Chinese Banks Are Finally Taking Retail Banking Seriously. Is It Too Late?

The market has already been handed over to the internet companies.

Over the past month or so, 32 A-share listed banks have released their 2018 financial reports one after another.

China's banking industry has hit a milestone: for the first time, two banks — China Merchants Bank and Ping An Bank — have gotten more than 50% of their revenue from retail (personal) banking.

China Merchants Bank, notably, also posted the highest return of any bank with assets over a trillion yuan, beating out the "big four" — ICBC, Agricultural Bank of China, Bank of China, and China Construction Bank.

Globally, this wouldn't be a big deal. Retail banking, centered on individual customers, generates roughly half of all bank revenue worldwide, with the rest coming from corporate banking, investment banking, and the like. A retail-dominant revenue mix is generally seen as one hallmark of a modern commercial bank.

But not in China. This is the first time any bank here has crossed that threshold since the banking sector began marketizing in 1995.

It's not just China Merchants Bank and Ping An — the big four state-owned banks still get more revenue from corporate banking, but they too are tilting increasingly toward retail.

According to McKinsey, Chinese retail banking revenue grew 23% a year between 2009 and 2016. The firm optimistically projects that by 2020, China's retail banking market will reach 3.2 trillion yuan in revenue and 32 trillion yuan in loan assets, making it the world's second-largest retail banking market after the United States.

China's banks are finally taking retail seriously — something they once looked down on.

Every domestic bank is moving into retail, but they're taking different paths

A bank's profit model, boiled down, is simple: trade interest for corporate and individual deposits, then lend that money out to other companies and individuals at a higher interest rate, pocketing the spread. That relationship with customers then lets banks sell them wealth management products and asset management services, earning fees on top.

But China's banking and insurance regulator bars banks from both lending and investing directly, which deepens their reliance on interest income. Last year, the 30-plus listed banks got 70% of their revenue from the interest spread. By comparison, Wells Fargo — widely seen as the best-run retail bank in the world — splits its revenue roughly evenly between interest and non-interest income.

Running on interest income alone, even the big four's retail share had climbed to nearly 40% last year. But they and the joint-stock commercial banks — led by China Merchants Bank and China Minsheng Bank — have taken two very different paths into retail.

The big four's rising retail share comes mainly from more mortgage lending. The joint-stock banks, by contrast, lean on credit cards and consumer loans. In 2018, mortgages accounted for 80% of the big four's retail revenue, with credit cards and other loans making up the rest.

Mortgages are the least frequently used form of consumer credit there is — most people buy a home only once in their life. By the time a 30-year mortgage is paid off, the borrower is in their fifties or sixties, and by then a bank might not even be willing to help finance a home for their son or daughter.

Mortgage rates are also much lower than other short- and medium-term consumer loans. But because loan amounts are large and default rates are low, the big four — which enjoy a low cost of capital — gravitate toward this business anyway. At the end of 2018, China's total credit card credit line stood at about 15.4 trillion yuan, with 6.85 trillion yuan in outstanding balances. Looking bank by bank, the big four state-owned banks and Postal Savings Bank, which have never leaned heavily on credit cards, still have credit card loans making up under 5% of their loan books — below the industry average of 5.03%.

The exception among the state-owned banks is Bank of Communications, whose credit card share of loans rose from 9.2% in 2017 to 10.8% last year.

The joint-stock banks, lacking the big four's cost-of-capital advantage, also do mortgages, but mortgages typically make up less than 40% of their retail revenue — credit cards, consumer loans, and other personal loans dominate instead. The People's Bank of China first disclosed domestic credit card data in its 2006 Payment System Report, when there were fewer than 90 million cards in circulation, under 6% of all bank cards. Less than three years later, in 2009, the number of credit cards had doubled to 185 million.

Last year, credit cards' share of loans rose broadly across joint-stock commercial banks. Ping An's credit card share rose 6.2 percentage points to 24.2%, the biggest jump, followed by China Everbright Bank, China Merchants Bank, China Minsheng Bank, and Shanghai Pudong Development Bank.

Data compiled by Tianfeng Securities shows credit cards are a major revenue source for joint-stock banks: China Merchants Bank, Shanghai Pudong Development Bank, China CITIC Bank, and China Everbright Bank booked credit card revenue of 66.7 billion, 55.3 billion, 46 billion, and 39 billion yuan respectively — roughly 25% to 33% of total revenue.

Beyond mortgages, short-term consumer loans have also grown quickly in recent years. By the end of February this year, outstanding short-term consumer loans among Chinese households reached 8.41 trillion yuan, up 20.24% year over year. Across all of 2018, banks sold roughly 110.779 billion yuan worth of consumer loans repackaged as wealth management products.

But much of this consumer lending may not actually be funding consumption — it may be flowing into the housing market instead. Data compiled by Caixin shows the rise in short-term consumer loans tracks closely with tightening mortgage policy. Down payments for home purchases in China rose by 3.4 trillion yuan between 2016 and 2017, while consumer loan issuance grew by roughly 3 trillion yuan over the same period.

Beyond expanding credit supply, new asset-management rules have also pushed commercial banks — somewhat passively — to ramp up sales of personal wealth management products. By the end of 2018, non-principal-guaranteed personal wealth management products made up 87% of the total, up 20 percentage points from the year before.

Banks used to make money too easily to bother with retail

China's commercial banking sector got a late start, tracing back only to 1995 — under 30 years ago. That year, the State Council transferred the policy-lending duties that state-owned commercial banks had previously carried out to the China Development Bank, the Export-Import Bank of China, and the Agricultural Development Bank of China.

Professor Wang Yijiang, a former World Bank advisor, has summarized that by 2000, Chinese banks generally suffered from high ratios of non-performing assets, weak profitability, and outdated corporate governance. Their main source of income was the spread between deposit and loan interest, and their performance metrics were essentially all built around corporate banking.

In terms of asset scale, the big four state-owned banks had a built-in advantage. Bank of China was long the main channel for foreign capital entering China; China Construction Bank managed the national housing provident fund accounts; ICBC managed the national social security fund accounts; and Agricultural Bank of China, with its deep reach into lower-tier markets, could better control its deposit costs. More importantly, these four banks were long the default choice for state-owned enterprises and government and public institutions to open their primary bank accounts — an advantage in gathering deposits that other joint-stock commercial banks simply didn't have.

Back then, financing in China's financial sector ran mainly through direct bank lending, rather than the indirect financing — via stock and bond markets — that dominates other developed economies. According to figures released by the People's Bank of China, bank loans made up 83% of total financing for domestic non-financial institutions in 2008, with the rest raised through bond and stock issuance.

A staffer who joined China Construction Bank in the 1990s once told QDaily that back then, the bank didn't even have a personal finance department yet — just a savings section, and voucher audits at every branch had to be forwarded to the back office for entry and review. For other fee-based businesses centered on individual customers, like funds, insurance, and precious metals, the bank barely had any targets at all. "I don't remember anyone actually buying funds — it was all employees buying them themselves," the staffer said.

In 2003, corporate banking made up 57% of China Construction Bank's revenue but contributed 74% of its gross profit.

Because early retail banking was unprofitable, for a long stretch Chinese banks expanded their retail footprint mainly through private banking services instead. But even catering to the wealthy didn't start until 2005, when Bank of China became the first to offer private banking services. China Merchants Bank's private bank followed two years later.

As in most of the world, personal wealth in China is distributed like a steep pyramid. China Merchants Bank's 2018 annual report shows that 30% of the personal assets it manages belong to just 0.06% of its clients — those at the very top of the pyramid hold an average of more than 28 million yuan each. Meanwhile, the bottom 98.12% of clients hold just 19% of the wealth, averaging 10,500 yuan each.

Average client wealth varies across other domestic private banking arms, but the overall distribution structure looks much the same as China Merchants Bank's everywhere. For instance, as of mid-2018, average assets per private banking client stood at 28.3384 million yuan at China Merchants Bank, 17 million yuan at ICBC, 19 million yuan at Shanghai Pudong Development Bank, and 10 million yuan at China Everbright Bank, with each bank's private banking client base ranging from 20,000 to 80,000 people.

Separate research data from Industrial Bank shows that the number of Chinese households with more than 6 million yuan in investable financial assets — stocks, bonds, futures, and the like — grew at 21% a year between 2007 and 2016, surpassing 2.1 million households by 2016.

As the easy money dried up channel by channel, banks began pivoting to retail

According to People's Bank of China governor Yi Gang, state-owned commercial banks carried heavy policy-lending obligations from their very founding, and most of that money went to inefficient state-owned enterprises, eventually turning into non-performing assets that were nearly impossible to recover.

"The sheer scale of non-performing loans left state banks severely undercapitalized," Yi wrote in a working paper. By September 2001, non-performing loans at the big four state-owned banks had reached 1.8 trillion yuan, or 26.62% of total loans. Add in the 1.3 trillion yuan in bad loans that had already been carved out earlier, and the combined non-performing loan ratio approached 45%.

That China's banking sector managed to keep going — and even keep printing money — was largely thanks to China's rapid economic growth and high savings rate. The high savings rate let banks gather deposits cheaply, while high levels of investment let them lend at higher rates, producing a wide interest spread.

Between 2002 and 2008, China's nominal GDP climbed at a compound annual growth rate of 16.5%. Over the same period, the national savings rate briefly approached its later historical peak of 51.55%.

The first wave of reform started in the early 2000s. It centered on improving corporate governance, bringing in foreign shareholders like Goldman Sachs and Citigroup, and carving out the massive pile of non-performing assets. Even though those bad loans carried an implicit government guarantee, their sheer scale posed serious financial risk.

After the bad-asset carve-out, the State Council pushed forward with converting banks into joint-stock companies. The banks received capital injections from Central Huijin, and by 2008 had all listed on the mainland A-share and Hong Kong H-share markets. "Discipline theory" from capital markets then kicked in: as public companies, Chinese banks now had to disclose their full financials, and how efficiently they converted shareholder investment into profit became a key metric capital markets used to judge a bank's investment prospects. To boost that metric — return on equity — the most direct route was to expand the asset base, grow net interest and fee income at an even faster clip, and keep non-performing loans under control.

But as the conditions that once let state banks earn easy money gradually broke down after 2000, they were forced to pivot toward retail.

First, according to research by Fan Gang, director of the National Economic Research Institute and a professor of economics at Peking University, state-owned enterprises' share of China's total industrial output fell from about 50% in 1998 to around 25% by 2011, as the private sector and foreign capital gained ground. China's 2001 WTO accession also opened the door for foreign banks to operate domestically — that year, Nanjing Ericsson repaid 1.99 billion yuan in loans from Chinese banks and instead set up a new factoring arrangement with Standard Chartered Shanghai and HSBC Shanghai.

Second, once China's economic growth slowed, the debt built up during the earlier expansion swelled the pool of non-performing assets, making banks more cautious about lending. Starting around 2006–2007, interest coverage ratios (earnings before interest and tax divided by interest expense) turned downward across the industrial, mining, manufacturing, and power sectors that had been the main engines of GDP growth, falling more than 20%. The overall interest coverage ratio for A-share listed companies fell from 9x in 2004 to 7.5x in 2008.

By 2009, nearly every A-share listed bank saw its non-performing loan ratio rise from the year before.

Third came interest rate liberalization, which began in 1994 and expanded significantly by 2004 — though its effects took longer to show up. A China Banking Regulatory Commission working paper shows that the interest spread at China's 16 listed commercial banks didn't really start shifting until 2013, when deposit rates rose faster than lending rates, narrowing the spread for two straight years. One driver was China's further loosening, starting in 2012, of the ceiling on yuan deposit rates and the floor on lending rates.

Finally, as rate liberalization eroded banks' low-cost deposit advantage, slower economic growth shrank corporate lending and raised default risk, and internet finance — led by Ant Financial and Tencent's Tenpay — disrupted the market, China's banks, sitting on 1.3 billion customers and a 50% national savings rate, found it increasingly hard to keep up with the changing market.

The real threat to banks came from internet companies

Retail banking in the 2000s and earlier ran mainly through branches, offering deposits, loans, and remittances. Services were highly homogeneous — customer experience barely factored in at all — and banks competed mainly by expanding their branch networks to capture customers.

Thanks to their branch-network advantage, the big four state banks could earn substantial income from fees alone. In 2000, net fee income made up over 19% of total revenue at ICBC, China Construction Bank, and Bank of China.

After 2003, banks kept adding new fees — dozens of them: charges to open an online banking account, charges for the USB security token needed for online payments, and a small monthly account-management fee if your average balance fell below a certain threshold.

Alipay, founded in 2004, first forced banks to modernize their payment systems simply by being free and offering a better user experience. But it wasn't until Yu'ebao launched in 2013 that an internet company posed the first real threat to the banks.

Yu'ebao is essentially a money market fund — easy to deposit into and withdraw from, and when it launched, it offered a higher yield than bank wealth management products. Banks had actually been selling similar products all along, but staff got no commission for selling them, so they never bothered pushing them on customers, and hardly anyone knew they existed.

Banks' wealth-management sales had long centered on third-party insurance and brokerage products, because those paid out the highest rebates to banks — and to the staff selling them. But complaints about opaque bank sales practices kept mounting, until China's former banking and insurance regulators issued new rules around 2015 requiring audio and video recording of staff selling third-party insurance products, plus a seven-day cooling-off period for investors.

Even today, some state-owned bank branches still chase this kind of income to hit profit targets handed down from above.

At one big-four branch in southwest Shanghai, nearly every time a customer opens a savings account, staff push them to activate extra services too — mandatory online banking, mobile banking, even a precious-metals trading account. Some of these bundled services require an initial deposit just to activate.

By contrast, Yu'ebao's superior user experience got 100 million people to park more than 500 billion yuan in it within a year of launch.

Yu'ebao then "lent" that money right back to the banks — except now, for the same money that banks could once obtain through a 0.35% demand deposit rate, they had to pay more than ten times that in interest after Yu'ebao got involved.

Yu'ebao pulled deposits away from the banks, then lent that same money back to them at a much higher rate. In the process, it also took away the customer relationships that used to belong to the banks. From there, once Yu'ebao started selling funds and other wealth management products, the banks were cut out of the picture entirely.

Having neglected retail for years, the banking industry effectively handed pricing power over wealth management products to the internet companies.

Banks are finally serious about retail — but is it too late?

China Merchants Bank's two mobile apps now have 100 million cumulative users combined. One focuses on basic banking and wealth management services; the other looks like a mashup of JD.com, Xiaohongshu, Alipay, Yu'ebao, Meituan-Dianping, and WeChat official accounts — but it's unlikely to out-execute any of the internet companies at any single one of those things.

That's partly because user habits are already locked in. If you want to find a restaurant, there's no real reason to open a bank app instead of Dianping. Building a genuinely good internet-style user experience also depends heavily on who you can hire — and banks' status as state-owned enterprises locks in rigid pay scales and rank structures that make it hard to recruit top programmers.

That's not to say banks are necessarily doomed to lose to the internet companies. After all, Yu'ebao — the product banks feared most — has already shrunk substantially under central bank oversight.

The consumer-loan asset-backed securities financing that Ant Financial later built was also temporarily halted by regulators. When it reopened in 2018, banks entered this market. Alipay stopped raising financing directly and instead became something more like a financial advisor — charging technology fees and providing access to lending channels — handing part of the interest income over to banks in exchange for banks funding the consumer loans.

The end result, after all that maneuvering, is that users still follow Alipay. Banks, meanwhile, earning the interest, now have to compete with each other just to access Alipay's traffic. The real leverage over users still sits with Alipay, which can keep selling new wealth management or insurance products whenever it wants.

Even when an internet company's foray into finance fails, the money doesn't necessarily end up with the banks. The clearest example is Meituan-Dianping: it bought a payments license and took a stake in a private bank, but its real money still comes from food and food delivery. On one side it shows consumers restaurant information; on the other it provides delivery. Based on those two services, Meituan charges merchants a commission and online marketing fees, and charges consumers a delivery fee.

Once users are used to finding restaurants on Dianping, a restaurant that wants those users has no choice but to pay for advertising there — which is exactly why restaurants with mediocre reviews often rank near the top of search results.

In the end, across the entire dining and food-delivery chain, the money still ends up with Meituan-Dianping.

Getting users to form habits doesn't happen overnight. Wells Fargo was one of the few American banks to come through the 2008–2009 financial crisis largely unscathed. By the end of 2017, it had over $2 trillion in total assets, the largest of America's four biggest commercial banks.

Unlike most American commercial banks, which pivoted toward investment banking after the 1980s, Wells Fargo has always kept retail's share of revenue high. Since 2000, retail has consistently made up over 50% of its revenue, hitting as high as 70% in some years.

Wells Fargo was an early adopter of digital tablets in its branches, letting customers choose their own services and tailoring recommendations to them based on backend customer data. It also used ATMs to build an emotional connection with customers — on a customer's account anniversary, for instance, the ATM screen would display a congratulatory message. Wells Fargo also launched an "Advantage" blog aimed specifically at fund customers, offering news coverage, investment-trend analysis, and advisory content.

Chinese banks have thrown a lot of money at this in the short term — air miles, hotel points, loyalty points, cashback — but none of it has gotten most users into the habit of opening a bank app. If anything, it's mainly attracted a wave of bargain-hunters gaming the rewards.

All of this takes a long time to build. Now that internet services have already worked their way into everyone's daily life, banks are just starting to think about these things — whether it's too late is anyone's guess.

Charts by Gong Fangyi.

Cover photo taken by the author.