One AI-powered MGA with hedge fund experience has created a model that allows it to accurately underwrite price protection insurance for raw materials
Insurance, at its very best, is a beautifully simple product that provides individuals and companies with the confidence to take measured risks, safe in the knowledge that, should the worst happen, some of the consequences can be mitigated or avoided.
When the very first marine underwriters emerged in the ports of Europe, it was that promise of restitution and risk mitigation that allowed ships to navigate the globe and plant the seed that turned into the global trade networks we all enjoy today.
Insurance has lost some of its simplicity since the early days but, at its core, retains this function. It covers everything from rocket launches to weddings, with the data underpinning this underwriting having become ever more complex.
However, few would argue that insurance trades in anything as complex as hedge funds do. Global mega-corporations definitely take out insurance for many of their activities, but their scale allows them access to markets where they can hedge their risk in raw materials by trading in futures or other derivative products.
This approach works for companies of a certain size that are sufficiently exposed to certain commodities or raw materials, but is incredibly complex and limited by both the existence of markets, which don’t exist for materials like recycled plastic, and the necessity of scale.
A new MGA has created a solution to this problem, however, opening up new markets for the insurance sector while also providing customers of all sizes with the ability to take decisions that were previously outside their risk appetite.
Democratised hedging
ChAI is an MGA that makes use of an artificial intelligence (AI) pricing model in combination with traditional data to predict the future pricing of a commodity and then underwrite price protection insurance products for customers – allowing them to effectively hedge against adverse price movements without handling the complexity of derivative products.
Read: TechTalk – Agentic AI presents a sink or swim moment for insurance sector
Read: TechTalk – Is the cyber insurance world prepared for quantum computing?
Explore more TechTalk content here, or discover other news analysis stories here
Tristan Fletcher, co-founder and chief executive at the firm, has experience working in investment banks and hedge funds and has lectured on AI at universities like Imperial College and Cambridge.
He explained: ”We built a product that would forecast commodity prices very accurately and communicate this information to corporates. But while doing this, we realised a lot of companies weren’t really agile enough to act on our forecast and, in a way, found it a little irritating to be told something was about to happen without the ability to do something about it.
“You have to be a very large, sophisticated company to get rid of price risk by buying a derivative, an option or a future and there aren’t even hedging mechanisms in many markets.
”So we’ve built an insurance product that is filling that gap, in both markets where there’s no ability to remove price risk at all and in markets where companies are too small to access them.”
Fletcher and his co-founders’ intention was to build something “beautifully simple, that companies of any level of sophistication would be able to buy and that would operate in any market”.
With experience of trading in commodities pricing, the ChAI team has secured backing from venture capital, a panel of reinsurers and is in the process of putting together a panel of primary insurers that provide it with capacity.
With this backing, Fletcher said he and the team are aiming to “democratise access to price hedging tools for smaller customers” and expand into markets where it can help support resilience and new economic activity.
New markets
Obviously, much of the success of underwriting price risk in materials will rely on the accuracy of underwriting. However, Fletcher noted that this is actually rather simple in underwriting terms.
He explained: “In some markets, like industrial metals, we are very accurate. In others, like energy, we are much less accurate. But the important part is that we’re more accurate than anything anyone else is able to do and are as competitive as the systematic hedge funds.
”What’s more important in insurance pricing is the distribution of possible prices and our ability to collapse that distribution into what it implies for an insurance premium. We’re doing it with AI, so that distributional forecast ability is incredibly accurate and the economics of that accuracy as very positive for people supplying insurance.”
What this level of fidelity in underwriting creates is new price risk insurance markets where they have not previously existed.
Fletcher added: ”We can really step into markets that are very idiosyncratically priced or very new – and recycled plastics are an example of both. Companies like Coca-Cola use vast quantities of this stuff and its very volatile in pricing.
”Also, companies are fairly agnostic about what raw materials they use, apart from being price sensitive, and the virgin versions and recycled versions are very similar. But, even if companies want to be green, they won’t use recycled materials if they can’t hedge out the price risk.
“Now that we’ve provided the ability to hedge price risks in recycled materials markets, these companies can be as green as they want. We’re very excited to contribute to the circular economy and so are our customers. They’ve been wanting to do this and now they can, because there’s a market mechanism to do so.”
The potential benefits of this sort of added resilience are plain to see – with new insurance markets providing companies with simple solutions and more options than they had previously.
Fletcher finished: ”Smaller companies have such thin margins and they do go out of business because their raw material costs go up. This is a real opportunity for the insurance industry to really step up and help companies that need it.”
And isn’t that what insurance is all about?

With a particular interest in regulation, technology, innovation and political stories, he has covered issues from the multioccupancy buildings scandal to the insurance implications of quantum computing and the growth of new markets.View full Profile
No comments yet