I spoke at a UBS luncheon today, entitled "Online lenders’ great future".

Joe Zhang, Nov. 2017.

Summary: The regulatory noise in recent weeks is due to populist backlash (mainly envy). But make no mistake: The sector has vast growth potential, and the underlying economics is compelling. The noise will subside over time. Expect the friendly regulatory framework to continue. As part of my book tour, I will speak again on this very topic at the Graduate School of the central bank in Beijing (Tsinghua Uni.) on 28 Nov.

Slide 1. Three types of lenders

* Unlicensed and semi-licensed: thousands of P2P operators, with 200 voluntary closures each year,

* Licensed: 11,000 traditional microcredit firms licensed by Provinces (210 of which have an online lending licenses),

* Thousands of non-P2P online or offline lenders operate mostly without lending licenses but make loans anyway, or lend via a trust structure or make bank entrusted loans.

* Online lenders equivalent to < 1% of bank assets (< 2 trn yuan vs < 200 trn)

Slide 2. Famine and feast in the industry

* Ability to access better borrowers,

* Ability to access cheaper funds,

* Ability to ramp up volumes,

* Ability to minimise delinquency

Slide 3. Regulation today and next year

* No single rule is in force,

* All rules are for consultation,

* Good guys behave but bad guys are not penalized: selective enforcement

* Potential rules to come: Online lenders must not take deposits, must get a lending license or use a qualified structure, Caps on rates, custodian banks, Caps on loan sizes of RMB 200k on personal borrowers and RMB 1m on corporate loans

* Regulation to evolve slowly

* China has the most sensible subprime credit regulation in the world, striking a balance between prudential supervision and space for innovation

Slide 4. Data data everywhere. Which bit is useful?

* All data shed light on a specific aspect of a borrower,

* PBOC Credit Bureau data most accurate but do not cover much of the subprime sector

* Zhima Credit, JD, WeChat and CTrip data are all narrowly-based. WeChat data may be of better quality but cannot be relied upon for big loans.

* PBOC tried in 2015 to gradually encompass microcredit sector data but gave up soon.

* Private-sector data services are growing to fill the gap but their quality varies.

* Hard to find credit data on borrowers’ activities in other omline lending platforms

Slide 5. The lure and traps of purchase-based loans

* Online purchases in credit analysis as a risk control mechanism are overrated,

* Qudian’s low delinquency owes more to its feedback loop with Zhima Credit (and its small loan sizes) than to its merchandise platforms,

* Lenders who targeted medical care, beauty parlors, vocational training, travel and weddings have lost big,

* Loans related to car-purchases have not seen waves of frauds, just yet, but it is early days: depreciating assets difficulties in collections,

* Home equity loans will fare better only because housing prices are still rising. Small-sized loans are key,

* Banks are eyeing home equity loans but are handicapped by CBRC

* Most SME lenders and guarantors have died a death by a thousand cuts

Slide 6. Issues for discussions

* The hangover of online binge: borrowing from multiple sources and social ills,

* Many operators pretend to have a sophisticated risk control models but are simply shooting in the dark,

* How many operators will the capital markets absorb?

* Subprime credit surge has made the banks safer and more profitable,

* Populist backlash against the sector has limited influence on regulation,

* The public suffers far more in the stock market, or due to alcohol, electronic games, bad medicine, and cigarettes,