Chinese Sterilization Watch

Three things to watch closely in 2006:

• The size of China's reserve increase.   Some say it is falling back from its recent $20 b or so monthly pace.  I am not convinced.   
• The amount of sterilization bills the PBoC issues
• And the interest rate the PBoC has to pay on its sterilization bills.  The stock of bills outstanding will grow fast as long as China keeps adding $250 b a year to its reserves.

In 2005, the pace of China's reserve growth picked up (once necessary adjustments are made for valuation), the pace of sterilization picked up commensurately and the interest rates the PBoC paid on its bills stayed quite low.  That combination put less pressure than I expected on China to adjust its currency regime.  The core question is whether 2005 demonstrates that it is possible, in fact, for China's central bank to sterilize (offset the impact of rising reserves on domestic money growth) $250 billion or so in reserve growth on a sustained basis. 

China seems to have pulled back a bit on sterilization during the middle of the year, allowing a pick-up in broad money growth and building up bank liquidity to buffer the Chinese economy against the impact of a 2.1% revaluation.   Or so the story goes.  Maybe China's leadership really was that worried by the tiny change in the RMB/$.  I would think they should have been far more worried by the changes in the RMB/ euro. 

Maybe they worried about deflation, and wanted money growth to pick up.  It certainly did.  Maybe the central bank had trouble controlling the pace of money growth.   Stephen Roach isn't sure.   I tend to think it was a conscious decision. 

No matter, UBS reports that the PBoC intends to tighten a bit and expects the pace of sterilization to pick up again.   But that the amount of sterilization needed will be manageable, because reserve growth is slowing. We will see.

I don't buy Keith Bradsher's argument that hot money flows into China have fallen off a cliff.   Are they slower than they were in the fourth quarter of 2004 and the second quarter of 2005?  Sure.  But too much is being made of the $9 billion November reserve increase.  That number needs to be revised up to reflect the $6 billion currency swap.  And then it needs to be revised up again to  reflects - I suspect - significant valuation losses on China's euro and yen reserve portfolio.   Add in the $6 billion currency swap and $8 billion or so in esimated valuation losses during October and November, and China's (adjusted) reserve accumulation over the past two months is around $39 billion, not $25 billion.   Or about $20 billion a month.

Valuation changes added to China's reported reserve growth in q4 of 2004, and cut into it both in q2 of 2005 and q4 of 2005.  And even if hot money flows tapered off a bit in the fourth quarter, the recent flurry of speculation that further changes in China's currency regime are in the cards may well be prompting renewed flows.

One indicator to watch, apart from the (valuation adjusted) change in China's reserves: the interest rate the PBoC had to pay to convince banks (and others) to hold the bank's sterilization bonds.  It is not exactly a market rate.  The central bank's restrictions on lending growth have created built in demand for cheap PBoC bills.  But it still tells us something.

That interest rate fell during the first part of the year.  For a while, the interest rate on one-year sterilization bills was closer to 1% than to 2%.  That is no longer the case.  But interest rates on one-year sterilization bills are still around 2% - a bit lower than where they were at the end of 2004. 

The 2% interest rate is important, because it implies that China's central bank pays less than on its bills than it gets on its euros, let alone its dollars. 

The interest differential was probably big enough to offset the capital loss associated with China's small 2005 revaluation - particularly taking into consideration all the zero-interest liabilities the central bank has issued.  

The standard name for zero interest central liabilities is cash.  Issuing cash against interest-paying assets is a decent line of business.  But most central banks prefer to follow the Fed's lead and issue cash against interest bearing assets denominated in their own currency - not someone's else currency.  Issuing RMB cash against maybe $550 billion in assets denominated in the currency of the world's biggest (net) borrowing and gross debtor (setting aside dark matter) and $250 billion in low-yielding euros and yen is not necessarily as good a business as issuing RMB cash against RMB government bonds that pay 3% (now).   

But issuing RMB against dollars and euros worked out for the PBoC in 2005.  China's central bank will probably end the year with around $800 billion in reserves.  And it will probably have issued about $250 billion in sterilization bills (I don't have the exact end of the year number handy, as I rely on the I-banks to provide the exact numbers.  Stephen Green, where are you?).  China has maybe sterilized another $100 billion in other ways.   That leaves about $450 billion in cash (20% of China's revised GDP) that has been issued against the central bank's foreign currency assets.    

This is all very rough ball park math.  The central bank has also issued cash for various domestic assets - including, according to Jonathan Anderson of UBS, for the bank's bad debt.   So overall base money is higher.  I suspect the China watchers will come out in force in the comments section, helping me fill in the gaps.  In other words, this is very much a back of the blog kind of calculation, one intended to lay out some basic parameters.

If the central bank earned an average of 4% on its reserves (more than possible, if most are in dollars since China holds a more diverse and higher-yielding portfolio than Japan), it earned $32 billion while paying no more than say $7 billion on its $250 billion in sterilization bills and $100 b in other types of sterilization bills.  That also is rough - the stock of both foreign assets and sterilization bills rose during the course of the year, and I used the end of year estimates rather than year-averages.      

No matter: the central bank clearly made money.  $25 billion by my rough estimate.

That would offset the capital loss China incurred when the RMB was revalued against the RMB.  Think 2% on the $700 billion or so China had at the time ... or $14 billion.  Add in maybe another $4 billion for the RMB's subsequent  0.5% appreciation against the dollar.

China took bigger valuation losses on its euros and yen from movements in the dollar/ euro (and thus RMB/ euro) than it did from movements in the RMB/ dollar in 2005.  But it can charge those losses against the capital gains on its euros and yens in 2003 and 2004.    Movements in the dollar/ euro won't leave the PBoC technically bankrupt.  Big moves in the RMB/ dollar (really, the RMB v. basket of dollars, euros and yen) certainly would. 

Of course, lots of folks think that it doesn't really matter when the central bank loses money betting on currencies.  In which case, other countries should stop supporting their exports with export credit agencies and let their central bank do the dirty work by building up more reserves than it needs.  

I, obviously, have a slightly different view.  Yu Yongding as well.  and I tend to think the rest of the PBoC does too.

UPDATE: After reading the latest commentary from Stephen Green and Jonathan Anderson, I want to add one key caveat.  China's trade surplus usually falls in the first quarter for seasonal reasons.  So if the monthly trade surplus falls from $10b a month to $5b, reserve accumulation of $15b a month is consistent with $10 b in net capital inflows ($5 b in FDI, $5 b in hot money).  The trade surplus then tends to expand for season reasons.  So $15 b a month in reserve accumulation in q1 works out to a higher number over the course of the year.

Last year, a big fall in imports in q1 kept China's trade surplus quite high in q1.  But there is no reason to expect something similar this year.