High Margin Non Cyclical Finance Business Is Good
At a high level I talked about XRP as the effective TAM for financial cryptocurrency. Which is about $230B.
And Post Fiat is a bet that this TAM is probably on the low side bc XRP represents the transaction bank (SWIFT / Payments replacement). When we all know that crypto is primarily in the investment Bank. My personal favorite financial category are the Apps that are low risk money machines that make their users money. A few examples:
- Moody’s Analytics
- S and P Global
- Factset
- Tradeweb
- IHS Markit (now part of S&P - but I feel like it’s an incredibly clean go to market for an intelligence product)
- Palantir (financial apps)
Other financial categories such as payments, consumer credit, mortgage lending, and retail trading / exchanges are interesting but they tend to
- Take a lot of economic cyclical risk whether that is through defaults or trading volume (it tends to all hit the fan at the same time)
- Be extremely reliant on balance sheet which is cyclical and capital intensive
XRP at some level is a bet on making transaction banking less capital intensive. Which is an expansive. But extremely difficult TAM. The compliance risks in the transaciton bank - specifically SWIFT - are getting more intense, not less intense over time due to the rise of Tariffs and the Cold War between the US and China, along with geopolitcal escalation globally. Historically – HSBC, and Stan Chart have faced major problems here – and it makes them risk averse implementing technologies that are ‘SWIFT alternatives’. The spectre of multi billion dollar fines makes that a no go. Even with generous pilot economics with Ripple Labs. It’s not to say that’s impossible, it’s more to say - that’s a slog. Ultimately XRP is probably going to succeed in this Slog but it’s a glass chewing exercise on the way.
Whereas Post Fiat is a bet on blockchains being useful in investing contexts. In the current environment that’s a lot more appealing - where investment banks and private wealth arms are heavily interested in both crypto and AI, it’s far easier to interact with them than the Transaction Bank
The TAMs of Crypto in Financial Apps Don’t match The Real World
Then in terms of understanding Post Fiat at a category level - it’s less an Exchange Product (i.e. IBKR or CME)
The TAMs are worth discussing
Exchange and Brokerage TAM: If you’re a winning, major exchange in Tradfi - you can hit $100b – and trade at 25x forward EPS
- ICE: $102B
- CME: $102B
- HKEX: $72B
- NDAQ: $45B
If you’re a brokerage - $90B TAM and trade at 30x forward
- IBKR: $88B
- HOOD: $56B
- FUTU: $16B
Transaction Banks - 9x earnings, huge TAM
- HSBC: $208B
- Stan Chart: $40B
Pure Investment Banks - 13x earnings – similar TAM to transaciton banks
- Goldman Sachs: $184B
Analytics - 30x
- Moody’s (34x) - $86B
- S&P Global (29x) - $161B
- Factset (26x)- $18B
- Tradeweb (41x) - $31B
- IHS Markit (30x) - $44B
- Palantir (200x) - $300B
I think what’s quite notable here – is that if you’re an absolute world class analytics business you can have a $150B enterprise value (1/2 of Palantir – the commercial, or S&P Global), which is bigger than any publicly listed exchange. But in the crypto space we are trained to think exchanges are the best businesses. When the reality is that they’re smaller in overall market cap than analytics and trade at roughly the same multiple.
You can also see the difficulty of XRP - which is basically in its best case interacting with organizations that are very much not ’tech first’ and therefore trade at 9x earnings. This can be a conceptual advantage but in practice, there aren’t banks on its validator list showing how hard it is to execute there.
So the ‘contrarian and right’ in the crypto space is injecting blockchains into the Analytics business, with an AI overlay – which gets you
- TAMs that are as big or larger than investment banks
- The highest possible multiples
- a non cyclical, non path dependent expansion plan
Which brings us to the question - in what sense are blockchains useful in the financial AI era
I have 4 answers to this question
- Blockchains can facilitate T0 settlement of securities which lowers capital requirements for trading (as opposed to outrageous, complex, securities clearing requirements)
- Blockchains are a non centralized source of truth so are excellent for indexing purposes – which are currently owned by centralized entities like those that I listed above
- Because Blockchains can’t be deleted they provide a major source of value for compliance – because a. they enable time stamped playback b. they can be private and non deletable without relying on a central 3rd party such as Bloomberg that follows its own rules
- Blockchains are pseudonymous – and the people interacting with them are extremely tech savvy. Creating the basis for a very powerful expert network/ or research business
So What Drives Convergence
So let’s say we are going to a Post Fiat reality.
My basic premise is that everything in the Fiat world will more or less linear port to a decentralized, digital analog of that world and
- Roughly correspond to TAMs and multiples described above
- Have a bias to enhance Digital and AI based valuations / TAMs- for example, if we are going to a digital post fiat reality then commercial lending probably is a donut whereas AI analytics is probably extremely valuable (this is discussed extensively, even in CBDC publications from Central Banks)
This means that you clearly want to be applying blockchains to
- Indexing
- Market intelligence
- Analytics
- With an AI overlay
Because
- Nobody else in crypto is doing this effectively
- The TAM is insane
- There is strong demand for financially focused cryptocurrencies due to the increasing acceptance of crypto in banks (see: why XRP is worth what it is)
- There is a clear use case for blockchains in this regard
The Expert Network / AI convergence
One of the most interesting things about AI – something I learned interacting with the folks at Story Protocol who are super deep on IP/ preserving what’s yours in the face of the model providers eating everything up. Is that it truly is a race to the bottom and nobody is doing very much about it. Model providers will devour all data and IP and have unlimited capital and will fight vehemently if you sue them.
So the only cogent set of actions are :
- Trying to architect a system that is not like this bc the current system is insane. Which is what Story is trying to do. But is super hard bc ultimately – the Model companies have all the cards so any interaction they have with any sort of IP protocol - on chain, or off chain- is going to be ultimately, voluntary
- Building a network of people who see that the Digital Behemoth is devouring all their information and want to do something about it. I.e. People who assume that - no Story Protocol won’t fix things, OpenAI and Anthropic will just be bad actors in perpetuity, and that we need to exist anyways.
So Story is the optimism bet. Whereas something like Post Fiat is closer to the ‘pessimism bet’. THe basic premise is that you can apply AI to make a hyper efficient information, skill and data market. By vetting individuals interacting with the system, and routing queries to the correct individuals using AI agents rather than peer to peer centralized communication as facilitated by Tegus.
So going back up to the TAMs / white spaces above:
- Indexing - you need a methodology and an ontology to capture what agents or individuals are actually saying and construct indices that have total return series associated with them to measure peoples’ credibility
- Market Intelligence - you need a vetting system similar to Bridgewater’s DOTS protocol - which assesses peoples’ credibility in different domains
- You need to have bifurcations and measurements between a. human analytics b. machine analytics c. the synthesis of human and machine analytics
And – for you to get people to participate in such a system they have to have a line of sight that DOING SO IS FINANCIALLY VALUABLE. So the role of the Post Fiat Token - in this regards, is to get people bought into the compounding network effect. Financially, such that they actually use the indexing rails, consume market intelligence, and do analytics.
And because Blockchains dovetail with existing financial institution compliance needs. Which are complex, painful and onerous - my role is basically to create a super scaled, compliant, financial hive mind. Because such a system is going to protect the economic interests of its constituents, and drive a potentially huge market value
Tying it All Back Together
If we’re going into a Post Fiat world everything that we have in the Fiat world today is coming with us, in roughly the same proportions except:
- It will be more digital
- It will be more AI driven
- It will have less capability to enforce property rights
So from first principles the question, if you want to go Post Fiat is “how do I architect a financial system likely to generate the maximum possible market value in such a world” – as this faith inducing act will drive token value, which will in turn drive data contribution, and then drive financial compounding effects that would be inaccessible in traditional hedge funds
Blockchains start as a compliance, and analytics function – but then eventually evolve into a settlement mechanism.
It’s all pretty complex. But it’s not like the world we are living in is simple.
I find the existing narratives about things like Hyperliquid or most crypto protocols are all about extracting money from their users. Whereas the exciting thing about Post Fiat - despite it being pretty ‘doomer’ in some ways is building real tech that helps everyone using it compound their capital more effectively
A giant, self compounding on chain financial hive mind. That - in my view, is how you make it going into AGI. Not as individuals, but through AI coordinated collective action
Concrete Product Things
So this was all a big complex discussion of what is actually being done but translating it into clean product features and tying it back to what I talked about yesterday
- You need a good, on chain indexing product that uses blockchains and encryption effectively
- You need a good compliance product for receiving information from internet anons – who are paid in Post Fiat Tokens to provide data or expertise
- This compliance and indexing product should be linked. For example - when there is an index update you shouldn’t just have the membership. You should have the reasoning used to generate the index components, the underlying compliance hashes,
So it’s basically Asset A, weight, reasoning, model selection, prompt hash, compliance hash
or something like that – in a big JSON blob with a protocol for when it’s published vs when it’s decrypted for the public (in case you’re running a transparent ETF)
I feel like this is an anticlimatic ending but this is a blog so I guess you just have to deal with that bc this needed to be recorded somewhere. But yes
Concretely: THIS INDEXING PRODUCT WILL POWER THE POST FIAT STRATEGY ETF