The latest from AARC.

This paper defines the verifiable program record and estimates the cost of its absence in the US delegated-authority insurance market. Motivated by the growth of US MGA premium to roughly $128 billion in 2025, of which fronting carriers support about $22.6 billion (Conning), and by the entry of InsurTech underwriters pricing on signals that capital providers cannot independently verify, this paper develops a quantitative analysis of the information load that every delegated program carries alongside its risk. In delegated-authority business, underwriting and claims authority sit with parties separated from the capital that bears the loss. The market prices this load by withholding capacity, locking collateral beyond what loss determination requires, and excluding new or small programs regardless of quality, so that the volume a broker can place is constrained below what the client's book would support. We define a verifiable program record by four properties: contract-bound terms, provenance on every row, an append-only ledger, and determinations any market can replay. We decompose these dynamics into an uncertainty margin we call the addressable information load, which we estimate at two to five points of premium, with collateral drag of one to three points and re-verification across markets of $150,000 to $400,000 per program per year. Applied to the Conning premium base, this implies an annual addressable load of $2.5 to $6.5 billion (central ≈$4 billion). Operating a verifiable record costs roughly $2,000 to $10,000 per program per year; a submission that markets can replay rather than re-audit removes an estimated 60 to 80 percent of re-verification and reduces bordereaux latency from 30 to 60 days after period close to days. The share of savings reaching policyholders depends on pass-through in each program segment, which we treat as an empirical parameter and report as a range of consumer welfare gains. We argue the record delivers these gains only if its specification is open, and we set out that specification in a companion paper.