Credit risk is an unavoidable component of peer-to-peer lending, so platforms have many ways of managing it, as Peer2Peer Finance News investigates…
THE PEER-TO-PEER lending industry’s reputation for doing things better, faster and cheaper than banks is often attributed to slick tech platforms, greater transparency and a friction-free customer experience. But success in the longer term also depends on the equally crucial if less glamorous discipline of good credit management.
If you are going to lend money, you have to deal with the fact that some borrowers will not be able to pay it back.
“For any lender, managing credit risk is the number one priority,” says Michael Hoare, head of risk analytics and retail credit at one of the largest P2P lenders in the UK, RateSetter. “If you don’t manage it well then you won’t last very long.”
Managing risk, he adds, is not the same as minimising risk. Although the perils of reckless lending may be fairly obvious, excessive caution can also bring its own difficulties. Credit risk needs to be balanced against investor returns and the growth imperative.
“You have to be commercial about it, you can’t always say no,” says Hoare. “You have to maintain a balance or you can stop a lot of business.”
RateSetter uses the tech that fintech is famous for to speed up loan processing and improve the customer experience – but the prospects who don’t pass with flying colours then get the personal treatment before a decision is made.
“Our decision engine makes automated decisions, a very high proportion of our loans are decisioned in under 60 seconds,” Hoare explains. “But then there is a chunk that is passed to underwriters. Experienced people play the vital roles.”
Referrals to the underwriters are made on the basis of RateSetter’s credit scorecard model, which is itself based largely on traditional data but with a dash of the contemporary thrown in.
“The data we use is from the application form itself, credit bureaux, have we had a relationship with the lender before?” says Hoare. “We also use some contextual data, such as what browser are they using?”
That’s because Mac users have been shown by some metrics to be more credit worthy than those who use PCs.
RateSetter has dabbled with third-party data looking at everything from psychometric and behavioural profiling to social media monitoring. Hoare remains interested in these areas but is as yet unconvinced that they make a big difference to lending decisions. “It won’t replace the traditional data any time soon,” he asserts. “The death of the credit report is significantly overplayed.”
Good credit decisions matter not only to the success or otherwise of individual platforms but are also a vital part of the wider economy – especially when lending to small- and medium-sized enterprises (SMEs), says Vicky Pryce, board member at economics consultancy CEBR. “Lots of SMEs survive on lending, and alternative providers are important to them,” she comments.
But there are those who fear that the credit management of some – P2Ps among them – may be found wanting if economic trouble strikes. “There have been concerns that some [alternative lenders] are not doing their due diligence properly and may collapse in a downturn,” Pryce adds.
There are signs that such a downturn may be on the way, she cautions. “There has been a substantial slowdown in the UK economy in the first quarter, will it continue? SMEs are more exposed and so are SME lenders. Households are borrowing less and mortgage approvals are down 20 per cent.”
Any P2P platform failures as a result of inadequate credit management would be very damaging to the reputation of the sector, even though the alternative lending market is still too small and heterogenous to pose a wider risk, she adds.
“Alternative lenders don’t pose a systemic risk, the lenders are mostly SMEs themselves,” she says. “But they could certainly suffer as a result of a systemic failure elsewhere.”
One SME lender which goes to great lengths to make sure it does plenty of due diligence is Crowdstacker. Not least because it is in the business of making a small number of larger loans (in excess of £1m at a time) to more established businesses.
Its credit process is very thorough, if relatively low-tech. “How do we manage risk? By stringent due diligence and active monitoring,” says chief executive and co-founder Karteek Patel.
Avoiding borrowers in the high-risk early stages is also a form of credit management. “We prioritise quality over quantity, and focus on medium-sized businesses that are not in the start-up phase,” he explains.
After automated pre-screening which weeds out “70 to 80 per cent”, the better prospects are treated to a full financial health check, plus interviews with the management team. “It’s important to be able to judge the character of the individuals,” says Patel.
It can take two or three weeks to complete but is a vital part of Crowdstacker’s USP. “Medium-sized businesses are not all the same, that’s why banks fail them,” Patel adds. “We take the time to understand a business well.”
Over at Zopa – best known as the P2P consumer lender which is turning itself into a bank – chief product officer Andrew Lawson says that the firm’s wellknown “obsession” with customer service has driven the search for newer, better ways to take credit decisions.
“We were the first prime lender in the UK to offer pre-approval on Moneysupermarket – a huge win for the borrower,” he states.
“To do that we need to take a great credit risk decision, and to do that better than the others we need to use more advanced techniques.”
Read more: Zopa revamps credit risk assessment model
20 years ago, he says, the technique that almost all lenders used was logistic regression. “With a bunch of variables and some coefficients to the variables, that would give you a likelihood of the customer defaulting,” he explains.
“But analytical techniques have expanded, and the availability of them has expanded through things like machine learning, too.”
At least if you have the tech to run them on. “If you are a big bank whose tech is based on a 1970s Cobol mainframe – it’s just too difficult for them to execute,” he asserts.
The raw material – the data – remains familiar however, including credit bureau files and the loan application form. “We have to be quite careful so we use fairly traditional data sources,” says Lawson.
“The variables in a bureau file have not changed much in the last 20 years, but along with some of the application form data they have historically been the strongest determinants of someone’s ability to repay a loan.”
What about new data sources like social media? “They can be useful for things like fraud, but less so for credit risk because they tend not to be stable long term,” he says. So for example, your data might tell you that iPhone users are a better risk than those with an Android phone, “but if Samsung brings out a new Galaxy tomorrow, that could all change”.
The type of loan being offered also has a substantial impact on the way in which the risk should be assessed, says Neil Faulkner, founder of P2P analysis firm 4th Way.
“On some property development loans, for example, loan monitoring is as important as loan assessment,” he says. “Has the development increased in value enough that you can deliver the next tranche?”
Nor is the provision of security against a loan as cut-and-dried a get out of jail free card as it can appear. “What about liquidity risk? That is something that many P2P platforms don’t mention much,” continues Faulkner.
“But if you are lending to a farm, that is not going to be as easy to sell in an emergency as a two-bed flat in a city centre.
“Or on a bridging loan, often it’s the exit that matters, not the cash flow. All the interest can be rolled up so that the exit pays the whole lot.
“Lots of retail investors don’t really know what they are doing with risk. You don’t need to be super expert to succeed, but you do need to know enough.”
As an institutional investor in the likes of MarketInvoice and EstateGuru, German bank Varengold runs its slide rule over a lot of alternative lenders. “The main role of our credit team is the assessment of other people’s credit assessment,” says Alison Harwood, senior vice president in the firm’s London office.
And while it is undoubtedly true that great tech can play an important role in facilitating lending, it’s not the first thing they look for when they are examining a prospective investment.
“Do they have management team experience? Do they know the asset class and its pitfalls? And do they have a track record of results?” she adds.
Taking stock of a prospect is a major undertaking, Harwood explains, involving interviews with the head of credit and other key execs, plus live examples and demonstrations of their credit processes.
All that facetime and on-site due diligence helps to avoid the temptation to place too much emphasis on high-tech bells and whistles, and focus instead on the underlying fundamentals.
“We see a lot of capable tech guys and fantastic systems, but the risk in any asset class will still be there and has to be addressed,” she affirms.
That way the P2P sector should stay hale, hearty and able to keep beating the banks for a long time to come.
This article featured in the June edition of Peer2Peer Finance News. Click here to read the magazine online.