The tbDEX Protocol

Grade: C

If the aim is to promote financial inclusion the protocol needs to recognize and consider the implications of mobile money as a settlement layer.
Identity negotiation for each ASK produces tremendous overhead and delays. Instead negotiate once per session and counter-party.
Prices should be streamed and be valid for a defined time period.
Making tbDEX compatible with the existing FIX protocol would simplify implementation of gateways to existing pricing engines and speed up adoption.
User managed digital reputations need to be addressed. In particular a way for users to convey sentiment towards PFIs to other users.
The protocol should add cryptographically signed timestamps to facilitate conflict resolution.

This protocol perpetuates the existing system of capital controls by extending it into the new arena of cryptocurrencies.
It is unclear how DID managed identities help at all if Alice needs to use a settlement layer that reveals her identity in full (bank account, credit card)
The protocol is asymmetric. The off-ramp protocol pushes all transactional risk to Alice without offering escrow release or conflict resolution when Alice sells crypto assets.
Claims it will increase liquidity for the end user but doesn’t define where the order book will be kept, how it will be populated or why an end-user would want to keep track of such a data structure in his wallet.
Initial BIDs issued by PFIs are non-binding and can be changed after the user has already revealed VCs.
Both the on-ramp and the off-ram parts of the protocol won’t scale if each transaction renegotiates the VC of participants.
The authors show no awareness of existing trading and pricing protocols (FIX protocol) and how they related to pricing engines used by financial institutions. Scaling the current draft protocol to billions of users is unrealistic.
Keeping a local order book for each wallet is a prerequisite for efficient capital allocation and narrow spreads, yet this is near to impossible without streaming price information.
Wallets pre-loading DID sentiments is a particularly bad idea given the track record of pre-loaded CAs in web browsers.
The protocol defines no cryptographically secured timestamps.