Who it is for
One submission, every carrier on your panel
An MGA does not rate a risk once. It rates the same risk against every carrier program the risk qualifies for, on the version genuinely effective for that policy date, and then has to be able to show each carrier that its own filed position and its own deviations were the ones applied. SelectRate treats that as the normal case rather than the awkward one.
Delegated authority means one risk has several right answers
You hold authority from more than one carrier. Each carrier has its own filed position and its own approved deviations, and each expects business written on its paper to be rated to that position exactly.
So a submission that would produce one number inside a carrier produces several inside an MGA. The work is placement: which paper will take this class, in this state, at this limit, on this date, and at what indicated premium.
Rating systems are usually built for the single-carrier case, and the panel gets bolted on afterwards. That shows up in two familiar shapes. Either the same submission is keyed and run once per carrier, which is slow and drifts as soon as one run is done from slightly different data, or the comparison lives in a spreadsheet next to the system and quietly stops agreeing with it.
If you want the platform view rather than the MGA view, the SelectRate platform overview sits above this page. AAIS programs on your panel are rated the same way; see AAIS rating.
One request, every program you are entitled to
The Rating API is synchronous JSON. Leave programCode null on a quote and one request rates every entitled program for that product, state and effective date. Each program comes back as an independent result, with its own premium and its own worksheet.
Independent is the important word. One program can return Rated while another returns Referred or Declined, or no premium at all because it is not effective for that date, and you see which and why per program instead of losing the whole call because one carrier on the panel did not want the risk. HTTP status describes the API call. Result status describes the insurance outcome, so a decline still comes back as a successful call.
What fans out is decided by entitlement, not by the request. Scopes gate verbs, grants gate programs, and there is no wildcard at request time. A program you are not entitled to returns 403 without disclosing that it exists.
Worksheets are not inlined, because a single rating run can carry seventy or more steps. Each result links to its own worksheet and you fetch the ones you want to read.
HTTP status describes the call. Result status describes the insurance outcome. A response of nothing but declines is still a 200.
Compare side by side before anything goes out
Underwriter Workbench is where your underwriters work: start a quote, rate it against every carrier program it qualifies for, read the worksheet, apply a debit or credit, bind, and then endorse, cancel, reinstate or renew the policy. It rates on published configuration, on the same engine as the Rating API. Rate once and see the indicated premium from every carrier program side by side, on the same submission, with the same effective date.
Workbench and the Rating API run the same engine. Identical inputs, program and effective date produce an identical premium and an identical worksheet, so what an underwriter approves in Workbench is what an authorized channel gets back from the API.
Judgment is kept separate from the rate. The engine's indicated premium and the final premium are reported separately on every result, so judgment applied on top of the rate is always distinguishable from the rate. A carrier reviewing your file can see where the rate ended and where discretion began.
Proving whose position was applied
SelectRate separates rating content into three layers. The bureau rating base, ISO or AAIS, is maintained by SelectRate. Carrier configuration holds carrier-approved deviations from that base, including company multipliers, by state. Program configuration holds the pricing and rules specific to your program.
The middle layer is the carrier's. It is the thing each carrier audits you against, and in a fan-out each program comes back as an independent result, carrying the configuration that program was published with rather than one house position applied to everybody. The same holds after the bind: every endorsement and renewal is rated on the program that bound the policy, with its own worksheet.
Underneath each result, every rating step is captured as the engine executes it, with the governing value and where that value came from. Provenance records which filed mapping was selected, why it governed, and the document and page it came from. On general liability, territory is derived from the filed mapping and reported in the worksheet provenance rather than being supplied on the request, which removes a category of dispute before it starts.
Every intermediate value is shown at the precision the engine used. There is no rounded summary standing in for the calculation. When a carrier asks what its deviation did to a rate two years ago, the answer is a document, not a reconstruction.
Adding a carrier should be configuration, not a release cycle
The trigger for most MGA rating projects is not dissatisfaction. It is a new appointment, or a new program, and a date it has to be live by.
You configure the program in Program Studio: deviations, policy writing minimums, fees, rules, eligibility and states. You configure it. Not a support ticket.
Every change is tested in Program Studio before it can go forward: rate complete test risks against the draft version and against what is published, and read the worksheet behind each result. A version cannot be submitted for release until at least one test has rated, and the review step shows exactly what it changes against the published configuration, state by state. A published version takes effect on its effective date and never changes afterwards. On its effective date the change is live in Workbench and on the Rating API at once. If a program is not effective for the quote date, it returns no premium rather than quietly answering from the wrong version.
The configuration is the part SelectRate speeds up. It is not your carrier's appointment paperwork, and nobody should pretend otherwise. What it does mean is that the rating system stops being the thing that sets your launch date.
Sub-producers without handing over the panel
If you delegate down, give each downstream partner its own API client. Each API client has its own credential and is granted the specific programs it may rate. Scopes decide which operations a credential can call. Grants decide which programs. There is no wildcard at request time, and a client never sees another client's quotes or policies. The channel-edge problem in full, including retail agents, is covered on the wholesaler and program administrator page.
What SelectRate records, and what stays in your system
SelectRate rates and records policy transactions. It holds the bound policy term, its current rated state and a ledger of every transaction, each with its calculation and worksheet. It is not a policy administration or accounting system: billing, invoicing, commissions and accounting stay in the systems you already run.
Endorsement, cancellation, reinstatement and renewal are rated in SelectRate from the policy's current accepted state, on the rating versions locked at the term's effective date. Every accepted transaction is kept, unchangeable, in the policy's transaction ledger.
Rating is synchronous, and worksheets are JSON. Knowing the edges before you integrate is worth more than a longer feature list.
Questions MGAs actually ask
Can we rate the same submission on carriers with different effective dates?
The effective date is the only temporal input the API accepts, and a fan-out rates every entitled program for that product, state and date. A different date is a different rating run.
Does the comparison pick a carrier for us?
No. Each entitled program returns its own independent result and its own worksheet. Ranking, appetite and placement remain your decisions.
What happens across the panel when a circular issues?
The base layer moves underneath you and your configuration stays intact. Rules, rates and loss costs are maintained against circulars as they issue, so you are not reapplying your own deviations carrier by carrier every time content moves.
What should we bring to a demo?
Bring a program. Pick one you would like to move, a state, a class or a whole program, and we will configure it live with you on the call. Bring the carrier whose deviations are the most awkward, not the easiest.
Bring the panel, not a single risk
The honest test of an MGA rater is a submission that qualifies on four carriers and is marginal on two of them. Book a call, bring a program you already write, and we will configure it with you while you watch.