Brad Setser

Follow the Money

Cross border flows, with a bit of macroeconomics

Print Print Cite Cite
Style: MLA APA Chicago Close


Hmmm – the PBoC still manages the reserves shifted to the state banks

by Brad Setser
September 23, 2006

The New York Times (emphasis added):

The People’s Bank of China, the country’s central bank, transferred $15 billion of the country’s foreign exchange reserves to the Industrial and Commercial Bank last year; a year earlier, the central bank transferred $22.5 billion apiece to the Bank of China and China Construction Bank, and both have since conducted initial public offerings here.

…  Although Chinese government officials continue to oversee the management of these foreign exchange reserves, the banks have been allowed to count the money as capital and have used it to offset large write-offs of bad loans. But Chinese banks have engaged in a surge of new loans over the last year that have diluted the proportion of nonperforming loans in their portfolios but have also raised questions about whether another large crop of bad loans may appear in the future.

There is a reason why I generally count the $60b in Chinese reserves shifted (nominally) to the state banks as part of China’s reserves, and add $60b to my global total (warning: the reserves link requires a RGE subscription).  It sure seems like SAFE – not the banks – manages that pool of money.

One point that I left out of my earlier post: when the PBoC shifted its reserves to the banks, it formally shifted the reserves to the state holding company that manages the PBoC’s investment in the state banks, and the PBoC in return received a claim on the holding company (I am not sure if the holding company pays interest to the PBoC though).  That kept the PBoC from taking a loss – if didn’t just give the banks its reserves, leaving it with more liabilities than assets.

Remember, those reserves (an asset) are offset by liabilities (cash, sterilization bills) on the PBoC’s balance sheet.   And any financial institution that gives its asset away will eventually have problems.  Giving reserves – the PBoC’s asset — to the state banks doesn’t eliminate the liabilities that the PBoC issued when it initially purchases the reserves. 

Of course, the PBoC now has so many assets (probably over a trillion now, not counting the $60b) than in some sense this doesn’t matter too much.   $60b is small …

But it does highlight some of the difficulties using the PBoC’s reserves to finance bank recapitalization – even setting aside the fact that the banks really need RMB-denominated assets to match their RMB deposits, not dollars.


  • Posted by HZ

    And the PBoC’s liability (banknotes or RMB) is in theory any way. There is no promise for you to get anything back in return for tendering your RMB to the bank. With the currency not convertible – not even foreign bank notes are promised.

  • Posted by Joseph Wang

    PBC is a central bank, and so I don’t think it makes sense for it to have a balance sheet any more than it does for the Federal Reserve to have a balance sheet.

    I suspect what the NYT means is that the PBC still “manages” Chinese foreign exchange reserves in the since that it still has regulations on foreign exchange.

    Also PBC doesn’t have any ownership stake in the banks. The banks are owned by Central Huijin which is owned by the Ministry of Finance. Central Huijing is basically a management company and has no assets of its own.

    When the PBC issues debt, it is acting as the agent of the Chinese government, and governments don’t have balance sheets.

  • Posted by Joseph Wang

    Government entities like the US government or the Federal Reserve just have different accounting than for-profit entities like Citibank or JP Morgan. In particular government agencies don’t have an accounting of assets and liabilities.

    One way of looking at Chinese economic reform is to figure out what entities would have balance sheets (the state banks, and SOE’s) and which one’s wouldn’t (the PBC, the State Council).

    The state banks are supposed to be for-profit entities and so they have balance sheets. The PBC is a central bank, so it doesn’t.

  • Posted by bsetser

    JWang — in my view, both central banks and governments do have balance sheets …

    the asset of most governments is its ability to tax more than its spends on things other than interest, and thus run a primary surplus. that offsets its liabilities. Google IMF and balance sheet approach or setser and balance sheet approach for the paper … and let me know what you think.

  • Posted by Steve Waldman

    Joseph, the Federal Reserve does publish a balance sheet, weekly, under the name “Consolidated Statement of Condition of All Federal Reserve Banks” in the H.4.1 release. The question of who does or does not “have” balance sheets is either a philosophical or political one. Categories like “non-profit” vs “for-profit” don’t cut it — private non-profits often publish balance sheets for example. Neither does government/nongovernment. Some government organizations publish balance sheets, and some don’t. PBoC publishes a balance sheet, though the latest a quick search turns up is for 2004. Thinking about government balance sheets writ large means thinking about whether accrual accounting is appropriate for governments, what kind of asset a tax base represents, what kind of liability, or equity claim, or none-of-the-above a modifiable entitlement program entails, etc. The lack of published balance sheets is a matter of lack of consensus on these topics, and the deadly politics of suggesting even tentative categorizations. But for subunits of government, especially quasi-independent units whose primary asset isn’t a changing, discretionary budget allocation, balance sheets are no problem, whether they are published or not.

    HZ’s point about that many central bank “liabilities” are analytically more like equity (in that currency holders can’t force any sort of redemption) only underlines Brad’s thrust that central bank balance sheets matter. The value of non-control equity is always a matter of trust (do the controlling parties intend behave in a way that maintains the value of my stake?) and ability (are the controlling parties capable of maintaining the value of my stake?). Central bank balance sheets speak to the second concern.

    I think the right way to judge a central bank’s balance sheet strength is to ask “what is the present value of all assets and liabilities that could be realized without taxation, new borrowing, or changes in the money supply?” For China at the end of 2004, by my reading of PBoCs balance sheet, that value was about 2.6 trillion CNY, or $325B in 2004, when FX reserves were worth about $587B. If the ratio doesn’t change and China hits reserves of $1T, that implies a present value of 4.4 trillion CNY or $554B. For comparison, an analoguous calculation for the US Fed as of last week puts its present value at $22B. I think Brad is right to point out that the potential volatility of PBoC’s in RMB terms is a problem. But it’s the raw strength of PBoC’s balance sheet that puts it in a great position to fix the banking system, if necessary by writing a check, without drawing on taxpayers or provoking inflation. The currency mismatch sucks, but the magnitude is a saving grace.

  • Posted by Steve Waldman

    (I should mention that my “pro-forma equity” guestimate above would be very sensitive to FX valuation change. A 30% appreciation of CNY against PBoC’s FX reserves would wipe out more than half of that surplus. If anyone wants to see where these numbers came from, I’ll write it up. These are back-of-the-napkin estimates based on a methodology I made up on the spot, so take it with boulders of salt.)

  • Posted by Joseph Wang

    A balance sheet for a commercial enterprise is very different animal from the “balance sheets” that we are talking about for a government. It’s not a theoretical item, but rather the central document of the company.

    For a commercial enterprise, the balance sheet is life or death, something that you have in mind day to day, and the whole point of the enterprise is to maximize the bottom line. What we are talking in terms of governmental “balance sheet” as far as assets is far more theoretical.

    It’s impoosible to run any sort of commercial enterprise for any extended period of time in a market economy without a balance sheet, and that provides a marker between commercial and non-commercial enterprises.

    Also in considering the “assets” of PBC simply can’t be considered in isolation, since PBC is acting as the agent of the Chinese government, and any liabilities issued by the PBC are by law claims against tax revenue and other assets of the Chinese government rather than claims against the PBC itself. This is radically different from the state banks whose liabilities aren’t legally claims against the Chinese government.

  • Posted by HZ

    Here is the PBoC balance sheet you were looking for:

    Your IMF paper seems to be saying that a large FX reserve like China would prevent externally triggered financial crisis (at a large cost of holding inefficient assets). Since the domestic balance sheet net out it does not address the potential problem in the Chinese case — financial sector balance sheet deterioation. But that is within the single government’s ability to address so it is probably out of the purview of IMF. It is more of a cost-benefit analysis with cost paid on moral hazards grounds, so there is no reason for it to devolve into a crisis if properly managed?

  • Posted by HK

    Brad and Joseph–Foreign exchange reserves must be liquid and usable. $60 billion transefered to state-owned commecial banks appear to be illiquid or unusable even if managed by the PBoC. So, I tend to think that these $60 billion should be excluded from China’s fereign exchange reserves.

    The balance sheet of the PBoC should be consolidated with that of the Chinese government since the fomer is a part of the latter. But this is not always the case–for example, the Centarl Bank of the Philippines is constitutionally independent, and the government cannot save the Bank even if it fails (its former central bank actually failed). The Bank of England until the 19th century and the two Banks of the United Staes in the 19th century should also not be classified as a part of the respective gevernments.

  • Posted by Steve Waldman

    Joseph, Even for a private, commercial entity, a balance sheet is a very theoretical beast. Commercial enterprises can and do have negative equity without problem, if they can meet current liabilities, and they can and do fudge balance sheets with iffy asset valuations. There certainly are important differences between government and commercial entities, but drawing a bright line at balance sheets doesn’t seem workable to me. Even the problems with measuring government “assets and liabilities” are not so different in theory from the judgement calls required to value the illiquid and sometimes intangible assets, and the contingent or implicit liabilities, of a firm. Most analysis, public or private, is based on hypthetical adjusted balance sheets, not directly on unadjusted published numbers.

    Brad, I like your (et al)’s balance sheet approach, and agree 100% that balance sheets “exist” regardless of whether they are published, and that government and sectoral balance sheets analyisis can provide a lot of insight into the strength and risks faced by a country. (It’s worth noting that the Fed’s Z.1 Flow of Funds includes lots of hypthetical balance sheets relevant to the US.) With HZ, there seem to be some ironies, in that from the point of view of the concerns addressed in the 2002 paper, China has addressed the risks you focus on quite well. You are now very critical of new, too-much-of-a-good-thing balance sheet risks. It’s also ironic that the context of the 2002 paper was using the balance sheet approach to guide IMF lending. Usually, taking on more debt increases rather than diminishes balance sheet risk, though it diminish solvency risk, at least temporarily. As you acknowlegde, a loan from a certain perspective can’t strengthen balance sheet equity, since any assets acquired are matched by an equal and offsetting liability. The strength of your paper, though, is its focus on using balance sheet analysis to identify risks, and on letting risk management be the sine qua non of lending decisions. Although taking on leverage usually implies increased risk (what you refer to as “capital structure risk”), you are right to point out (esp. with respect to “maturity risk”) that if the value or liquidity demands of an assumed liability vary with an those of an entity’s asset portfolio, taking on a new liability may cut risk, and that would be a sweet spot for new lending.

    HZ, thanks for the link. (Replacing “2004” in the URL above might have been a good thing to try!) The same calculation I made above, which basically excludes central-bank-to-government claims (both ways) and currecy-outstanding and capital or equity accounts on the liability side, implies a “pro-forma equity” for PBoC of 3.7T CNY, or 464B USD as of July, 2006, less than I’d guestimated above, primarily because bond issues have grown faster than reserves. I guess PBoC really is steriziling.

  • Posted by bsetser

    Tis very true that the 02 paper was light on “too much of a good thing” risks — my CESifo paper has used the 02 analytical framework to look more closely at those risks.

    While domestic assets and liabilities net out on a country’s external balance sheet, the balance sheet approach pays a lot of attention to the transmission of shocks across internal balance sheets — Thailand’s firms couldn’t pay their fx debt to the banks, the banks couldn’t pay their depositors, the government had to bailout the banks … and so on.

    HK — consolidated or unconsolidated, the PBoC is in a position to provide very large amounts of fx liquidity if there is a need. there isn’t a “too little of a good thing” risks here — only the set of risks that come with too much.

    I rather suspect the $60b is quite liquid — maybe it is in agencies. and if needed those could be sold — sure, it damages the banks’ equity capital, but the external gov. bonds could be replaced by domestic government bonds if the gov. really needed the cash. the point is — it doesn’t need the cash. which lets it do somewhat unorthodox things like develop structures that count some of the PBoC’s fx reserves as part of the banking system’s capital.

  • Posted by FR

    Joseph wrote

    It’s impoosible to run any sort of commercial enterprise for any extended period of time in a market economy without a balance sheet, and that provides a marker between commercial and non-commercial enterprises.

    I think some of us are confusing balance sheets and income statements. I ran a small business for 30 years and never drew up an accurate balance sheet. (of course, I never got rich either). But I had to draw up an income statement every year, if only for the IRS.

    I also worked in a bank, and I can assure you that any bank – even a central bank – needs a balance sheet – preferably daily.

  • Posted by FR

    It just occurred to me that the US IRS may ask for balance sheets, but the French “fisc” doesn’t for certain small businesses

  • Posted by DOR


    SAFE – the State Administration for Foreign Exchange – does in fact guarantee convertibility.

    If you recall the ITICs blow-up, the foreign lenders were guaranteed being able to convert to hard currency any Renminbi they succeeded in prying out of the cold, dead hands of those pesudo-banks. Those foreign lenders with SAFE guarantees were protected, at least on the convertibility issue.

    There are similar guarantees for repatriating dividends, royalties, capital and other money.


  • Posted by Guest

    The reserves transferred to the state banks were in the form of US Treasuries, which were loaned, not given, to them. The banks can use their value to beef up their capital ratios but do not have the power to sell them.
    So yes, I agree that they should still be counted as reserves.

  • Posted by HZ

    Thanks for the insights. FX denominated debts — how significant they are in China nowadays? A few sovereign issuings and World Bank/ADB/Japanese development loans?

  • Posted by HZ

    But you never know. There could be quite a lot of players borrowing in FX and play the RMB appreciation game. Doubt SAFE would condone these, not even to mention guarantee these though.