My Blog

September 20, 2007

Dan Rosen was the first to alert me to the fact that short-term rates in China have soared in recent days.  The benchmark rate is the 7-day repo rate, which hit a pre-September high of 4.77% in April.  Today, however, it went to 6.69%.  The one-year repo rate, which had peaked at 4.45% in April was even more volatile, reaching 7.21% today.

My first thought was that this was a reaction to some recent large IPOs, notably the China Construction Band deal which priced on Monday with $300 billion of demand.  Combine this with a 50 basis point increase in the bank reserve ratio last week, and the recent and planned sales of long-term MoF bonds to finance the sovereign wealth fund, and it seemed like the run-up in rates was caused by a normal excess demand for funds in the repo market.  This is how it was described last Thursday (September 17) in the South China Morning Post:

Traders dumped mainland bills and bonds on Thursday as a money market squeeze triggered by tighter policy and China Construction Bank’s huge Shanghai initial public offering worsened. “There’s panic selling - people are scrambling to raise funds,” said a trader at a Chinese bank in Nanjing. “Liquidity is much tighter today than yesterday.”

 

The money market has suffered half a dozen other squeezes this year during big initial public offers of equity, and each time short-term interest rates have quickly pulled back near their previous levels after the IPO has passed. But the current panic appears to be more serious because it is occurring at a time of unprecedented monetary tightening, and after Tuesday’s announcement that last month’s inflation jumped to a 10-year high of 6.5 per cent, several traders said.

 

Over the past week authorities have announced a reserve ratio increase, a 151 billion yuan issue of special bills and a plan to sell 200 billion yuan of special bonds to the market, while an interest rate rise is expected as soon as this week. That has left the market unsure about how much liquidity the central bank intends to leave in the market after the IPO is completed, and how far short-term rates will eventually come back down. “There are too many uncertainties about policy now. The special bond issue really scared the market,” said a trader at a leading domestic bank in Shanghai.

 

Trading volume has shrunk dramatically because of the squeeze but some individual bonds were quoted early on Thursday at yields at least several basis points higher than Wednesday’s levels, traders said. “Nobody is buying bills or bonds,” said the trader in Nanjing. The weighted average seven-day repo rate, the key measure of short-term liquidity, jumped to a fresh two-month high of 4.0173 per cent from Wednesday’s close of 3.8437 per cent. Traders think it could hit a multi-year high of 5 per cent in the next couple of days as subscriptions are taken on Friday and Monday for CCB’s IPO, which is expected to attract a record total of more than 2 trillion yuan in subscriptions.

The central bank appeared to be trying to ease the squeeze last Thursday by canceling its regular weekly issue of three-year bills. That means it injected RMB 140 billion into the market, against a net drain of RMB 145 billion last week.  But that wasn’t the end of the story.  Today, as Dan pointed out, the deals have been put the bed and the unsuccessful bids returned, but instead of coming down, rates have actually increased.

So what is going on?  Quite a lot, it seems.  Market rumors are that this is all part of a short squeeze engineered by the PBoC because of their worries about inflation and rising stock prices, which may be linked in their minds.  Instead of acting to balance short-term changes in the supply of and demand for funds, as they normally do, they are simply walking away from the short-term markets and allowing rates to rise sharply.  By causing such volatility in market and letting short term rates rise so sharply, albeit temporarily (money is expected to seep back into the market over the next few weeks), they are hoping to undermine speculative fervor and, by slowing down the stock market, even take wealth-effect pressure from consumption (and thus reduce inflationary pressures).

This squeeze might have a dampening impact on the stock market, but since margin buying is illegal, I am not sure how the transmission will work.  At any rate the market has been slow but not panicky in the last three days, and in fact was up 1.40% today.  I suspect high repo rates may dampen demand for new issues more than they affect secondary trading levels.  

There may also be a very negative and unintended consequence.  With the Fed having lowered rates by 50 basis points, this may not be the best time to raise short term RMB rates if the PBoC wants to discourage speculative inflows.  After all, if you can get 7% in RMB, plus 5% or more in expected RMB appreciation against the dollar, the resulting 12% return in dollars is high enough to make all but the wildest hedge funds pretty happy.  Of course anything that encourages more hot money into the country will simply exacerbate the very process that is at the root of the imbalances the PBoC is so desperate to end.

Will any of my former students or other readers of this blog who are active in the local money markets weigh in?  What exactly is going on?

Comments (15) for "What is the PBoC doing to re...

At this point, I don't see how the PBC has much choice except to bite the bullet and sharply revalue. I should make it clear that my objections to revaluation in the past have been operational rather strategic. Moving from a peg to a semi-floating currency has a lot of operational issues that are sometimes underestimated. I still think a one time reval of 15% overnight would be disastrous, but a 15% reval over one to three months shouldn't have much problem, and the PBC can add random volatility into the mix to confuse/annoy/bankrupt speculators. (The reason an overnight reval of 15% would cause chaos is that banks do risk analysis and position calculations overnight. A sudden 15% shift and all of those numbers are meaningless.) The PRC has had over two years of experience in dealing with unpegged rates, and at this point my feeling is that the remaining operational risks are less severe the what happens if you don't revalue given that the US and China have wildly different monetary objectives at this point. Time to take off the training wheels..... By Twofish - 9/19/2007 9:47 PM

The problem with the faster appreciation expectation story is that the returns acheivable for China's dollar FX recylced back to the markets is much lower now, given Treasury yields, and the reluctance to further explore ABS/MBS etc in the current climate. Just beating a 5% appreciation pace will be tough. And SAFE fx managers and RMB rate setters are, of course, on the same team. Dan Rosen By DHR - 9/19/2007 10:32 PM

The short-term liquidity squeeze is a result of the sheer size of the IPO subscriptions (2.26 trillion yuan for CCB = 6% of total banking system deposits), the fact that we have had 4 of these large IPOs this month (and may have PetroChina next week), and the cumulative effects of reserve requirement hikes, which affect smaller banks to a greater extent than large banks. Custodial accounts for brokerages, however, are all held primarily at the larger banks. Therefore, these huge IPO subscription periods result in big flows of deposits from small banks to large banks, leaving small banks scrambling for funds to rebuild the levels of reserves they need for regular transactions, pushing up 7-day repo rates (one transaction yesterday was executed at 10%). You see these spikes all the time around large IPOs, the difference now is that they are happening one after the other. During these times, there's no demand for PBOC paper- smaller banks don't have the funds, and larger banks can make more lending to the small banks than buying PBOC paper in the short term. The cost of an apparent plan to expand the supply of shares to take some steam out of the equity market is the efficacy of the PBOC's regular sterilization operations. At the end of all this, however, the IPOs don't actually suck up that much in liquidity, and the PBOC can't reduce excess liquidity any other way but to hike reserve requirements again (400 billion yuan in PBOC paper matures in October), which they will have to do at the end of next week, to be effective next month. Alternatively, they could keep placing punitive bonds with the commercial banks at below-market yields, which they have done 5 times this year (followed by interest rate hikes a week later, in every instance). It's a pretty vicious cycle for the smaller banks in this environment. By Logan Wright - 9/20/2007 9:49 AM

Logan, thanks a lot. This is really interesting and useful stuff. It seems from what you are saying that you don't think there is anything new here. Do you think the only difference between the recent spikes and all the others is the size and number of fund-draining transactions? Hasn't the PBoC been more accomodative during other such periods? By Michael Pettis - 9/20/2007 1:37 PM

In my opinion, it's a combination of the massive subscription funds (which are all-time highs) and the cumulative effects of higher reserve requirements this year (350 bps in hikes so far this year). In the past, squeezes would last a few days, but there wasn't as much interest in subscribing to the IPOs and the flows didn't impact small banks' reserve positions as significantly. Absent issuing significant volumes of reverse repos, there's not much the PBOC can do here, simply because of the sheer size of the funds being locked up. They're injecting 94 billion yuan this week, 140 billion yuan last week, but you're still seeing liquidity shortages like this, primarily because of how attractive the equity market is and the recent performance of IPOs. I have also heard traders complain that there doesn't seem to be that much coordination between the PBOC, MOF, and CSRC on issuance schedules and IPOs- this just makes everyone a little more anxious. The MOF sold 35.09 billion yuan in 10-year bonds in the midst of this squeeze, and is selling 32 billion yuan more next Friday, even while the PBOC can't sell much of anything, two more IPOs will occur next week, and banks will need more cash on hand ahead of the holiday. This squeeze is likely to persist all of next week, and maybe even beyond the holiday... By Logan Wright - 9/20/2007 2:08 PM

Twofish, I agree that a maxi-reval is very risky; it's just that I think the alternatives are riskier. If China were to reval by 15% over several months (over six months, by the way, the annual return would be almost 39%), I think that it would inspire an orgy of speculative inflows even long after the PBoC had reached the target level. After all, how would we be sure that the appreciation had really stopped, and as long as Chinese interest rates are higher than US rates, investors would earn a premium making the bet even if the bet was "lost". These inflows could even become self-fulfilling because if they are too great, they could force the PBoC to revalue even beyond its target -- and since everyone knows this, there would be an incentive to join the bet no matter whether or not you believed the PBoC had reached their target. The only defense the PBoC would have is to lower interest rates substantially – not a good idea during a liquidity swell – and focus all its attention on controlling capital flows, which is difficult, slow, and very inefficient to the economy. Even introducing random volatility into the upward crawl would not discourage anyone since we will all know why it is occurring and how it will end. If the PBoC allows the appreciation to occur much more quickly, say over two or three moths, there will still be some speculative inflow (and it will almost certainly continue after the PBoC has achieved its target since no one will know the target was achieved), but the banks will have all the same risk-management issues that they would have had with a one-off maxi-reval. The accounting for the banks might be a little easier, but even after a maxi-reval they can get an extension on monthly reports, to get them to the same place. Aside from not encouraging speculative inflows during the appreciation, the one-off reval has the important advantage of making it very clear that the PBoC has achieved its target level. By revaluing and pegging, the PBoC signals to the world where their target is, and they can immediately begin to sell one year dollar puts both to absorb further speculative demand for RMB and to signal their determination to resist further appreciation (and of course they would earn the premium for free). Dan, I know perceptions at the PBoC may differ, but the loss they take on their reserve position is not as important a consideration as many believe. First, reserves can only be used to purchase foreign goods and pay external debt, and any RMB loss on the reserve position would be met exactly by a reduction in the RMB value of foreign goods and external debt. The danger, of course, is that the PBoC will have a balance sheet loss and so its creditworthiness will be reduced, but since the PBoC doesn’t borrow externally and its domestic borrowings are underwritten by its ability to create RMB, that shouldn’t matter too much. At any rate if they are worried about the losses, the last thing they should be doing is to accumulate even more overvalued dollars. Things will be tough for the CIC, who has to earn a higher return in a rapidly depreciating currency, but either way someone in China is going to have to absorb that loss. By Michael Pettis - 9/20/2007 2:19 PM

Pettis: The accounting for the banks might be a little easier, but even after a maxi-reval they can get an extension on monthly reports, to get them to the same place. What worries me is not the monthly reports, but the nightly ones. A one off 15% reval moves everything and everyone so far off the previous days positions that all of the overnight risk analysis comes completely undone. It would generate total chaos. The other problem is that I don't think that the PBC can credibly say "we've revalued once and we will revalue no further." The future of the CNY/USD rate is going to be determined by factors that the PBC has absolutely no control over, the US fiscal position being the big thing. If the PBC resets the rates, and the US dollar keeps dropping, the PBC will have to appreciate the currency even more. It's interesting to try to compare this with the situation in Latin America in the 1980's. People abandoning the dollar in another age and in another context would have been called "capital flight." The big difference is that capital that wants to run away from the United States, has nowhere to run to. I think one place to begin is to figure out how big and how fast the speculative inflows are likely to be? $100 billion? $500 billion? $1 trillion? Once we have that number, we can then look at out quickly the PRC can push capital back out of China. If we are looking at $300 billion over 3 months, then I can see the PBC being able to push that money back outside of China. On the other hand, if China were able to accelerate capital outflow, then maybe it isn't necessary to revalue the RMB quickly. One other point. I think we are seeing one limitation of the corporate balance sheet approach to national finance. The operating assumption that most corporate managers make is that "money is good" and that the goal of the corporation is to maximize share holder equity. There are cases in which a corporation could have too much cash, in which case the solution is to declare a dividend,. (And if the board of directors won't declare a dividend, then there are corporate raiders who are willing to do a leveraged buy-out to "free the cash.") The problem here is that I think we are seeing a situation is which "money is bad." So I'm wondering what the national balance sheet equivalent is of "declaring a dividend"? It occurs to me that by offering US consumers, low, low interest to buy cheap goods, that China Inc. is effectively "declaring a dividend" and that have US consumers spend and build like crazy might be economically the "best of possible worlds." The converse of that is that if having US consumers spend like crazy is a good thing, if we get into a situation where US consumers can't spend, this is going to have bad effects on the Chinese economy. By Twofish - 9/20/2007 4:09 PM

One other note. I doubt there will be any major economic moves before the party congress in mid October. In particular, the next party congress will be where we will likely see the China Investment Corporation unveiled. CIC represents a major restructuring of personnel and power relationships, and those just will not be formally finalized until the party congress convenes. At this point the details of CIC are likely to have been set up, but there is likely to be some huge back room negotiations over the details. If RMB revaluation needs to be coordinated with capital outflow policy, this means that nothing interest is also going to happen before the party congress. The upcoming party congress also probably accounts for the huge numbers of IPO's that are in the system right now. All of the major state owned enterprises have to have their boards and executive staffs approved by the Organization Department of the Communist Party, which means that they couldn't go IPO too far in ahead of the upcoming Party Congress. Stepping back a bit, and thinking about what I just wrote. I have to say that we are living in a very bizarre world. By Twofish - 9/20/2007 4:22 PM

Also about deliberately introducing volatility, I'm taking a page from derivatives pricing theory and portfolio theory. What the PBC wants to do is to keep people from doing arbitrage. From derivatives pricing theory, the only way you can do this is to introduce risk (beta) that corresponds to return (alpha). So what the PBC could do is to figure out what alpha they want, and the figure out what the corresponding volatility should be, and then introduce some randomness into the system. Ideally, they could do so via a random number generator so that even *they* don't know what the next days rates are going to be. A basket peg also does the same thing. By Twofish - 9/20/2007 4:36 PM

Michael, you say in your post ......." After all, if you can get 7% in RMB, plus 5% or more in expected RMB appreciation against the dollar, the resulting 12% return in dollars is high enough to make all but the wildest hedge funds pretty happy"........yes I would be very happy with that. But tell me how I can get it? The RMB market is tightly regulated. The conventional method of speculating in RMB is the offshore interbank NDF market. Currently, you PAY about 5% per year to own the RMB in that market. I am sure there are loopholes in the Chinese FX system somewhere but it is likely that to exploit them is diifcult and expensive and can't be done in any size that would make a difference to the PBOC. If I am worng please tell me. By Max - 9/20/2007 8:08 PM

The suspicion is that a lot of the hot flows may be overseas Chinese with family/relationship ties that can be used to informally move money in and out. You wire money to your nephew. and he uses it to buy stock. People doing this really don't care that much about currency restrictions, and there are lots of ways of moving money back and forth. This is aided by Chinese regulations which encourage financial transactions between people in the PRC and their relatives overseas. For a Western company to get money into China is easy. You send in money that goes into wholely foreign owned enterprise or a joint venture. While WFOE's and JV's aren't supposed to be used for speculative purposes, the restrictions for getting money in are so loose that you can find a way of effectively do currency/real estate speculation. Buy land meant for a factory, for example.... The hard part in this situation is getting your money back out. You can take your money into China and buy land which is meant for a factory to produce widgets, but after a while, the government is going to want to see the factory and the widgets. By Twofish - 9/20/2007 9:57 PM

Twofish, I agree. What you descirbe happens all the time...I just don't think any hedge funds are doing it. It is way too illiquid and cumbersome. So when people speak of hedge funds speculating in the RMB, it is not because we are getting 7% yield on RMB assets. It is because we expect the RMB to revalue more than the 5% we are paying to own it forward. Still a worthwhile bet in my book. By Max - 9/20/2007 10:55 PM

Max, I think this is a case of bleeding of 1000 cuts. It's not that billion dollar deals are happening but rather that hundreds of thousands of small deals are happening that add up. For example, most of my friends who live in China (and me too for that matter) have accelerated our rate of money inflows to protect ourselves from revaluation. There are in theory limits to how much you can bring in, but those limits seem imposed mostly at the bank level, and if you open two or three accounts you can double or triple the limit. Of course I a not able to earn the 7% locally (I get deposit rates) but it is still a hell of a lot better than leaving my money in Citibank New York, and I know that Chinese friends with better connections than mine are able to get higher returns from more "informal" arrangements. Anecdotal evidence suggests, as Twofish mentions, that a lot of families are bringing money into the countries, and Standard Chartered has ettimated that as much as $100 billion comes in through the trade acoount via under- or over-invoicing. When I said 17% was good enough for a hedge fund, I didn't mean that hedge funds wered doing it (although I think some may be through their QFII quotas). I meant rather that if a hedge fund is happy with that kind of return, I am sure lots of other less sophisticated investors would be delighted. In the end, the PBoC doesn't need to distinguish between a single $100 milion investment or 10,000 $10,000 dollar investments. the monetary impact is the same. By Michael Pettis - 9/21/2007 1:44 PM

Thanks Michael. I understand well the fact that there is under/over invoicing and retail flows and its effect on monetary policy. I was just responding to your post because I thought you believed that hedge funds could just borrow dollars and convert them into RMB deposits. I see now you don't. What I do wonder is why the PBOC doesn't do a more effective job of sterilization. Perhaps the recent surge in repo rates is the PBOC finally getting aggressive on sterilization? By Max - 9/21/2007 4:31 PM

Max, I don't think the PBoC can do much about sterilization. First of all, they try to limit the yield at which they sell their bills, so that often means failed auctions. Secondly, I don't think exchanging cash for central bank bills, an almost perfect substitute for cash, has much impact on underlying money. I would like to see them use longer-term bonds and less liquid paper -- like the new MoF bonds used to purchase the reserves from the PBoC -- since they are likely to have a greater impact, but when you have to sterilize $30-40 billion of reserve growth every month, there are not a lot of things you can do, especially if you don't want rates to go up too much and too quickly. I have heard that part of the reason repo rates rose so much is because, as you suggest, they are being less accomodative, perhaps in order to fight inflation and speculation. Who knows? By Michael Pettis - 9/21/2007 6:23 PM