Superior knowledge is the holy grail of the used-car market — an industry still beholden to salesmen and auction houses, where buyers fear “lemons” — and it is this that Hakan Koc, co-founder and co-chief executive of SoftBank-backed Auto1, thinks he has in abundance.

The start-up has 35,000 cars on its balance sheet and its new credit business, Auto1 Fintech, that it established with Deutsche Bank and Allianz earlier this year, is set to expand from Germany to France and Poland.

“Of course there is a lot of risk,” Mr Koc told the Financial Times. “Only the superior knowledge allows you to take controllable risk.”

The German start-up, which raised €460m earlier this year from SoftBank’s Vision Fund in one of the biggest tech funding deals in Europe, uses machine learning technology to buy tens of thousands of cars each month, then sells them to dealers in more than 30 countries. The fintech business allows it to extend credit and insurance to car dealers.

“Frankly I don’t know how a dealer in Calais, let’s say, who has a certain kind of car he’s struggling to sell . . . could possibly know that a dealer in Naples has people coming in all the time and asking if they could buy that type of car,” said George Galliers, an automotive analyst at Evercore, of the traditional used-car market. “There is room for the application of new technology.”

Auto1 was founded six years ago by Mr Koc, a former employee at start-up Home24, and his co-chief executive Christian Bertermann, who previously worked at the voucher website Groupon. Hungry for funds to buy cars, the start-up struggled in its early days to raise debt, relying on funding from early stage investors such as DN Capital and Piton Capital.

A graphic with no description

Fast-forward to 2018 and the company is awash with funding. By connecting buyers and sellers using machine learning technology, and handling car shipments, Auto1 has united two hallmarks of trendy technology companies: data analytics and logistics. However the start-up is different in one key respect from businesses such as Amazon, with which Mr Koc compares its model: it takes risk on to its own balance sheet.

In this respect, Auto1 is one of only a small clutch of start-ups that does so and is therefore exposed to downturns in consumer demand. Others include US-based Opendoor, which buys and sells houses. But the rest of the technology industry has largely stayed away from the model in favour of a platform approach, connecting users with services without taking on risk.

The balance sheet, in Auto1’s model, is “a technical function”, according to Mr Koc.

“We can see clearing prices, we can see what people sell to us for, what people are buying for . . . the algorithm comes up with a price thesis and says at this moment of time is it relevant that this seven-year-old car is green, or does it matter whether it has leather seats or not,” he said. “You have the selling side certainty, the buying side certainty and the balance sheet becomes a technical function.”

Auto1 has not disclosed annual accounts, but the SoftBank investment for a 20 per cent stake in the company valued it at about €3bn. In previous rounds it also raised money from Goldman Sachs and JPMorgan and two years ago lured JPMorgan’s European head of technology, media and telecommunications, Markus Boser, to join the company as chief financial officer.

SoftBank’s tolerance for risk is well-known, but Auto1’s expansion has caused concern to some observers. Because the start-up’s machine learning technology has been trained during a growth cycle, one investor, who spoke on condition of anonymity, said it might not be able to spot signs of a downturn. “The tech uses the marketplace to make decisions, but it will find it difficult if the marketplace changes in a fundamental way,” the person said.

Analysts of the used-car and auto-financing markets said both have proven relatively resilient, even after the financial crisis. Germany is Europe’s largest market for automobile asset-backed securities, according to Volker Laeger, senior director at S&P Global Ratings, who added that there were vast gains to be made from selling cars across borders.

If a downturn in one country caused dealers to default on loans, Auto1 could sell the cars elsewhere. “In one way or another [lenders] have to deal with the fact that cars come back to them when borrowers default,” he said. “If you broaden the region where you can sell the car then you can achieve a higher likelihood that you can sell the car at a reasonable price.”

For Mr Koc, this turns risk into an advantage. With superior knowledge of trade data and customer-buying patterns, and a platform for selling cars from defaulted dealers, he said Auto1 could survive a collapse in demand. “We are convinced that we might be the only party that can very profitably lend to car dealers in a downturn environment,” he said.

Additional reporting by Silke Richter in Berlin

Get alerts on Retail sector when a new story is published

Copyright The Financial Times Limited 2019. All rights reserved.
Reuse this content (opens in new window)

Follow the topics in this article