My Blog

October 15, 2007

Knowing this rather brutal history, I think one of the main goals of the CIC should be to act as a balance sheet corrective – to take on positions that may deteriorate when China’s underlying conditions are good but appreciate when things turn badly.  At the very least this can smooth out government creditworthiness, which is a major source of instability because of the impact of perceptions of government credit on investment decisions and capital flight. 

To a certain extent that is the purpose of central bank reserves – they are there for liquidity purposes – but the CIC should manage this risk much more aggressively.  This means identifying possible reasons for a crisis in China and then to take countervailing positions.  A crisis reduces Chinese government ability to service debt, which can both cause and be exacerbated by capital flight (another, among the most damaging, self-reinforcing balance sheet structure).  If the CIC invests in a way so that its debt servicing capacity increases at exactly that time, it would reduce investor concerns and eliminate one important source of volatility.

It is a complex process to discuss all the possible major risks China faces and what the countervailing corrective balance sheet positions might entail, but I think everyone would agree that a banking crisis is one obvious disaster scenario – even if we disagree on its probability.  Chinese banks, already insolvent or barely solvent if loans were correctly marked, have seen tremendous loan growth during optimal liquidity and GDP growth conditions, and knowing what we know about how inexperienced Chinese banks are in managing major economic volatility, it is hard to imagine why they would be an exception to the almost unbroken history of banking systems making bad lending decisions in times of excess liquidity.  This risk is highly pro-cyclical because of course a slowdown that caused a rise in NPLs would also cause an economy-wide hoarding of liquidity and a contraction in lending, as in Japan in the 1990s, which would exacerbate the slowdown as well as the rise in NPLs..

I know, I know, many people who assume that there are only two banking systems in the world – that of the US and that of China – will point out that banking history isn’t relevant since the two countries are radically different. In China, we are told, the banks are state-owned, and so will not contract their lending because the government can simply order them to keep lending. 

Rather than explain why this is unlikely to be the case, I should just point out that banking crisis have actually occurred very often in countries other than the US, and in many of these countries the governments also owned the banking system.  Government ownership of the banking system (or of anything else) is almost certainly not a useful indication of immunity from crisis.

If you agree that there is a real possibility of a banking contraction, you would probably also agree that a banking contraction may cause a surge in government debt, partly to cover rising NPLs and partly because of increased fiscal expenditures to counteract a contraction (along probably with reduced tax collection).  You would also probably agree that a position in which the government benefits during a banking crisis and suffers during a period of banking improvement (in other words, a hedge to the banking system), would reduce the country’s balance sheet risk by hedging the government’s debt-servicing capacity.

There are many possible ways to hedge – for example a serious banking crisis in China would almost certainly see a rise in the value of US Treasury bonds – but of course the simplest hedge would be to go short Chinese bank stocks.  We cannot expect the CIC to take a massive short position in Chinese banks stocks (which the government can do more efficiently anyway by privatizing its shares), but we can argue that for it to take a massive long position just does not make any sense from a balance sheet perspective. 

As I have written elsewhere, Chinese bank share prices contain a lot of time value and very little intrinsic value, so they are enormously susceptible to changes in expectations.  During any sort of economic or banking contraction, experience from other developing countries with high-time-value banks suggest that drops in value of as much as 50-75% are not implausible, so if China were to experience a banking crisis, one consequence would almost certainly be a collapse in bank share prices.  It would not boost investor confidence much to see the value of the CIC’s investment drop – perhaps by as much as 50% or more – just when the country was experiencing trouble.

If the CIC is truly to be useful to the long-term growth prospects of China, it should not simply be an investment fund but should combine some of the interests of a central bank (stay liquid in case of a repayment crisis), a stabilization fund (smooth out the impact of commodity price volatility on the economy), and a balance sheet stabilizer.  Under none of these three cases should it invest in Chinese banks.

I realize that on the one hand China has such high levels of reserves that protecting the value of reserves in a crisis may not seem like a particularly pressing need, and on the other hand that through its purchases of bank stocks the CIC is not increasing government exposure to the banking system – it is merely receiving the transfer of existing ownership – but I still do not think it should own the banks.  The ownership of the banks should be funded by domestic borrowing, not by reserves, and good risk management practice should always be put into place not when conditions are bad but precisely when conditions are so good that risk management seems like a stupid idea.

It is always dangerous to make predictions, but I have absolutely no fear of making one prediction.  As JP Morgan famously said when asked which way the market was expected to move, markets will fluctuate, and good times will inevitably be followed by bad times.  Developing countries with poor governance frameworks, unstable banking systems, weak information disclosure and rigid political structures (sound familiar?) have a history of veering violently from good times to bad times, and balance sheet structures that exacerbate volatility are always one of the prime culprits.  China could become a real innovator in developing country liability management if it used the CIC to attempt to correct these balance sheet imbalances, rather than exacerbate it.  Of course that means acknowledging the possibility that things can go dramatically wrong, and this is not always an easy idea to sell to a politician.

Comments (6) for "The CIC should not invest in...

Michael -- good points as always, tho in this case, the cart seems to have left the barn. the CIC is going to start out with a lot of exposure to Chinese banks via Huijin. part of me wonder is shifting the big paper profits on huijin's investment in the banks over to the CIC is a means to offset expected paper losses on its $ holdings from the expected RMB appreciation. its rmb balance sheet need not look bad so long as it has a lot of bank stock bought low that can be valued high ... More generally tho, my read is that the CIC seems interested in buying assets that are highly correlated with chinese growth, notably resource stocks along with the banks stocks, rather than following your advice ... By bsetser - 10/15/2007 2:37 AM

Brad, I think the investment strategy arises from a combination of factors and many of them (most of them?) do not necessarily make economic sense. This is as much about internal politics and controlling resources as it is about anything else. You may be right that shifting all those unrealized bank profits to the CIC may be about protecting them from more Blackstone-style criticism, but it is hard to know one way or the other. One thing about which I am pretty sure, however, is that those paper profits will evaporate very quickly when the economy turns. I would guess that your instinct is correct -- much of what they buy will be highly correlated with their own growth, which may seem like a great strategy now but will lead to losses just when they can least afford it. Still, as you know, they won't be the first government to design their balance sheet in a way that allows them to double up their bet on their expected growth. I think only the Chileans have really hedged their balance sheet. By Michael Pettis - 10/15/2007 11:21 AM

In order to run an economy you need different organizations doing different things. CIC can't be all things to all people, and CIC *shouldn't* be all things to all people. The way that CIC has been structured seems to make a lot of sense to me. It's a diversified pension fund and its role as shareholder of the Chinese banks will be as a shareholder trying to maximize value. The role of regulating the banks and preventing systematic crisis has do be done by someone *other* than CIC (like the PBC). The job of a shareholder and the job of a risk regulator are fundamentally in opposition to each other, and you get a mess if you try to have the same institution do both. The thing about shareholders is that they cannot and should not be intimately involved in the day-to-day running of a bank or for that matter any other corporation. That creates too many conflicts of interests. Shareholders do not have the incentive to put in too many risk control measures because controlling risk can eat into profits, and shareholders's liability are limited in case the bank goes under. If you allow shareholders to control day to day risk management, you will end up with a mess. The people you want driving the day to day risk management are the regulators who are going to be cleaning up the mess if something bad happens. By Twofish - 10/15/2007 7:38 PM

I had some previous objections to CIC taking over Huijin but these evaporated once I say that they were also bringing the assets of the National Social Security Fund and that the director was NSSF. It's pretty clear that they are modeling CIC after Calpers, which is why CIC should *not* be involved in macroeconomic regulations of banks. Something that the CIC managers might see once this thing gets off the ground is that CIC might realize that it is too heavily invested in financial services and too lightly invested in something else like manufacturing. At this point CIC could arrange some deal to do a stock swap and sell some of its bank holdings in exchange for holding in some other unrelated industry. As long as CIC remains an investment company, it can do this. If you put any sort of non-investment fund responsibilities into CIC, then it can't do this. By Twofish - 10/15/2007 7:50 PM

I think you miss the point, Twofish. This is not about how the CIC should regulate the banks. It is simply to say that the CIC must define its purpose -- like any fund should -- and it is hard to believe that its purpose is to enhance government credit volatility. Far smarter to act to minimize volatilty -- which benefits the country much more than almost any other use of external funds besides those used for normal central banking purposes. You don't maximize shareholder value by ignoring the relationship between assets and liabilities -- on the contrary, one powerful way to maximize shareholder value in a volatile economy (or company) is by minimizing financial distress costs, and the CIC can play a useful role here. Doubling up on the "China" bet doesn't make much sense to me, but of course almost every developing country has done this during the boom times and regretted it only after the boom reversed. As Edmund Blackadders'retainer would have put it: It is the way of things... By Michael Pettis - 10/15/2007 8:28 PM

Pettis: It is simply to say that the CIC must define its purpose -- like any fund should -- and it is hard to believe that its purpose is to enhance government credit volatility. It's purpose is to serve as an investment corporation to make sure that China has enough reserve to pay the pensions that are promised people when they retire. The logical thing for CIC to do in order to hedge some of those risks is to sell some of its holdings in the big banks in exchange for ownership in other sectors, since from CIC's point of view, this would be an exercise in portfolio diversification. I think that we may be in agreement that it is unwise for a government investment corporation to be heavily invested in financial services, but the issue here is that we may be in disagreement about what happens next. I don't see the transfer of Huijin to CIC to be a signal that the Chinese government wants CIC to manage the banks forever and ever, but rather then transfer of Huijin to CIC creates a structure in which CIC can change the ownership structure of the banks. As long as Huijin controlled the banks and the MOF controlled Huijin, there would not be the institutional structure to have pretty radical changes in bank ownership. Huijin couldn't plausibly swap Bank ofChina shares for some other sector. With the bank share owned as part of a diversified investment company, these swaps become possible. I suppose I agree with your analysis of credit volatility and the problem that you brought up, but I think that moving the banks so that they are "owned" by CIC is part of the solution. Selling parts of the banks in exchange for counter-cyclical sectors seems like a good idea, as does boosting foreign ownership to spread the risks of a downturn and getting some cash now while things are hot. But without putting the banks under CIC, there is no institutional framework to do this. By Twofish - 10/16/2007 12:39 PM