Claims bots assemble

Hey Fintech Fam!

Welcome to the last edition of FintechFT. Sid and I are excited to bring you a new and improved newsletter beginning next month. Stay tuned!

For this week, insurance correspondent Ian Smith takes a look at the chat bots that are increasingly handling your claims. We also highlight a profile of Jake Freeman, the amateur day trader some corners of the internet believe is too good to be real. (He is real. We checked.) And for our final Q&A I spoke with the CEO of an AI-powered search engine for wealth management products that has attracted more than $200mn in investments from the likes of JPMorgan and Franklin Templeton over the past four years.

This newsletter will be taking a one-week break in observance of Labor Day in the US and will be back with a new name and format on September 12.

Until then, you can reach Sid and I at imani.moise@ft.com and sid.v@ft.com.

Happy reading!

Rise of the claims bots

Of all the ways technology is changing the insurance sector, arguably the biggest shake-up has been to one of the sector’s more prosaic areas: claims. Things going wrong, and putting them right again, is the essence of insurance. How companies manage their claims is crucial not only to profitability but also to keeping customers onboard.

Artificial intelligence is fundamentally changing this process, promising more accurate predictions of financial loss, enhanced fraud checks and ultra-fast payouts. For physical damage, a crucial component is the use of computer vision systems, which mine images and videos for information to gauge damage and spot red flags.

China’s Ping An is at the forefront of implementing this technology across the motor insurance sector. Its algorithms can analyse car damage based on images and video that customers can send to the insurer. The average claim is investigated in less than five minutes, and nine out of 10 claims are paid within one hour, the group says.

Global insurers that do not have the tech or scale of Ping An — its auto app, has more than 150mn registered customers — have seized on the technology through partnerships.

One big winner has been Tractable, which branded itself the “UK’s first computer vision unicorn” when it reached a $1bn valuation last year. It works with dozens of insurers worldwide including the UK’s Aviva, Geico in the US and Japan’s Tokio Marine.

Tractable’s technology, trained on hundreds of millions of images, can assess car damage at the roadside or assess a garage’s repair estimate, among other applications.

The platform allows insurance customers to submit photos of the wreckage that AI bots can scan instantaneously to determine if more photos are needed or calculate a repair cost or proposed payout within about 30 seconds.

Automation at this scale has been in use at Tractable for the past couple years — and it has continued to work on the model and broaden its use. The goal is to provide a more accurate prediction of the repairs than a human specialist, and in a fraction of the time.

People “get tired, we miss things, we can’t zoom in and zoom out constantly”, said Mohan Mahadevan, Tractable’s chief science officer and a former Amazon executive. “We’d like to go one step beyond where your average human performance is.”

Zurich is another insurer using computer vision tech in some of its smaller markets. “The key for us is saving the assessor going on site,” said Ericson Chan, Zurich’s group chief information and digital officer. Chan also highlighted the fraud prevention benefits: image mining can show up if the event pictured matches the claim, for instance. Insurers are increasingly combining this with natural language processing of policies and other documents to get to an automated decision on whether a claim is covered.

Still, looking beyond surface damage is a challenge. Stripping out human interaction from the claims process, after something as anxiety-inducing as a car accident, has downsides. And there are perennial concerns about what is going on in the black box.

For motor services, computer vision has implementation far beyond insurance: Auto parts providers use Tractable’s scans to help them decide which salvage vehicles to buy, and rental companies use the technology for inspections.

Within insurance, its applications go well beyond motor. Zurich’s Chan highlights home insurance, where image analysis is helping insurers to be more proactive in spotting risks coming down the track.

“We use satellite images to look at every home and assess the risk even before the policy is issued,” he said. “Then we can advise the customer — this bush next to your house, that could be a fire risk.” (Ian Smith)

Fintech fascination

Digital transformation reaches legacy players St James’s Place, the UK’s largest wealth manager, joined the 21st century and launched its first mobile app. Wealth and asset management incumbents have proved their low-tech, high-touch strategies could be resilient even in the face of digital challengers, but Covid and changing generational preferences have prompted some of the slowest-moving giants in the industry to react.

Payment pressure Card network giants Visa and Mastercard blamed increasing fraud and competition in the payments space as they were forced to defend rising interchange fees to government authorities last week. The UK’s Payment Systems Regulator is also investigating the companies after finding processing fees have increased fivefold since the UK left the EU.

The man behind the meme stock The FT’s Antoine Gara and Madison Darbyshire profile Jake Freeman, whose $110mn gain on Bed Bath & Beyond stock launched him into fame last week. Digital brokerages such as Robinhood have reported declining engagement this year, but captivating stories like Freeman’s could convince more day traders to try their hand once again.

Quick Fire Q&A

Every week we ask the founders of fast-growing fintechs to introduce themselves and explain what makes them stand out in a crowded industry. Our conversation, lightly edited, appears below.

Last week I spoke with Vinay Nair, founder and chief executive of TIFIN, a wealth management intelligence platform that counts some of the largest US asset managers and banks as clients and investors. The company, which bills itself as a data and AI-powered search engine for the world of wealth and investments, has raised over $200mn in capital since its founding in late 2018 from the likes of Franklin Templeton, and JPMorgan.

Why did you start the company? We felt that wealth advice could be really individualised and personalised a lot more. Think of it as a Spotify, Netflix-type of transformation but in the world of wealth. And we felt that the engines to do that were still not well developed in the industry. The second problem we saw was that the distribution the all of the various products offered by asset managers — thousands of ETFs, tens of thousands of mutual funds and hundreds of thousands of indexes — was getting more and more competitive and inefficient as we saw the industry shift more towards digital distribution.

What’s the business model? We have two divisions under TIFIN: One is a consumer-facing division that we call Magnifi, and we have a B2B division, which we call TIFIN Wealth that takes all [those] personalisation capabilities and works with intermediaries — financial advisers, wealth enterprises, workplaces, even other consumer fintechs, anyone who’s trying to solve the same problem. We are very open architecture. On the consumer side, Magnifi, our mission is to simplify investing success. We do that by bringing in layers of intelligence on top of a transaction or a brokerage-type capability. Those two divisions, they feed in data into our centralised client data platform and all of that is what powers digital distribution algorithms for investment managers.

Why a diversified model as a start-up? We didn’t start off this way. We continue to believe that when you start something new, you have to be specialised. But at the same time, there were thousands of fintech [individual] solutions out there, and we saw platforms that had tech debt that were broadly not up to date. So we felt there was an opportunity to build a platform of innovative solutions, and go omnichannel. So after the first year and a half, we decided to integrate everything into what has become TIFIN today.

How will rising interest rates impact your business model? There will be cost pressures for asset managers, or market adjustments which will accelerate the shift towards digital distribution and efficiencies everywhere. Just like how the Covid outbreak accelerated digital adoption, we think that you’ll see a lot more attention paid to sales and distribution efficiencies spend. So our business model, perhaps a bit counter-intuitively, benefits from more efficient distributions. We feel pretty good about anything that gets asset managers very cost-conscious, because the current spending that many asset managers do is really weak in terms of return on investment, and it’s getting worse over time.