Canada's marine insurance market is being pulled in two opposing directions at once, and the tension between them is pushing prices down in a way that cannot last, according to a source who leads a major marine underwriting practice at a global insurer, speaking to Insurance Business Canada on condition of anonymity.
On one side, capacity is pouring in. New entrants are opening for business, and insurers with no history in the line are moving into marine for the first time.
"There's a lot of capacity being pumped into the system," the source said. By their count, there are now 23 markets offering a marine product in Canada, though only a handful write the full range. "Of those 23, maybe the top five or six will offer all lines of insurance."
On the other side is a shortage the industry did little to prevent. The people who can actually underwrite the business are not being replaced as fast as the capacity is arriving.
"There isn't enough talent out there to drive the business properly," the source said. Existing underwriters are being approached with competing offers, they said, while the pipeline behind them was never built. "In the marine world, we never really spent a lot of time investing in talent."
The result, the source said, is that inexperienced underwriters are being handed portfolios they are not equipped to manage, and they fall back on the only lever they understand. Marine is not a single product but a category spanning more than a dozen distinct coverages, and reading a complex risk takes knowledge newer entrants often lack. Without it, they said, "they'll just rate on price."
That dynamic is feeding a race to the bottom. Non-traditional insurers bring capacity that pushes pricing down, which helps clients in the short term but sets up a harder fall later.
"The cheaper we are in the marketplace, the sooner there's a correction," the source said.
Marine also sits in a structurally weak position when a client's wider program is under pressure, they said, because it is almost always the last line to be renewed. When property, casualty or other major lines are not delivering the reductions a client wants, the pressure lands on marine to make up the difference.
"Marine gets a phone call and they'll say, we need a reduction of 30, 40%," the source said, even when the larger lines are not moving. "And if you don't do it, someone else will."
Compounding it is how marine is often housed inside a company. Because a marine book is small next to a property book, the source said, many insurers fold marine into the property division so its results disappear into a larger number.
"They take the marine department and they have them report into the property, so their numbers don't become as critical," they said. The effect elsewhere, they said, is that marine is not always required to stand on its own profitability, which lets loss-making pricing persist.
The correction will come, the source said, but late. When other lines such as property or cyber begin to harden, marine will lag well behind. Other lines may firm in around 18 months, on their read, but marine will stay soft longer because the pressure to keep cutting will still be there and the excess capacity will still be in the market. They put the lag at two to two and a half years.
In the meantime, business is flowing out of the domestic market to where the expertise sits. A broker weighing a junior underwriter with limited capacity against an established international market will often choose the latter, the source said, because it brings the depth the risk requires.
"They have more capacity, they have expertise, they have a claims team, they have the people that we need to support the business," they said, describing the broker's reasoning.
Capacity can be raised overnight, the source said, but judgment cannot, and the market will not stabilize until insurers rebuild the expertise they stopped developing. Too many companies, in their view, are reaching for automation as the cheaper answer to what is really a people problem.
"When it comes down to the intuition that is developed by a seasoned underwriter, AI will never match that," the source said.
Underwriting authority is earned, not programmed, they said, and that judgment is built only through repetition. "You have to practice underwriting every single day to do that," they said, something a model cannot shortcut.
"I think executive leaders think that's a quick way of saving," the source said, "but I don't think it's going to be that successful in the long term."
