Sam Bankman-Fried (SBF) ex-CEO of FTX

Triple Entry Accounting, the “Independence of Irrelevant Alternatives”, and Blockchain

Jon Gulson
4 min readNov 13, 2022


The recent collapse of cryptocurrency exchange FTX has provided easy pickings for cryptocurrency critics, since estimated losses run into billions of US dollars with customers facing little immediate prospect of redress.

These events raise questions on the accounting and reporting practices used by FTX and its ex-CEO Sam Bankman-Fried — it is said a “backdoor” existed in FTX’s books, created with bespoke software. This means (if true) Bankman-Fried could have made unflagged changes either internally or externally to the company’s financial record.

This is similar to “single entry accounting”, which is heavily influenced by sole parties (in usual circumstances small businesses) who record the “bare essentials” — it means theft or losses are less likely to be detected, compared to “double entry accounting”, which provides for a more methodically rigorous procedure in balancing debits and credits (as double entries), and for which larger entities or corporations will more likely use, mitigating potential for fraud among other benefits.

Triple Entry Accounting (TEA) and Blockchain

There is an irony presented in all this, in that the bitcoin white paper speaks to financial institutions being unable to avoid fraud or mediation — and that in addressing the fraud problem (as described by double spending), bitcoin removes a “trusted third party” or centralised mint of the coins the blockchain issues as incentive. For Satoshi, it was all to be about fraud prevention.

It’s this prevention idea which leads into the “triple entry accounting” (TEA) method, whereby a third party exists, but where that third party is a timestamp of a digital ledger entry — for example if Bob and Alice transact, the third party (Charlie), in today’s world, is “the blockchain”.

The TEA method is close to the “Ricardian contract”, which are contracts designed to be efficient structures expressed and executed in software, creating a definitive document, digitally, ensuring that document becomes unique by means of hashing and stamping. This is graphically demonstrated here:

The Ricardian Contract

Contracts, Game Theory, and Blockchain

The Ricardian Contract, so named by Ian Grigg after David Ricardo’s contributions to international trade, where Ricardo believed “comparative advantage” in industry specialisation and free trade produces value — and in a generalized sense of contract theory, the contract will then define and determine the value, be enforceable if necessary, inducing cooperation (in the game theoretical definition) — as opposed to non-cooperative games, where there is the absence of an external authority or contract, and where such (non-cooperative) games are self-enforcing through credible threats, focusing instead on predicting individual players’ actions and payoffs and analyzing Nash equilibria.

Notwithstanding questions on contract quality, or varying grades of cooperation, or identifying appropriate and trustworthy external authorities (or umpires) to enforce contracts in international trade scenarios (internet commerce would seem analogous to Ricardo’s economics in this regard), there is sufficient evidence to say bitcoin was consciously made as a triple entry accounting system.

Relating “Triple Entry Accounting” to “Independence of Irrelevant Alternatives”

As bitcoin was designed to support “a tremendous variety of possible transaction types”, this possibility (bitcoin as TEA) can’t be completely discounted. Alternatively, there is nothing explicit or publicly available to say Satoshi was an accountant or interested in accounting methodology — but there is evidence available showing Satoshi interested in contractual implementations of bitcoin, including “Escrow transactions, bonded contracts, third party arbitration, multi-party signature, etc”.

This opens a potential reconciliation in understanding the similarities between triple entry accounting and a social choice and bargaining axiom known as “Independence of Irrelevant Alternatives” (IIA) which takes us into the realms of individual behaviour, coalition formation, aggregated welfare, and majority decision making: IIA is used to understand voting preferences, where adding a non-winning (or irrelevant) candidate, shouldn’t change the outcome of an election.

With TEA, where Alice and Bob choose Charlie (the blockchain) as their preference (or representative) in recording transactions, then in relation to IIA, where Alice and Bob are in an election, the outcome won’t change by adding Charlie. Or if there is a choice between vanilla or chocolate ice cream, then adding strawberry ice cream as an alternative option shouldn’t change vanilla or chocolate from winning. In respect of bitcoin, the proof of work is defined as a voting mechanism with the most proofed chain accepted the most valid as it has the most computational power invested, (which bitcoin itself) says solves the problem of determining representation in majority decision making — in other words, in the TEA analogy used, Charlie is to be the most trusted representative, and in the IIA analogy, Charlie is creating Alice and Bob as peer-to-peer.

Thus, both TEA and IIA are related to issues of fraud — either accounting fraud, or election fraud (an observable chain of events is established). Where the issue then becomes the contract, both TEA and IIA suggest prevention is preferable in the first place, which will be scant consolation to the customers of FTX and their now seemingly inevitable and time-consuming litigation to recover losses. It at least however may provide some opportunity for learning new lessons regarding the origins of blockchain and its relation to contractual certainty and reliability, where the quality of the money unit the contract is expressed can make a big difference.