Twenty years before it was a company.
The weather risk economy is worth more than a trillion dollars, and almost none of it runs on software built for the job. Most of the industry still prices, transfers, and manages parametric weather risk the way it did twenty years ago: spreadsheets, phone calls, and whoever happens to remember how the last deal was structured. ClimateDelta exists because five people got tired of that, one desk at a time, over two decades, before any of us called it a company.
It starts in 2006. Marty was rebuilding a weather trading desk at a global reinsurer after its previous trader left, and he needed someone who could think in numbers the way he thought in risk. He hired Vivek — his third hire — a mathematician who'd spent the previous five years building the hurricane models an entire industry relied on without knowing his name, writing statistical curves in Fortran because the tools that would eventually replace that work hadn't been invented yet. Marty and Vivek have worked together every year since. Twenty years, this year.
Their first real test came almost immediately. Katrina hit, the book took losses like everyone else's did, and the reinsurer decided to leave the business. What stayed with Vivek from that year wasn't the loss — it was a rule he's carried through every model built since: it's not what your model does that matters most, it's what your model doesn't do. Understand the limitation, and build the policy terms, the limits, the payout structure around it. That rule is still load-bearing in ClimateDelta today.
The two of them spent the better part of two decades moving through the industry together — pricing insurance risk, pricing financial risk, sitting through a decade of audits on both sides of the business — and came out the other side with a conviction most of the market still doesn't share: that weather risk in an insurance policy and weather risk in a financial contract are the same problem wearing different paperwork. Almost everyone in this industry has built for one side or the other. Nobody had built for both, in one place, until this.
At one point, the team needed a real system and got a quote from an outside vendor: roughly a million and a half dollars to build it. Ling looked at that number, decided it was absurd, taught himself the languages he didn't yet know, and built the first working module himself. He showed it to Marty. Marty didn't say anything right away — he took a long walk, came back, and said we're bringing this in-house, all of it. That decision is the reason ClimateDelta is built the way it's built today: designed by the people who use it, not licensed from someone who doesn't.
Diana has been doing the unglamorous, essential work the whole time — back office, settlement, the reporting that has to be exactly right every single time because nobody notices it until it isn't. Vivek says the back office matters more than the front office, not less, because the front office is a profit center and can ask for whatever resources it wants; the back office is a cost center and usually gets none. ClimateDelta was designed to fix that math, and Diana is a large part of why it had to be.
Ralph brought the piece the two of them didn't have: a career spent actually trading energy, not just modeling weather's effect on it. Insurance people and energy traders tend to talk past each other, using different words for the same risk. Ralph is one of the reasons ClimateDelta doesn't.
For a long time, this was internal — the tool a small team used to run its own book. Parameter Climate, founded during Covid, used it to originate, price, and underwrite risk across the industry: a firm that has since worked through more than $500 million in limit and $100 million in premium, built by people whose combined careers span over $10 billion in limit. But the software kept getting bigger than the business that used it, so we did what the software was already telling us to do: we spun ClimateDelta out as its own company, built to serve the whole industry, not just one desk.
Every dollar of profit since founding has gone back into the platform. It has never taken outside money, and it has never stopped being used, in production, by real risk. This website is the first time any of that has been said publicly. We built it quietly for twenty years. We're done being quiet about it.