When a commercial submission arrives missing loss runs, the standard response is to ask the retail broker to provide them. The retail broker asks the insured. The insured contacts their current carrier and requests a loss run report. The carrier issues it on their own schedule, which typically runs between three business days and three weeks. The document then travels back through the same chain. That is four handoffs, each with independent latency, before the submission can move forward on the one piece of information the underwriter often waits for most.
Loss runs are structurally different from nearly every other required submission document. An insured can produce most missing items from their own records: a completed application, an updated statement of values, supplemental questionnaires, information about operations that wasn't in the original submission. Loss runs require action from a party that has no role in the transaction being processed. The incumbent carrier is not part of the distribution chain, has no financial interest in the new submission moving quickly, and issues loss run reports according to its own service standards. Some carriers have automated the process. Others require written requests. A few still route them through the agent of record, which adds another step.
The result is that loss run requests are routinely on the critical path for commercial submissions, and the timeline for resolving them is largely outside the control of anyone processing the file. Operations teams that understand this manage it by requesting loss runs at the moment of submission receipt, before the file has been reviewed for other gaps. The reasoning is straightforward: if loss runs take two weeks to arrive, they need to be requested immediately so the delay runs in parallel with the rest of the intake process rather than after it. Teams that wait until the rest of the submission is complete before identifying the loss run gap will consistently lose that two-week head start.
Even when loss runs arrive, they present a parsing problem that other submission documents don't. Each carrier formats them differently. Some organize data at the policy level. Others present it at the occurrence or claimant level. Some include reserves alongside paid amounts. Others show only what has been paid out and leave reserves off entirely. The distinction matters: a file with three open claims at $40,000 each in reserve looks very different depending on whether the reviewing underwriter can see those reserves or only the zero paid amount next to each claim. Two loss runs from different carriers, covering the same insured for the same period, can produce different apparent loss ratios because of how each carrier accounts for open claims.
This is one of the cases where improving the format of the document wouldn't fully solve the problem. Structured data from two different carriers still requires reconciliation because the underlying conventions aren't consistent. What one carrier calls total incurred, another calls gross incurred before recoveries. The field names suggest equivalence that the accounting doesn't always support. Extraction can capture the fields as presented. Interpretation requires knowing whose conventions you're reading.
The years requirement makes this more complicated. Some carriers underwrite general liability on three years of loss history. Others require five. If the submission arrives with three years of loss runs from the incumbent carrier and the target market requires five, the broker has to go back and request additional years from whatever prior carrier wrote the account during that period. That prior carrier may or may not still have accessible records. A submission can have complete, accurate, recently issued loss runs and still be incomplete for the market it's targeting.
Most operations teams track NIGO volume by category rather than by specific cause. That aggregation tends to obscure how often loss runs, specifically, are the blocking item. A submission flagged as incomplete because of missing loss runs represents a different operational situation than one flagged for a missing supplemental questionnaire. The supplemental can usually be produced within a day. The loss run is on an external timeline. Knowing which accounts are held up waiting for loss run requests to come back, and how long each has been waiting, is different information than knowing that the submission is in an incomplete state.
The follow-up cycle looks the same from the outside in both cases. The mechanics and the leverage available to resolve them are not.