.  Rather than turning in your work to a teacher to grade, the teacher lets the class exchange their work with another student.  They grade the work themselves and the results are then communicated back to the teacher.  A DAG can be set up so that if I want to submit a transaction, I have to grade the work of other transactions at the same time.  When I verify that work I include that information with my transaction.  As my transaction is verified by the next user in the system they receive my verifications from the previous transaction and pass along that information as well when they make their next transaction. This provides verified information to the network in virtual real time.  This is also referred to as tangle by one of the DAGs currently in the market.

Hashgraph takes this idea and builds even more improvements on top of it.  There are two key features that make it superior to other DAGs and blockchain; gossip protocol and voting protocol.  In a distributed ledger you can reach consensus by telling someone everything you know and then telling another person everything you know and everything you learned from the last person you spoke to.  Then you ask everyone in the system to vote on what is known and whether it is accurate. The Hashgraph does this with a voting algorithm that knows before you respond how you will vote.  The creator of the Hashgraph, Leemon Baird, explains this as "kinda weird," but it is simply gossip about gossip and a voting algorithm that knows what your vote will be.  This voting algorithm can assume your vote because of the information the algorithm has received from the network.  If you are running a node on the Hashgraph you reach consensus in this way with tremendously high throughput, very low latency, with asynchronous Byzantine fault tolerance.   The Hashgraph can achieve throughput up to hundreds of thousands of transactions per second without sharding or sidechains or any of the other gimmicks needed to achieve these speeds on a blockchain.

So, what is the downside? Hashgraph is not a public ledger.  It is patented and privately owned by a company call Swirlds.  Swirlds was founded by Leemon Baird and Mance Harmon.  Swirlds brought Hashgraph out of "stealth" in 2017.  They are a small startup tech company with venture capital backing including a seed round in September 2017 that brought in $3million in new capital.  Their entire company is built around Hashgraph and leasing the technology to organizations that want to build programs/applications that require security and consensus in a shared world.  Their largest success to date was beating IBMs Hyperledger to secure a contract with over 5000 credit unions that built a private system to share information amongst themselves.  

If Hashgraph is the technology with the competitive advantage in the distributed ledger race, how do you get your hands on it?  I think these are your four options:

    1.  Lease from Swirlds
    2.  Buy Swirlds
    3.  Hijack Hashgraph with improvement patents
    4.  Create a clone

Leasing from Swirlds is the simplest way to go, but I don't think it's the right answer for you.  You could build and own some very exciting programs like:

    •    Faster more secure payment systems
    •    A crypto currency that could be used to crowd out many of the other currencies - a crypto killer
    •    A title company that greatly automates the process of recording deeds/liens/etc
    •    Create smart contracts to start a new era of derivatives trading
    •    Create your own stock market

The problem with leasing the tech is you would not really control your future and you would lose a slice of the pie that may be a small percentage, but it could be a small percentage of a very big pie.  I think you could easily pull off one of the other options and eliminate this cost.

Buying Swirlds is a pretty straight forward idea.  The cofounders, Baird and Harmon, are not millennial dreamers with grand ideas to bring down the system and replace it with a utopian free currency.  Trust me, you find a lot of these people in this space and they would never sell to JPMorgan.  Instead, Baird and Harmon are seasoned veterans in the tech sector.  They both worked for the military.  They have both founded and sold companies in the past.  There is also VC money in the company and everyone knows that those folks are looking for 10x+ on their money and a smooth, quick exit.  I'm confident a purchase would be possible, but at what valuation is difficult to assess as an outsider.  It may be $30million or it could be $300million.

Hijacking Hashgraph would be possible if you pursue improvement patents over the current technology.  Swirlds offers a software development kit (SDK) for Hashgraph.  The end user agreement would prevent a user from trying to reverse engineer the software and build on it for their own purposes, but I'm sure it would be easy enough to find a copy of the source code for the demo apps in the SDK without signing up as an end user.  This would give you a good jumping off point.  You could easily employ a few software developers to get into this code and find algorithm improvements and efficiencies that could enhance the technology to a point where you could secure improvement patents.  You may face legal battles with Swirlds if you choose this approach, so a cost/benefit analysis would be necessary before choosing this over a Swirlds purchase.

Finally, you could clone Hashgraph by creating your very own DAG with voting and gossip protocol.   Even Baird admits these are both old ideas that have been brought out of the cupboard to create this new software.  This would be the safest way to control the tech, but you do run the risk of creating a subpar DAG and losing out to Hashgraph.  Swirlds is going to continue to improve their software and the number of developer hours working on the project is growing every day.  I would only consider this a last resort.  You have plenty of capital, financial and intellectual, to buy or improve Hashgraph without trying to start from scratch.

I think that's about as far as I can take this as an outsider.  In reality, I would want to work with Amber and her existing team to hash (pun intended) out the best option for JPM based on what priorities you have attached to this project.  

I can't say with any certainty how serious you are about distributed ledger technology, but you are a smart man and you have an incredible responsibility and opportunity in front of you.  If you deploy a strategy that gives you even a miniscule advantage over your competitors, the leverage that you will have is impossible to overstate.  This is an arms race that may very well define our financial system for the next century.  Someone is going to win this race.  There may be multiple winners.  You cannot afford to be on the sidelines, tinkering with this idea, and let the tech move so far ahead of you that you become a consumer rather than an owner.

Now is the time to step up and become a leader in the most important fintech of our generation.  Thank you for the taking the time to read this letter.  Please reach out to me when you decide it is time to put JPMorgan Chase at the forefront of this revolution.

Sincerely,
Aaron Julian
ToTheBoss.com