How asset managers like FinEx Asia are using AI to disrupt traditional bank lending

12-Dec-2017 Intellasia | South China Morning Post | 6:00 AM Print This Post

Banking disintermediation essentially, taking out the middle man has taken a new twist. While in recent years peer-to-peer (P2P) lending has become the poster-child for threatening banks’ lending business, a new type of hybrid disrupter is apparently starting to emerge: asset managers backed by financial technology.

One such firm attempting to cut banks out of the consumer-lending equation is FinEx Asia. The newly-licensed asset manager connects Asian investors with American consumer-credit assets, using artificial intelligence to select the loans based on risk appetite.

Founder and chief executive Maggie Ng said the company’s three funds now have $100 million under management. They are backed by a portfolio of more than 10,000 US-based borrowers who have obtained loans from multiple online lending platforms, she said without specifying which ones.

FinEx Asia is leveraging the machine-learning and blockchain technologies developed by Dianrong, a Shanghai-based P2P platform, with whom it recently partnered.

Dianrong, which was established by LendingClub co-founder Soul Htite, has a 20-strong technology team supporting FinEx Asia in loan selection using machine learning. Using Dianrong’s blockchain technology, investors can also monitor in real time how the quality of each loan in the fund has changed over its tenor, and check out a borrower’s credit data.

“By applying artificial intelligence in our risk modelling, we will now have more parameters in performing more refined analysis on consumer loans’ credit quality than those run by banks,” said Ng. She said that Dianrong’s platform is capable of approving loan transactions of up to $500 million every month.

By applying artificial intelligence in our risk modelling, we will now have … more refined analysis on consumer loans’ credit quality than those run by banks

Maggie Ng, chief executive, FinEx Asia

With 25 employees in Hong Kong, FinEx Asia was licensed last week by the city’s securities regulator as an asset manager and securities adviser. It has recently also opened offices in Singapore and Taiwan.

A consumer-loan banker for close to 20 years before co-founding FinEx Asia, Ng was most recently Citibank’s Asia and EMEA head of unsecured lending. She had also served as chief risk officer for Citibank in Hong Kong.

Ng said she prefers US consumer loans over their Chinese equivalents because they are a highly transparent asset class whose performance data is closely tracked over the long term by regulators. As a result, they are a mature asset class that has been tested through several economic cycles over the last four decades.

“With the long history of consumer-lending development, regulations of consumer lending in the US are consistent and mature. Borrowers are protected by fair lending laws and lenders also given protection with recourse to borrowers,” said Ng.

In the US, the Truth in Lending Act is a federal law that ensures consumers are treated fairly by lenders by, for example, requiring lenders to disclose credit terms in an easily understandable manner. There are also state usury laws stipulating the maximum interest rates.

Ng did not disclose which online platforms FinEx Asia’s funds invest in. The ones that are more established in the US consumer lending space include LendingClub, Prosper and Avant.

US household debt has continued to expand to record levels, with data from the New York Federal Reserve showing that aggregate household debt grew to $12.96 trillion in the third quarter, $280 billion higher than the previous peak seen in the same period of 2008 when the global financial crisis hit the US economy.

In the first quarter of this year, student loans had the highest delinquency rate among types of consumer credit, with about 10 per cent of the total loan balance more than 90 days in arrears. Credit cards were the next highest at just over 6 per cent.


Category: FinanceAsia

Print This Post

Comments are closed.