China’s NPLs: Another financial time-bomb?

By Arthur Kroeber

Is China’s credit binge a financial time-bomb waiting to blow the country’s much-vaunted economic miracle to smithereens?

Beijing has long bet that the problem of bad loans can be solved by pushing off the day of reckoning into the future, with rapid economic growth reducing the size of the problem.

So far that calculated bet has proved a sound one.

But the unprecedented expansion in bank credit this year, coupled with last month’s decision to roll over for another decade the bonds used to finance the first non-performing loan (NPL) workout of 1999, make it a good time to submit this policy to a stress-test.

Chinese NPLs can be divided into four tranches. Tranche 1 is Rmb1,400bn removed from the “Big Four” commercial banks – Bank of China (BoC), China Construction Bank (CCB), Industrial and Commercial Bank of China (ICBC) and Agricultural Bank of China (ABC) – and China Development Bank in 1999.

Tranche 2 is Rmb1,300bn removed from the first three of the Big Four in 2004-05 as they were being prepared for international stock market listings, plus some other bits and pieces from other banks in 2007.

Tranche 3 is Rmb816bn removed from ABC in 2008, presumably preparing the ground for it to list.

And Tranche 4 consists of whatever NPLs still sit on the banks’ balance sheets, especially as a result of this year’s lending binge.

The treatment of Tranche 1 is illustrative of Beijing’s calculated bet. If the government had simply financed a write-down in 1999, the cost would have been an unbearable 18 per cent of GDP. Government revenue was only about 12 per cent of GDP at the time – so explicit recognition of these liabilities would have blown a very damaging hole in Beijing’s books.

Instead, Beijing devised an accounting dodge: the NPLs would be purchased by special asset management companies (AMCs) set up by the Ministry of Finance, paid for via a combination of cash and 10-year bonds.

The AMCs would then spend the next decade trying to recover what they could, with the final loss being absorbed by the Ministry of Finance – by which time that loss would be vastly reduced because the economy and MoF’s resources would have grown enormously.

And yea, verily, so it came to pass: the net liability to the government of that original NPL tranche at the end of 2009 will be less than 4 per cent of GDP, and less than 20 per cent of expected government revenue for the year.

Indeed, the bet has paid off so handsomely that CCB’s decision to roll over an Rmb247bn bond from Cinda Asset Management Co for another decade looks set to be followed by the other big banks.

Will the bet pay off again – or have officials simply set the timer on a financial time-bomb and decided that someone else will be in the room when it explodes?

To begin estimating the full cost of writing down China’s NPLs in 2019, we assume that recoveries on NPL Tranche 1 are 15 per cent, which the distressed assets specialists tell us is reasonable.

For Tranche 2, 72 per cent has already been written off. We assume no further recoveries on the outstanding balance of Rmb348bn, all of which seems to have been financed by the central bank.

For Tranche 3, we assume a recovery rate of 15 per cent – but this may be too optimistic. To balance that out, we make a very aggressive estimate for Tranche 4, assuming that one-sixth of the Rmb20,000bn in bank loans likely to be issued in the three years 2008-2010 will go sour.

This is bold: it means that China will have to deal with an as-yet unrecognised NPL liability of Rmb3,300bn, almost equal to the face value of all NPLs hitherto recognised (Rmb3,500bn).

So just how big is China’s NPL time-bomb? That is largely a function of economic growth rates.

Average annual nominal GDP growth of 11 per cent, 9 per cent and 7 per cent over the next decade would generate net fiscal NPL costs of 6 per cent, 7.2 per cent and 8.7 per cent of GDP respectively in 2019 – substantial, but not catastrophic.

There is no necessary reason why existing NPLs, even including bad loans arising from the 2009-2010 monetary stimulus, should threaten the viability of the system. In short, the calculated bet of letting NPLs shrivel through time and growth can safely be placed one more time.

But this bet absolutely cannot be placed a third time.

The above scenarios only work if the financial system generates no net new NPLs in 2011-2019 beyond the banks’ own ability to provision and write down.

After 2020, demographic and other factors will turn unfavourable, and the structural real GDP growth rate will fall from 10 per cent in 1980-2008, and an expected 8 per cent or so in 2010-2020, to around 5 per cent. When that happens, growth can no longer erase past NPLs – only inflation can.

The solution is that banks must start lending on a commercial basis; they must be permitted to deny credit to state entities that are obviously un-creditworthy; and enterprises that cannot sustain themselves in a market-based financial system where capital is properly priced must be allowed to go bankrupt.

If China wants to avoid a hard choice between financial crisis and rampant inflation in the 2020s, it must start fixing its financial system, right now.

Answers may be published in the Financial Times newspaper. Please include your first name (surname is not needed), email address, gender, age and job title. Want to report an offensive or unsuitable comment? Our comment guidelines, including reporting procedures, can be found here.

Dragonbeat is no longer updated but it remains open as an archive.

Readers of the Dragonbeat blog can now go to www.ft.com/dragonbeat to read the new Dragonbeat weekly column for insightful commentary and analysis on China.

Full list of the FT’s blogs