Can bots take the pain out of mortgages?

Robot adviser © Getty images
He won’t be all that much help, but it’s a good start

Finding a suitable mortgage can be a tough job – and it’s got tougher in recent years. In 2014, the Financial Conduct Authority (FCA) put in place the Mortgage Market Review (MMR). This was designed to prevent a repeat of the careless lending that contributed to the 2008 financial crisis. Self-certification mortgages (the so-called “liar loans”) were banned. Interest-only mortgages could only be offered if the borrower had a “credible repayment strategy”. Lenders had to consider not just an applicant’s income, but their overall financial status, with all outgoings taken into account. All of this had to be “stress tested” to ensure that a borrower would not default if rates rose.

It’s all sensible stuff. But it has also made getting a mortgage more tricky. In the year after the introduction of the MMR, 2.5 million people started the mortgage-application process, but only about half of them ended up with a mortgage. Roughly half a million were rejected, but a bigger problem was that up to three-quarters of a million simply abandoned the process, for reasons that included “overwhelming paperwork”.

This is where technology aims to help. Mortgage Gym – the company that carried out the research mentioned above – is one of a new breed of “robo-advisers” that hopes to profit by making the process of getting a mortgage less painful and a lot quicker. Mortgage Gym aims to streamline the application process. The “portal will be integrated with consumer credit files provided by Experian,” notes FT Adviser. As a result, consumers “can expect a detailed insight into the mortgages they can afford within 60 seconds based on live affordability searches from 12 of the top 20 lenders”. Meanwhile applicants will be able to upload their documents online, getting rid of the need for piles of physical paperwork to be sent to various addresses before any progress is made.

Once the initial search has been done, the borrower is passed onto a “digital network” of human brokers who can advise on which mortgage to go for. The application is then sent to the lender – at which point it will still take two to three weeks to get the loan. Mortgage Gym will make its money by charging brokers a fee for access to its users. The company – which is backed by comparison site Go Compare – plans to launch this summer.

Existing online brokers that offer similar services include Trussle, which has partnered with property portal Zoopla – and Habito, which is backed by a number of venture capitalists including Silicon Valley-based Ribbit Capital, and Funding Circle’s founder Samir Desai. Again, you input your details and a bot scours the market for a suitable product (Habito claims to use 70 lenders, while Trussle claims 90) before passing you on to a human mortgage broker.

Both Habito and Trussle seem very popular among users, with a 9.5/10 rating on Trustpilot, an online business review site. The two sites make their money by charging referral fees to lenders – the services are free at the point of use to the borrowers.And once you have your mortgage, if you keep your details up to date, both Habito and Trussle will monitor the market, and notify you if you could get a better deal elsewhere. Trussle’s founder, Ishaan Malhi, reckons two million people are “needlessly languishing” on a standard variable rate mortgage costing an average of £4,900 a year more than it should.

Frankly, beyond the idea of only having to input your details once, the benefits of using a “robo-adviser” may seem small, given that the big sticking point – the final decision by the bank – still takes time. But as the government’s “Open Banking” initiative forces banks to open up their systems to rivals and startups, we can hopefully expect to see deeper integration between front-end apps such as these and providers of banking services – and maybe, eventually, a genuinely faster mortgage decision.

In the news…

• More than $30m-worth of ether tokens (a digital currency that operates on the Ethereum blockchain platform) were stolen by hackers last week, says Richard Chirgwin in The Register. The tokens – about 153,000 in total – were stolen from digital wallets created by software developer Parity Technologies, and used by companies that had recently raised funds via initial coin offerings (ICOs). The thieves exploited a coding error which resulted from a single missing word. A further 377,000 ethers were pre-emptively “rescued” by “white hat” hackers (supporters of Ethereum) to stop the tokens from being stolen. The hackers promise to return the ethers once the “vulnerability” has been dealt with.

• RateSetter, the P2P lender, has stepped in to protect investors from losses on £80m-worth of loans spread across three borrowers, says Emma Dunkley in the Financial Times. The platform has taken over two borrowers – advertising firm Adpod and car-finance firm Vehicle Trading Group – and taken a minority stake in a third, consumer lender George Banco. RateSetter noted that the problem deals arose from its wholesale lending activities (lending to other lenders) which have now stopped (partly due to the Financial Conduct Authority deciding this was beyond the remit of P2P). RateSetter has given savers with the platform the option of withdrawing their money without incurring any fees.

The bot that can pick your next flatmate

Those not yet ready – or able – to apply for a mortgage (see story above) can still use bots to locate the perfect flatmate. Ideal Flatmate, “the UK’s first algorithm-based flatmate matchmaker”, was set up by 20-somethings Tom Gatzen and Rob Imonikhe, who note that “who you live with affects your happiness far more than the features or size of your room”. Users are asked a set of 20 questions to assess their personality, and are then matched with like-minded souls. Although most users are between 20 and 35, there is a “significant number” of over-40s, says Gatzen. The service launched in London this year and currently lists 2,000 flats and 15,000 flat-hunters.