16 September 2007
In spotless Dalian 10 days ago, where all signs of industrial activity and construction had ceased for the "Summer Davos" session of the World Economic Forum, nobody wanted to mention the "B" word. Hear no evil, speak no evil and see no evil in the world's largest country without free media. Even western bankers and economists, who should know better, bit their tongues. Take, for example, the session on Outlook for China's Capital Markets.
In China, it's become the great urban myth of 2007: "the government won't let the market fall before the Beijing Olympics". But the Government wasn't able to control the market on the way up, so why should they be able to control it on the way down? All that they can do is manipulate the state-controlled media in increasingly futile efforts to jawbone the market in a sideways direction, accompanied by tweaking of bank reserve ratios and interest rates. They must be absolutely terrified.
Back on 23-Mar-07, the People's Daily opined China has no serious stock market bubble. Rest assured, they said, the State Council had made "important strategic plans and brilliant policy decisions" and quoted former CSRC Chairman Zhou Zhengqing as saying "there is no serious bubble in China's stock market." and "generally speaking, the development of the stock market is healthy". That has become the line-to-take among officialdom.
Almost 6 months later, and with the Shanghai Composite Index (SCI) up another 72%, on 8-Sep-07, the official Chinanews tells us "Generally speaking, Chinese stock market is still developing soundly at present and there are no major bubbles in the stock market...". The latest article proudly tells us that the mainland market capitalization has hit CNY24trn (USD3.2tn) and the average (historic) price/earnings (P/E) ratio is now 61.2.
Of course, the "E" in that P/E includes a substantial component of stock market gains as the dragon chases its own tail. Since the end of 2005, the SCI has increased from 1,161 to 5,312, a gain of 358% (or 4.6-fold) in just 20 months, allowing many companies to report bumper earnings for the 2006 calendar year. Accompanying the market bubble has been a property bubble which again has been included in the "E" through increases in the valuations of property assets. But these are capital items, reflecting the net present value of future net income from property - so including them in a P/E calculation is wrongly squaring the multiple. Excluding stock market and property gains from earnings, a fairer estimate of the market P/E would be 80-100x core earnings.
The vast majority of individual investors in China today have never experienced a serious market correction, and have only the simplest understanding of the market. All that they know is that almost everyone they know has made gains on their stocks, so they jump on the bandwagon. What they don't know is that the bandwagon is heading for the edge of a cliff. Like a giant Ponzi scheme, when everyone has bought in, there will be nobody left to pay them a profit.
By definition, bubbles are markets whose valuations are unsupported by fundamentals. There is nothing to fall back on. Bubbles never just plateau and go sideways - because that would not satisfy investors who bought in expectation of continued rapid gains. As soon as the momentum runs out, those investors head for the exit, and with nobody willing to take their place, the market crashes, usually overshooting fundamental value on the downside. Just a return to the index level of 20 months ago, when valuations of some stocks were beginning to look reasonable, would be a drop of 78%.
Like avalanches and other non-linear phenomena, nobody can exactly predict when a bubble will burst. All they can do is look at the accumulating snow on the mountain, and decide that it is not a good time to go skiing. By staying indoors, they might miss out on some great skiing, but they can be certain of avoiding burial in an avalanche.
The A-H Gap
As some of China's more sophisticated investors look out of the chalet window at the rest of the world, one thing that must have become increasingly obvious to them is the discount on which shares of some companies trade in Hong Kong as "H" shares, relative to their A-share counterparts. "One company, two share prices" mirrors "One country, two systems". Since the end of 2005 and up to 17-Aug-07, the H-share index had gained "only" 106% from 5,330 to 11,003, a pale imitation of the bubble over the border, where the SCI had gained 301% to 4,657 in the same period, outperforming the H-share index by about 95%.
Along the way, there has been increasing hand-wringing by bureaucrats in Hong Kong about the city's market being sidelined as companies naturally prefer to list where they can get the best price for their shares. For the same reason, mainland government-controlled companies listed in HK during the earlier "red chip" phase, when the companies were incorporated in HK, are now seeking to cash in on the mainland bubble by issuing shares there, although none has yet been approved by the CSRC.
HK Government fails at home
In our view, the concerns of HK bureaucrats are misplaced. Rather than looking for ways to spread the bubble from the mainland to HK, and seeking to prevent red-chips from listing in the mainland, they should allow the bubble to run its natural course and burst. Instead, they should be focussing on a backlog of legislative reform, such as statutory backing for our Listing Rules (under discussion since the Penny Stocks Incident of 2002), scripless registration for our stocks (on the drawing board since the SCEFI report in 1999), and class action rights for investors, to privatise and enhance the deterrent to abuse of minority shareholders. The HK Government has never even acknowledged the need for the latter, preferring to rely on a system that goes little further than name-and-shame, and leaving the courts inaccessible to investors individually, due to the prohibitive cost of a single action. Even if class actions were permitted, the current law leaves auditors with no duty of care to investors, and prospectus liability goes no further than those who subscribe in an IPO and still hold the stock, not secondary market purchasers.
HK authorities should, as we said a year ago, focus on building a value proposition for HK by upgrading HK's legislative and regulatory software for the future. When the bubble bursts, HK will, as part of the free world, continue to be open for business, and mainland companies will still come here to list, while the mainland will likely go into one of its paralytic phases of banning new IPOs and will likely launch a "stabilisation fund" to cushion the fall. So the bust will buy HK more time to get its act together. In the 4 years of mainland market disruption from its 2001 bubble down to the trough of 2005, HK did very little to build its competitive advantage.
The yawning A-H price gap, as well as the increase pressure from foreign exchange inflows, may have been a catalyst for the premature announcement on 20-Aug-07 by the State Administration of Foreign Exchange, under the mainland's central bank, of pilot plans to allow mainland individuals to convert their domestic savings into HKD and buy shares in HK via the Tianjin branch of Bank of China Ltd.
The HK market reacted euphorically, with the H-share index gaining 33.7% and the broader Hang Seng Index gaining 22.1% in the last 4 weeks, while the SCI gained 13.2%. Still, A-shares trade at an average 65% premium to corresponding H-shares as of Friday 14-Sep-07. That's without a single share being bought under the so-called "through-train" scheme, which media reports say will not start until October at the earliest, more likely in 2008. For now, the authorities have succeeded in talking up the HK market, partly by being vague about the scheme's details.
In our view, the HK market's reaction was wrong, because either:
1. the scheme is going to be very limited, with operational barriers such as Tianjin Binhai residency requirements or an inability to deal in real-time, having little effect on HK-listed shares; OR
2. the scheme is for real, in which case the capital outflow will crash the mainland market and take HK down with it.
The CSRC, responsible for the stability of the securities market, knows the consequences of scenario (2) and is not keen on the scheme. In Dalian, Chairman Shang Fulin described it as "a trial, an experiment". When your own stocks are on 80-100x core earnings, and the rest of the world averaging 20x or less, the consequence of lifting outbound capital controls is obvious.
Those who say that Hong Kong prices will rise to meet China's prices have failed to appreciate that HK stocks are priced relative to the other investment opportunities for global capital, whereas China's stocks are priced relative to no other choices for mainlanders but real estate or cash. Even the recent run-up since the "thru-train" was announced is making some foreign investors nervous, as demonstrated by the block-sales last week of certain H-shares.
HSI past draw-downs
Whatever the eventual catalyst, when the mainland bubble bursts, it will certainly take the HK market down with it, since most of the market capitalisation is now either mainland stocks or stocks with a large component of mainland business. It would not be at all unreasonable to visit 15,000 or even 12,000 again on the Hang Seng Index, despite the depreciation of the dollar and the time value of money since we were last at those levels. Here is a look at previous draw-downs (bold text indicates a then-record):
Looking just at the last 6 major draw-downs since the HK$/US$ peg was introduced in 1983, the average is 47%, the smallest was 36% in 1984, and the largest was 60% from 1997 to 1998. So a mid-sized drawdown would take us back to 13,200. For sure, the market could stretch even higher from today, but it will be ugly when it corrects.
The mainland market today shares many of the characteristics of the Taiwan market in 1990, although the Taiwan market was of course much smaller. The market was young, speculative and underdeveloped, as Taiwan emerged from it's post-war period of martial law. The capital account was largely closed, with foreign investors first granted permission for fixed amounts of investment quota as Qualified Foreign Institutional Investors (QFII) in 1991. Market turnover in 1990 was 500% of market capitalisation. The Taiwan Stock-Weighted Index reached a high of 12,495 on 10-Feb-90, and then crashed 80% to a low of 2,560 that year. The two years before that also saw huge swings. 17 years later, that 1990 record high still stands, with the index now at 9,032.
Burst bubble will hit the banks
When the mainland bubble bursts, it will accelerate the onset of the first post-IPO banking crisis. Before listing 3 of the big 4 mainland banks (BOC, ICBC and CCB), the Government injected some US$20bn each of fresh capital and transferred out the bad loans to special vehicles. That cleaned out the bad loans, but it didn't remove the bad lenders. With China only part-way through the transition from planned economy to market economy, the large old banks are still really just collections of thousands of smaller banks, with inadequate credit analysis, and with a large measure of policy-driven and corruption-driven lending. In the current bubble, many of the corporate loans are being diverted into the stock and property markets, and the collapse of the bubble will create loan-losses.
In the past, with the banks wholly-owned by government, bail-outs were relatively simple affairs, but next time, they will have to choose between renationalising the banks by injecting fresh equity, or diluting their control with outside equity, something which is an anathema to a control-obsessed state.
The common market
HK Government representatives have in recent days been in full spin mode, seeking (but failing) to justify its purchase of a 5.9% stake in HKEx. They talk in vague terms about possible future co-operation with the mainland exchanges, cross-listings, share swaps, or full mergers. This is really stating the obvious. When mainland capital controls are fully abolished, and the RMB takes its place as an international, fully convertible currency, then China's exchanges will be competing on a level playing field, and it will then make sense (subject to any competition laws) for them to cross-list securities or merge at a corporate level. With full convertibility, the financial systems can be connected, and HKEx can trade and settle shares in RMB just as easily as HKD. However, we are still some years away from full abolition of capital controls, and that day may recede further when the bubble bursts.
While we wait for capital controls to be abolished, there have been calls from the Hong Kong Monetary Authority and the Government since Mar-06 for China Depository Receipts (CDRs) to be listed in the mainland representing HK-listed shares, and for Hong Kong Depository Receipts to be listed in HK representing mainland-listed shares.
This, on its own, would be largely cosmetic. Putting a wrapper around the shares would not change the pricing dynamics. The CDRs would still amount to A-shares, available only to mainlanders (and QFII investors) and traded in RMB, while the HDRs would still amount to H-shares, traded in HKD and available only to the rest of the world (and QDII investors). Unless there is freedom to move capital between the markets, the A-H price differentials that vary from time to time would be unaffected by the introduction of CDRs/HDRs, as they would trade in isolated pools.
However, the most worrying aspect of the HK Government's calls is that the HKMA has been pushing to establish an "A/H arbitrage" mechanism (apart from the thru-train) in which an entity (let's call it "ArbCo") jointly owned by the two central banks would have exclusivity to arbitrage by freely moving capital between the two markets. Joseph Yam hinted at this in his column of 23-Mar-06:
"Before the relevant controls are dismantled and the invisible hand can be relied upon, an arbitrage mechanism might be needed to permit buying and selling in the two markets to achieve price equalisation. Since foreign-exchange transactions would be involved, the People's Bank of China and State Administration of Foreign Exchange might have to play a role in this arbitrage mechanism."
and again on 1-Feb-07:
"Another example is the creation of derivative instruments in the form of, say, certificates of ownership of shares listed on the Shanghai, Shenzhen and Hong Kong stock exchanges, and have them traded in both markets with an arbitrage mechanism to equalise prices."
and this from Donald Tsang in the FT on 15-Jun-07:
"There's no reason why through arbitrage arrangements that shares, stocks listed in Shanghai cannot through some financial instruments be traded in Hong Kong... Similarly I do not see why Hong Kong stocks cannot be co-listed in the Shanghai stock market through an arbitrage arrangement...
All you need is cooperation between the [HKMA] and the [PBOC]. And we are pals. We work together very closely already so there is really not too much difficulty for doing the arbitrage arrangement but this would be a big step for the mainland. They quite rightly so have to be very cautious how it might impact on the mainland market... I do not want to characterise [China's reaction] in any way but I'm very enthusiastic..."
How could this possibly work? Whenever there was a price differential in a dual-listed stock (or DR) between the two markets, ArbCo would buy from the cheaper market and sell to the more expensive. To settle those trades, there are two possible approaches, but as we will show, only one of them is viable:
ArbCo would be allowed to transfer shares between the mainland and HK registers, converting A-shares to H-shares or vice versa, to deliver for settlement of the sale; OR
ArbCo would buy shares in the long market and borrow shares to sell short in the higher-priced market (currently, the A-share market).
However, the mainland does not allow short sales and does not have a well-developed stock-lending market, and even if it did, ArbCo would incur stock-borrowing costs, would have to post collateral, and could end up with huge short positions in one market, with correspondingly huge stakes and voting rights in the other side of the pair-trade. That's also very risky, because once they had borrowed every available share, there would be no further way to hedge the long position, and they would have to stop intervening in both markets, so the spread would widen again. If stock-lenders started to recall their loans, then ArbCo would have to buy back shares in a short-squeeze, pushing the price up, and unwind the other side of the hedge, pushing it down and incurring a loss. So option (2) is essentially unworkable, and we have to assume that option (1) is the intent, i.e. that ArbCo would have the exclusive right to move shares between the HK and mainland registers (or depositories) for settlement of offsetting trades, but would not hold any long or short positions in either market.
There are two possible approaches to the trading; we'll call the first one a "clean arbitrage" in which ArbCo seeks to eliminate price differences by bridging the two markets, simultaneously hitting equal and offsetting bids and offers in the two markets whenever the two cross over, and not seeking a profit. The second approach is a "dirty arbitrage", in which ArbCo profits from its unique privileges by allowing a spread between the two markets, and only intervening to take advantage of that spread when it is at or above a pre-determined limit.
The dirty arbitrage would be similar to the way the HKMA runs its thick peg HKD/USD currency board, purportedly only intervening at the boundaries of 7.75 and 7.85 (a 1.3% range) rather than running a clean peg at 7.80. The difference, however, between the currency board and the stock markets is that in the currency board, the HKMA can issue an infinite amount of HKD (at 7.75) to satisfy demand, because it is the issuer of the currency. In the stock market, you are dealing with a fixed amount of shares in each company, and that only changes if the company issues new shares (or buys back shares).
In either the clean or dirty arbitrage scenario (unless the target spread was very wide, say a 50% discount), mainland prices would plunge towards HK prices, which are determined in competition with investments in the rest of the world. ArbCo would sell in the mainland and buy in HK until prices were equalised (clean arbitrage) or reached the target spread (dirty arbitrage). In a clean arbitrage scenario, and with enough companies in the program, that would crash the mainland bubble. In a dirty arbitrage scenario, there would remain an imbalance between demand in the mainland and HK, reflected by the differential pricing, and this could draw shares off the HK register and onto the mainland register, gradually delisting the shares from HK. If this reached its extreme, liquidity in the HK stock would dry up, and there would be nothing left to arbitrage until mainland prices fell enough to be attractive to overseas investors again.
In either scenario, as more and more companies were dual-listed (whether in share form or depository receipt form, it makes no difference) this would amount to the gradual abolition of capital controls, as mainlanders would be able to use their RMB to buy shares in foreign companies on the mainland register, and ArbCo would funnel the foreign exchange out through the conduit as shares came in. By proposing an arbitrage system, the HKMA is, in effect, calling for a gradual abolition of capital controls.
We can see no good reason for giving the HKMA/PBOC a monopoly on the relaxation of capital controls. It is surely not in HK's interest to allow mainland companies to be delisted from HK by converting H-shares and red-chips into A-shares through a dirty arbitrage conduit, and it is not in China's interest to trigger a precipitous collapse of its bubble through a clean arbitrage conduit.
Given the precarious state of the mainland bubble, they have simply left it too late to open the capital floodgates without triggering a crash. This is something they could have more safely done in mid-2005, but it is not a realistic option today, any more than a full-blown thru-train is.
Socio-political risk and reform
The most unpredictable aspect of all this is how the Chinese public will react to the eventual bursting of the bubble. Cosseted in a state-controlled balm of soothing media, with assurances that the markets are "healthy" and that there are "no major bubbles", the revelation that stock markets are not centrally-planned and state-directed, and the loss of substantial chunks of life savings, is going to be a huge shock. At a dinner in Dalian, we were joined by postgraduate students from Tsinghua University, the cream of China. We asked one of them, pursuing a PhD in economics, whether he invested in the stock market. "Oh yes", he said. "And what about your classmates?" we asked him. "Well, last year, I was the only one, but now all of them are doing it."
The rising prosperity of China's urban masses under economic reforms, together with the State's iron grip on the media and oppression of free speech, has kept the population largely content with the Government's leadership, but it comes at a cost, because efficient and free markets depend on information flow. The near-absence of public debate on the market risks, and the complacent propaganda in the media, have allowed the bubble to grow larger than it otherwise would, sucking in more people with more of their money. The collapse will leave a lot of people questioning that system, and that could in the end be the silver lining, leading to greater freedom of speech and the media, greater governmental accountability, and ultimately political reform.
© Webb-site.com, 2007