Each additional layer of yield enhancement compounds technical, smart contract, and tail risk. We propose a simple decision framework.
The proliferation of yield-bearing wrappers around ETH and stablecoins has accelerated since the spring. Among them: liquid restaking tokens, wrapped staking tokens with leverage embedded, stablecoin LP tokens auto-compounding across protocols, and tokenised treasury exposures with embedded redemption mechanics. Most of these are recent enough that their behaviour in stress is not well understood.
We have evaluated 23 such products in the last six months. We have approved two. We share here the framework that produces such a high rejection rate.
The number of distinct protocol dependencies in a single token determines the slashing or operational surface. A liquid restaking token that holds stETH that is restaked across three AVSs is, in dependency terms, four layers deep. A failure at any layer compromises the position. We do not approve any product above three layers, and we evaluate two-layer products on case-by-case basis.
Sustained yield must be attributable. We require issuers to disaggregate the source of yield by counterparty, by protocol, and by economic mechanism. Where yield is described in aggregate ("4.5% APY") without underlying composition, we treat the product as opaque and decline.
Tokens that are not 1:1 redeemable for their underlying carry duration risk. We evaluate redemption windows, queue mechanics, and historical exit behaviour during stress. The acceptable spread between token price and underlying NAV is, in our mandates, below 80 basis points on a six-month average.
We require at least two audits from named, independent firms with sufficient sample of the codebase. Audit reports must be readable in the original language. Marketing summaries are not acceptable substitutes.
We do not allocate client capital to products whose operating team is anonymous. The crypto industry's tradition of pseudonymous development is admirable but not compatible with the legal and fiduciary structure of family wealth management.
A product with less than nine months of mainnet history is, by default, declined for client mandates regardless of its other qualities. Exceptions exist; they are rare and named.
The two products that have passed this framework since spring are, unsurprisingly, ones operated by named institutional teams with multi-year audit histories. We will not name them publicly here — clients in scope have been informed via individual updates.
A general note. The temptation to capture an additional 100 basis points of yield through layered products is real, and it is amplified in a low-volatility environment. The history of finance suggests that such temptations are most reliably resisted in periods when they are most acute.