The declinature drawer
The hardest placements don’t die of price — they die of routing.
Underwriters keep a drawer for this. So do brokers. In it are the risks that would have been written at the right terms but died in somebody else’s inbox — and, facing them, the risks declined on a Tuesday that were exactly what the pen two desks away had been asking for. The tenth paper in this series argued that time is the single physical variable running through the delegated-authority business. Nowhere does it bite harder than here: a placement is a race between a renewal date and a market’s attention, and the market’s attention is not organised to run races.
Watch a hard placement die and the post-mortem is almost always the same. The risk was insurable. Somewhere, a market had genuine appetite for it — at a price, on conditions, but genuine. The risk and the appetite simply never met before the clock ran out. We tell ourselves these files failed underwriting. Mostly they failed geography: the wrong inboxes, in the wrong order, with no way of knowing which right inbox existed. That is not a pricing failure and it is not a broking failure. It is a routing failure, and the market treats it as weather.
Four failures, one habit
| Where it fails | What actually happens | What it costs |
|---|---|---|
| Appetite is a document | Appetite guides are published annually and go stale the week they are printed. The market’s real appetite lives in the heads of line underwriters and changes with every quarter’s results | Brokers market to last year’s appetite. Submissions arrive at pens that closed months ago while open pens hear nothing |
| Submissions are prose | The risk travels as an email and a PDF. Nothing about it is matchable — occupation, class, limit and territory are buried in attachments a reader must open | Every match depends on a human noticing. Attention, not appetite, becomes the scarce resource |
| Silence is free | No convention obliges a market to answer. The submission that doesn’t fit simply sits, and the broker cannot distinguish “considering” from “never opened” | The broker’s margin is spent chasing. The client’s renewal is spent waiting. Certainty — the thing a deadline buys — is unpurchasable |
| The no carries nothing | Declines arrive, when they arrive, as a sentence: not for us. The reason — price, occupation, limit, capacity — stays with the decliner | The single richest dataset in the market, revealed preference at the point of decision, is discarded daily |
The economics of silence
Price the failure and it stops looking like weather. On the broking side, a hard placement consumes hours of chasing per market approached, against a fee that assumed a routine renewal — the margin of the toughest files is spent on telephony. On the underwriting side sits a stranger paradox: markets spend real acquisition money — BDMs, sponsorships, lunches — to attract submission flow, while submissions matching their exact appetite die unseen in a competitor’s inbox for want of anything better than folklore about who writes what. Both sides pay for the same missing information: who, today, actually wants this risk?
And underneath both runs the clock. A renewal has a hard date; every day of silence narrows the options and worsens the terms. The client experiences the failure as their broker “still trying”; the broker experiences it as unbillable hours; the market that would have written the risk experiences nothing at all — which is precisely the problem. In many other financial markets, an order that finds no counterparty at least fails visibly, and the failure itself is information. Here, the failure is silent, so nobody learns anything, so it repeats.
Declines are data
The waste compounds in a second way. A decline is not an absence of information — it is a revealed preference, delivered at the moment of decision by the person who holds the pen. A market that declined on occupation has told you its occupation boundary; a decline on limit has priced its ceiling; a decline on capacity has dated its treaty. Collected and kept, a year of reasoned declines would describe the market’s true appetite more accurately than every published guide combined. Instead the reason is rarely given, never structured, and never kept.
The delegated-authority world has spent a decade learning that premium data compounds when it is captured once and coded properly — the ninth paper made the same argument for claims. Placement data has had no such decade. The submissions, the quotes, the declines and their reasons — the market’s own map of itself — evaporate at the moment of creation.
What a functioning market would require
There is no mystery in any of this, and effort alone does not fix it — by better BDMs, longer lunches or more diligent chasing. The properties of a placement process that doesn’t leak are structural. The risk would be described once, in terms a machine can match — occupation, class, limit, territory — rather than prose a human must notice. Appetite would be a standing commitment to respond, not a brochure. Every submission would carry a deadline, and every match would owe an answer by it: a quote, or a no with its reason attached. And the answers would be kept, so that each placement leaves the market knowing slightly more about itself than before. These aren’t behaviours. They’re infrastructure. Which is the optimistic point — because behaviour problems persist, but infrastructure problems, once named, tend to get solved.
The routing test
Two questions measure a placement process. When you decline a risk, does anyone — including you — learn anything that survives the week? And when your appetite changes, how long before the brokers who need to know actually know? If the answers are “no” and “whenever the guide is next printed”, the drawer will keep filling. Not because anyone priced badly — because nothing carried the message.
The conclusion
The hardest segment of the market is not hard because the risks are unwritable; it is hard because the routing is unbuilt. Risks with willing pens die of silence, declines that would map the market’s appetite are thrown away, and both sides pay — the broker in margin, the market in acquisition costs, the client in terms — for the same missing piece of information. Ten papers in this series have argued that delegated authority’s costs hide in its infrastructure and surface in its latency. The eleventh observation is simpler and, for once, ahead of the fix rather than behind it: placement is fundamentally a routing problem, and routing problems can be fixed with better infrastructure. The firms that treat every submission as a routed, deadlined, evidenced transaction are best placed to win the segment everyone else calls hard.
Cuttleflow Systems · Perspective 11 · 33°53′S · 151°16′E · Sydney