Thirty Minutes of Coverage: Rob Schenone on How Space Risk Really Gets Priced
Most insurance runs on volume. Thousands of similar risks, decades of loss history, a model that turns all of it into a number.
Space insurance has none of that. And it works anyway.
Our Co-President of Aerospace and Head of Space, Rob Schenone, recently joined Amrish Singh, CEO of Liberate AI, on the Everything Is Insurance podcast to explain how. The episode starts with a specific question — who paid for the $200 million AMOS-6 satellite destroyed on the pad at Cape Canaveral in September 2016 — and ends somewhere much bigger than satellites.
Rob has spent 35 years underwriting this class of risk. Here’s what came out of the conversation, and why we think it matters well beyond the space community.
Coverage begins at ignition — and often ends half an hour later
“We’ve all seen a rocket go up,” Rob told Singh. “As soon as that ignition, as soon as that separation takes place, that’s when our coverage begins.”
On a Falcon 9, that window is roughly thirty minutes.
Which produces one of the stranger economics in specialty insurance. As Rob put it: “You work for a long time, you put this policy together and it’s over in thirty minutes.” Months of technical review, engineering diligence and market-building, resolved in the time it takes to drive to the airport.
That compression is not a quirk. It’s the defining feature of launch underwriting, and it explains why the discipline looks so different from the rest of the industry.
There is no model — and that’s not a failure of sophistication
Asked how you price something like this, Rob’s first answer was a joke: “We just put our finger in the air and do what we’re told to do.”
The joke lands because the underlying point is serious. The machinery that makes conventional insurance work — mountains of data, decades of loss runs, credible actuarial frequency — simply isn’t available. There are only a few dozen insured launches in a given year. Every satellite is effectively a one-off: a bespoke machine, on a specific vehicle, into a particular orbit, for a single mission profile.
So the work shifts. Instead of a formula, it runs on accumulated judgment: what this manufacturer’s heritage looks like, how this vehicle has performed across its flight history, where the failure modes cluster, what a given configuration has done before.
Rob has been building that judgment — and the underlying record — since 1991, tracking the market by hand from the beginning. That’s not a charming anecdote. In a market this small, institutional memory is the dataset.
2023 didn’t have a tidy explanation
Rates in the space market had been sitting at historic lows. Then 2023 happened.
“Boy, I wish we knew,” Rob said when asked what caused it. “A lot of bad things happened, that’s what I know happened. We saw a number of launch vehicle failures. We saw a number of satellite failures.”
Was there a root cause? “I just think it was bad luck. It was bad timing. A lot of things happened at once.”
That answer is more useful than a neater one would be. In a market with this few risks, correlation and coincidence are genuinely hard to distinguish, and a single bad year can reprice everything. Rates moved sharply and capacity left. There’s no post-mortem that makes the volatility go away — there’s only the recognition that in low-frequency, high-severity lines, one year can carry the weight that a decade carries elsewhere.
The bigger story: risk that nobody is writing down
The most consequential shift Rob described has nothing to do with pricing. It’s about who buys coverage at all.
“If you’re launching a large number of satellites, your risk management is built into your constellation,” he explained. “So if one goes down, you’re not in that bad of trouble.”
That’s a rational position, and it is now the dominant one. When you operate hundreds or thousands of satellites, redundancy does the job a policy used to do. Lose a unit and you replace it on the next launch. Filing a claim would cost more attention than the asset is worth.
The result is that a growing share of what’s in orbit is uninsured — not because operators can’t afford coverage, but because the architecture of their business already absorbs the loss.
When an operator declines coverage, the exposure moves onto its own balance sheet — and from there onto its lenders, its investors, and its shareholders. The episode traces exactly where that leads, including through public markets now that the largest launch provider in the world completed the biggest IPO in history in June 2026. We’ll let Singh land that argument himself; it’s the best few minutes of the episode.
But the principle is one we’d underwrite anywhere: the difference between a company that absorbs a catastrophic failure and one that doesn’t survive it is rarely the failure itself. It’s whether the risk was priced and placed or quietly carried.
Why Novacore built an aerospace practice
We launched our Aerospace segment because this is precisely the kind of market where specialist judgment is the product. Low frequency. High severity. Thin data. Rapidly changing risk architecture as constellations replace one-off satellites and new operators enter with new failure profiles.
Those conditions don’t reward volume. They reward people who have seen enough launches to know what they’re looking at — and who are candid about the limits of what anyone can know.
Rob’s thirty-five years are the reason we can be in this market at all.
Listen to the episode
Everything Is Insurance with Amrish Singh, featuring Rob Schenone and Chris Quilty of Quilty Space.
