Fosun-backed AI-driven adviser sets its sights on China’s US$7.84 trillion interbank bond market
Israeli company BondIT hopes to leverage Fosun connection as well as partnership with Chinese financial data provider Wind as it expands its coverage
BondIT, an Israeli artificial intelligence-driven bonds adviser backed by Fosun Group, said it will speed up plans to include the 51 trillion yuan (US$7.84 trillion) Chinese interbank bond market as part of its “algo-driven” portfolio building tools this year.
Donald Chan, head of Asia-Pacific at the company, said it intends to leverage its existing shareholder relationship with Fosun and expand its service coverage to China’s vast domestic yuan bond market.
First launched in Asia in August 2015, the Herzliya, Israel-based start-up has been targeting the private banking arms of global investment banks and asset managers in the region with its portfolio building tools. Using machine learning, BondIT helps private banking advisers and portfolio managers to search their own internally approved bond shortlists by matching specific investment criteria – such as targeted yield levels, credit ratings, duration, geographic and sector allocations – to come up with a recommended bond portfolio for clients through automation.
In October, Fosun invested US$14.25 million in series-B funding in the company and has since become a major shareholder with representation to BondIT’s board. Chinese financial data provider Wind also has a partnership with BondIT.
“As a major shareholder of BondIT, Fosun, together with Wind, will bring their investment experience and China market expertise to BondIT. The combination of financial market and bond data know-how for China helps us design a tailored product for both domestic investors and foreign investors in the China interbank bond market,” said Chan.
Most BondIT’s 35 employees are based in Herzliya and focus on research around data science and algorithm development. Three staff members, including Chan, are based in Hong Kong and Singapore for expansion of its business in the region.
Chan said, on average, an adviser at a private bank or a relationship manager could reduce the time spent on building a portfolio comprising 20 different bonds from three hours to merely a minute.
The company, founded in 2012, did not provide details about clients, or whether it was profitable. However, it did say it counted “top-tier wealth managers, which are among the top 10 private banks in Asia, that operate in multiple jurisdictions” as its clients.
Not all asset managers, however, believe that AI works well for the fixed-income market, a largely over-the-counter market where trading can often be illiquid.
A portfolio manager working for a French investment bank said the idiosyncratic nature of the bond market – where every bond is different even by the same issuer, and the fact that there is a lack of historical data for tracking performance of each individual issuing entity – makes it challenging to build models around fully optimising portfolio returns through algorithms.
The latest data by the Bank of International Settlements shows that as of the second quarter of 2017, international debt securities outstanding totalled US$22.7 trillion, comprising securities issued by banks, financial and non-financial companies, government and international organisations. The bank defines international debt securities as debt securities issued outside the local market where the issuer resides.