Crypto is not that different from the Model T. Photographer: Jeff J Mitchell/Getty Images
Citigroup Inc. has just released a research report making a bull case for cryptocurrencies – an asset class that hasn’t had a lot to cheer about the last two years. The thesis is that blockchain and related technologies will grow to having a billion users and trillions of dollars in value over the next six to eight years. What’s missing is the implications for investors, which I’ll try to outline.
The 162-page report starts with the example of someone in 1900 predicting the eventual massive economic and cultural change caused by the automobile. It was not practical to invest directly. Most automobile start-ups at the time failed. Ford was not founded until 1903 and didn’t go public until 1956. Shorting buggy-whip companies sounds clever but it wasn’t any more practical than investing in car companies. Plus, buggy-whip companies had little trouble shifting to making similar consumer goods.
What would have made sense is realizing cars need gasoline, highways and gas stations, and would create demand for motels, drive-ins, suburban homes, rubber, glass, concrete and many other things. You didn’t have to bet on one car company or technology. Today there’s small reason to believe massive growth in the crypto-technology sector will increase the value of most existing crypto assets. A more promising investment approach is to think about what general technologies will be needed if crypto takes off.
The Citigroup report zeroes in on four key technologies. All of these exist today. Invention is not required, but the trick will be to deliver known technologies in the right package that becomes a widespread standard in the crypto ecosystem. (I am an active crypto investor, and he have venture capital investments and advisory relations with crypto companies including the types mentioned below.)
DECENTRALIZED DIGITIAL IDENTITIES:
Most people have partial digital identities smeared over the Internet in accounts at different sites and perhaps a master identity managed by Google or Microsoft. It’s little problem for hackers and identity thieves to link these together and compile much of your personal information. It’s also easy for people to fake or steal identities. On the other hand, it’s often difficult transferring information from one place to another, say to prove to a wine shipping site that you’re over 21 or to verify your identity on your Twitter account.
There are places that offer you the ability to create multiple secure digital identities that cannot be linked together. You might use one for work, one for shopping, one for friends, one for financial transactions and one for social media. Each one reveals only the information you want for that purpose, and all information is verified so others can trust it.
If crypto-technology expands, there’s little doubt that DDIs will be in great demand. The problems now are few individuals take the trouble to use them, and therefore few sites accept them, and therefore it’s difficult to verify most of the important information. But once the chicken-and-egg issue is overcome by rising user demand, it seems likely that a few big providers will dominate. Some might offer greater protection and be attractive to undocumented immigrants, adulterers and recluses. The big technology companies today could well get in the business, but likely with lower levels of privacy. Some DDI providers might specialize in the simplest interfaces, some might offer the most advanced features.
A friend of mine who was pursuing a doctorate in social work and a law degree griped that everything she learned in social work classes — gather all possible information about clients to provide the best help — was contradicted in law school: meticulously protect and segregate all information about clients. Zero-knowledge proofs are the solution. They give the benefit of collecting and provably summarizing information without revealing the components that went into that summary.
A major use case for ZKPs at the moment is cryptographic proof of solvency. An ancient problem that has popped up more than usual in the last 12 months is financial institutions — both traditional and crypto — suddenly revealed to be insolvent. One solution is more disclosure, but this can endanger client privacy and confidential business information. Moreover, few people have the ability to analyze even a simple financial institution’s books. Regulators and auditors can try to do the job, and to promise to keep all information confidential, but they are generally too slow and coarse to prevent the problems.
The solution is for the institution to run a ZKP algorithm on its own books. The ZKP app sees all the firm’s internal information, but it outputs a code that cannot be used by anyone to figure out the information. That code can be revealed to the public and proves the institution’s cash assets are greater than or equal to its liabilities — or any other desired proof, such as that market value of assets are at least 110% of liabilities. The institution cannot simply make up a code or feed false information to its ZKP app.
Crypto applications require real-world information: how old an individual is, what degrees she holds, what an Uber driver’s rating is, who owns a property, etc. Some of this information can come from authoritative sources, but much of it must be encoded in distributed public ledgers by private parties. Widespread use of crypto will require building vast numbers of secure, trusted oracles.
The weakest link in crypto are the bridges that connect different blockchains. Just as embezzlers concentrate on money transfers rather than the underlying accounts themselves, crypto hackers have often stolen money or information in transit between applications. The problem is not in fundamental technology; we know how to build secure bridges. The problem is cost. Proof-of-work networks require lots of expensive work, such as the electricity needed to secure Bitcoin. Proof-of-stake networks require large stakes. Validator networks need lots of validators, each of which are expensive to crack.
Today the resources necessary to maintain secure bridges are lacking. But the need for secure bridges goes up with the square of the number of major crypto applications, so a 100-times growth in the latter could mean a 10,000-fold growth in the former.
Personally, I have accepted the basic bull case Citigroup lays out for many years, but I have not bought a lot of crypto tokens. I prefer to invest in technologies that will be necessary whichever specific projects win or lose, including the four mentioned above.