With the visit of President Hu to the United States, Chinese currency misalignment is at the top of the agenda. What is “misalignment”?
This question is not rhetorical. In a paper written with Yin-Wong Cheung and Eiji Fujii, presented at a Sept. 2005 SF Fed conference on Pacific Basin imbalances, we exploited the well known cross country relationship between the real exchange rate and relative per capita income (expressed in PPP terms).
Figure 1: Scatterplot of Relative Price Level against Relative Per Capita Income (in PPP terms), Conditional Mean, and Serial Correlation Adjusted Prediction Intervals. Source: Cheung, Chinn and Fujii (2005).
As can be seen,the Renminbi (Rmb) appears to be substantially undervalued, even after taking into account the fact that absolute PPP doesn’t hold. The regression line is for a pooled sample encompassing over a 174 countries over the 1975-2003 period, with plus/minus 1, and plus/minus 2 standard error bands (adjusted for serial correlation). The point estimate indicates that each one percent increase in relative (to US) per capita income induces a 0.25 percent real appreciation (these real exchange rates are the Penn World Tables “price level”).
One key result is the confirmation that according to this criterion, the Rmb is indeed undervalued, and undervalued by a very large amount (in log terms, over 70%; in level terms about 50%). A second, equally important result, is that the 2003 estimate is within one standard error of the conditional mean. Hence, the corresponding p-value for the hypothesis that the Rmb is undervalued is larger than conventional significance levels. In other words, by conventional levels of statistical significance, the Rmb is not undervalued.
Speaking for myself, I do not take this as proof positive that the Rmb is not undervalued. After all, failing to reject the null hypothesis is not the same as accepting the null hypothesis. Rather, I would say it tells us something about the difficulties of defining exchange rate misalignment. One notices that the degree of measured undervaluation was not that much smaller in 1997-98, when the Rmb was widely acknowledged as overvalued, or not misaligned.
Brad Setser would likely argue that the Chinese trade surplus (which has ballooned since 2003, and is now running around $100 billion a year) is clear indication that the Rmb is undervalued. I would tend to agree, although I (like Brad) would take to heart Jeff Frankel’s key observation that it is not really interesting to discuss misalignment without discussing both internal and external equilibrium. That nuance is often missed in the discussion of Rmb misalignment.
Now, most of us now agree that, under conditions of a capital control-ridden economy laden with many distortions (a legacy of central planning), the Rmb is undervalued according to both the price criterion (discussed above) and the internal-external equilibrium criterion. The question I have is what is the equilibrium value if the capital controls were not in place. This might seem to a hypothetical question, but I suspect that the initiative to acclimate the markets to a more flexible Rmb might presage greater capital account openness. Then, we may discover the equilibrium Rmb rate is much weaker than what we currently see.
The second part of the title asks whether Rmb undervaluation matters. Of course, yes. But partly no. Even if the Chinese Rmb is allowed to appreciate substantially, and the other East Asian currencies follow, the new set of imbalances is with the oil exporting countries (see the IMF WEO Chapter on oil prices and global imbalances.)
Figure 2: Current Account Balances. Source: UN ESCAP, Key Economic Developments and Prospects in the Asia-Pacific Region 2006 (Dec. 2005)
Chinese revaluation will do little to affect these balances — much of which is run with the U.S. For real change, I reiterate, the central responsibility lies with American policymakers: cut the budget deficit and decrease oil imports. (For everything you wanted to know about evaluating the Rmb, see this website).
Technorati Tags: misalignment,
undervaluation,
China,
Renminbi
Isn’t the real problem that the U.S dollar is overvalued, rather than the RNB being undervalued. I realize on a bilateral basis it amounts to the same thing, but we are running deficits with just about everybody, not just China. If the Fed really is just about done raising short term rates, and the JCB and ECB are going to continue/start raising, it seems to me that the natural path should be for the dollar to fall, perhaps significantly.
Tremendous discussion, and I agree with your policy points. Thanks Menzie.
– Tim
I would suggest that the question should not be so much about the Rmb versus the dollar but rather the Rmb against the currencies of other less developed producers of consumer goods.
the US current account deficit is driven by the us savings shortage and over the last quarter century it has been against many countries and has moved from country to country over time with the chinese just being the last in a long line of suppliers.
Maybe the real question is not how much the US imports from China, but how much China has displaced other countries exports to the US —
Mexico comes immediately to mind as it is easy to see the damage Chinese competition has done to the Mexican border industries.
nice discussion. I am inclined to think that even if China eliminated its capital controls, the RMB would rise not fall. Why? Right now the current account surplus is 7% of GDP and the FDI surplus is between 2-3%. So you need net outflows of 9% of GDP (roughly $200b) to offset current account/ fdi inflows. That is a huge outflow. And remember, right now China restricts both inflows and outflows, and is loosening the controls on the outflows while tightening the controls on the inflows. So tis true that Chinese private savers have too much concentrated exposure to Chinese banks. But it also true that most of the rest of the world is massively underweight Chinese financial assets. I don’t see a net hot money outflow of 10% of GDP.
Menzie — I would be interested in your take on whether China’s trade surplus is “real” or it represents disguised inflows. If it is disguised inflows, i would be a bit less confidence that the no intervention/ no controls (and higher deposit rates) equilibrium would be an RMB appreciation.
In my purely personal capacity, my opinion is that everything points to the CNY being substantially undervalued and to this being a real problem for China and for its trading partners.
China’s current account surplus has risen from close to balance to a surplus of 7% of GDP despite three years of boom in domestic demand that has its government worried about overinvestment. Consumption growth has been less spectacular than investment growth, but retail sales growth of 12%+ annually is hard to call weak. Adding to the evidence that the relative price signal from the currency is that China’s manufacturing base is cheap, China’s export share in GDP has risen every year for the last five years and is now at a historic high. It should be about 45% of GDP this year, spectacular for such a large domestic economy. China’s global export market share has soared over the last five years. Other than Korea, whose exchange rate has been undervalued until recently, China has been the only country in Asia to see its global export market share rise at all, much less significantly over the last five years. Clearly, China is taking market share from otherwise very competitive exporters whose currencies most models show to also be undervalued against the dollar.
And as Brad points out, China’s basic balance surplus (C/A surplus + FDI) is so large it is hard to imagine Chinese capital outflows that would neutralize pressure for the yuan to appreciate. Anecdotal evidence supports this claim. First, much of the capital inflow has been by overseas Chinese, i.e. by quasi-insiders. Second, some of the capital inflow has been by Chinese themselves who have previously transfer priced profits out of China in order to round-trip them back in under the guise of being foreign to gain incentives for FDI. These guys are definitely insiders. Third, the recent Unicol debacle shows that even if the Chinese had no Chinese government constraints on investment abroad, they would face significant foreign constraints. These geopolitical constrains on Chinese outward FDI mean much of the burden of recycling China’s basic balance surplus falls on pure financial investment. The clear evidence from the rest of the world is that a mix of home bias and simply the time it would take to mobilize Chinese savings in unit trusts and other such vehicles for investment abroad would greatly limit pure financial investment outflows from China for some time.
Finally, I think this undervaluation does matter. China’s physical size means that it is a price setter of sorts in terms of the global production possibilities frontier. What I mean is that it matters much, much more that the CNY is undervalued than say the MYR, the PLN, the MXN, or the HUF, or even all of these put together because the size of their land area and populations means it is physically impossible to fully replicate the investment that has taken place in China over the past 5 years in these alternatives. A cheap China can replicate all of the manufacturing capacity of Malaysia. A cheap Malaysia cannot replicate the manufacturing capacity of China.
I think this is causing distortions in the structure and distribution of global manufacturing and creating serious hysterisis risks for countries competing with China. I don’t believe the studies that say that China has not diverted FDI are accurate.
From discussions with Asia’s central banks, I am certain that these competitive effects, combined with China’s reluctance to allow the yuan to appreciate, has led other Asian countries to be resist unwinding of their currencies’ post-Asian crisis undervaluation. So yuan undervaluation is producing a more general Asian undervaluation.
To be sure, this generalized Asian undervaluation is a key point that the US Treasury makes. The question of adjusting relative prices as a means to adjust the US savings-investment imbalance is not just a question of yuan levels, but of pan-Asian FX levels.
Following from this, I struggle to believe a relative price shock of the magnitude of something like a 20 – 30% appreciation of the pan-Asian FX block, including the yuan, would have no effect on the US savings-investment balance. US import prices would almost certainly rise, reducing US real incomes, probably increasing pressure to slow consumption and increase savings. Asian real income would rise and Asian imports would rise, including from the US. Certainly this was the experience as Asian real exchange rates appreciated afer the 1984/85 Asian FX crisis (Malaysia and Thailand). This real apprceciation in the late 80s and 90s contributed to a steady increase in Asian domestic dmeand and deteriorationg in Asian current account deficits. I see no reason to believe similar dynamics would not be the future response to Asian FX appreciation.
Yuan undervaluation matters to China because it is distorting the structure of the economy away from domestic demand creating a growing dependance on the external sector. Ironically, this makes China more reluctant to appreciate. And to the extent that yuan undervaluation has led the other Asian countries to keep their currencies undervalued it has produced the same distortion in their economies. Note export shares in GDP have risen sharply in ALL Asian countries over the past decade. Asia is today hugely more dependant on US growth than it has ever been. Some may retort that China is now Asia’s largest trading partner. Yes, correct on paper, but this is large due to supply chain dynamics, not Chinese domestic demand. Asia exports to China to export to the OECD. Its like saying that GM parts supplier Delphi is not dependant on the US consumer because it sells to GM, not the US consumer.
I was not surprised to receive many insightful comments on this post. Let me try to do justice to them.
Brad Setser and rfarris1 on capital flight. According calculation by my former colleague Mike Dooley indicate that 1994 capital flight was about 6% of GDP. If 6% capital flight could occur at a time when much more restrictive controls were in place, why not 10% in a period of looser controls. (By the way, I count the 2006q1 figures as 4.2% of GDP, rather than 7%, using Chinese trade figures). Regarding the question of whether there are large amounts of disguised capital flight in under/over-invoicing, I don’t have any particular opinions (i.e., I don’t have any information).
rfarris1 on China’s impact on global prices. I once again refer you to the only exhaustive study I know on this subject: Kamin, et al. “Is China ‘exporting deflation’,” International Finance and Discussion Paper No. 791 (Jan 2004). Recent work by Deutschebank and in the IMF WEO argues that whatever downward pressure arising from globalization will likely disappear as output gaps around the world go positive.
Spencer and Dirk van Dijk on dollar overvaluation. I agree — and in addition the profit repatriation act probably temporarily appreciated the dollar above equilibrium. That being said, the adjustment process would be facilitated if the dollar depreciation could take place in a broad-based manner (i.e., not just against the euro). Hence, Rmb flexibiity would be a useful adjunct in the adjustment process. (Although I continue to believe that Chinese reserve accumulation of U.S. Treasurys is more important than misalignment per se).
I wonder what the bar graph would look like if we eliminated oil imports. As there isn’t much that can be done to change supply patterns of energy in the short run, the revised graph might cause us to think again about what policy to follow.
Menzie — I don’t think q1 data on the current account is out, but certainly the trade data would put the current account surplus for q1 at under 7% of GDP (If you have a q1 estimate for the current account, let me know!). However, I don’t think China’s numbers are seasonally adjusted, and q1 is traditionally the weak quarter for China’s current account surplus. So for the year, I strongly suspect China is on course for another 6-7% surplus. Certainly that is what you get if you extrapolate q1 export and import data through the entire year.
RE: 6% of GDP capital flight in 1994 … wasn’t that either just before or just after China devalued the exchange rate and during the period when the investment boom of the early 90s was turning into a bust? It certainly strikes me that conditions are rather different now. And even so, with a 6% of GDP capital outflow and a basic balance of 9% — the rMB appreciates. And if China had a fully open capital account, folks like me would put money where our mouths are, and would bet on appreciation … bringing some funds in …
The only wildcard is that China’s trade data may be distorted, with a significant amount of unrecorded inflows coming in through the trade account. that story has some problems (i.e. flows coming in via the trade account go up in 05, even as other measures of hot money suggest a fall off in hot money flows after q2). but it does give me pause, since it implies a smaller basic balance providing support for the rMB no matter what and more volatile inflows which might reverse.
I also agree with rfarris that the longer this goes on, and the more China orients its economy toward export assembly/ component production for export, the harder it becomes to change — as more domestic interest groups are adversely affected.
p.s. really like the scatter plot, even if the scale is small …
jim miller: Haven’t seen this exact same graph adjusted for oil, but Chapter 2 of the IMF World Economic Outlook has relevant data on this.
Brad Setser: These are figures reported by Dragonomics (formerly The China Economic Quarterly). I defer to you on the issue of seasonality; I’m confident you’re right that Chinese q1 figures understate the trajectory of the 2006 current account balance.
You’re right that 1994 was just before the exchange rate unification. And overall conditions are surely different. But my main point is that conditions can change, and against a backdrop of enormous contingent liabilities (pensions, NPLs) one shouldn’t forget the lessons of the East Asian crises of the 1990’s. Equilibria depend upon expectations, and those can shift quickly when the economy’s workings are not transparent.
Thanks for the compliment on the scatterplot. I’ll see if I can fix the small size.
Roach has a commentary today
http://www.morganstanley.com/GEFdata/digests/20060424-mon.html
to the effect that the Chinese are exactly right not to be pressured into re-valuation because look what happened to the Japanese.
Is there something wrong with my views here? It seems obvious to me that part of the rationale for Chinese re-valuation is that it would be good for the United States. Why is that not a consideration?
It is my understanding that there has recently been a pretty major loosening of the capital controls that China has had in place, although they are not completely gone. Unfortunately I am not sure of the extent of these.
In any case, we are in a somewhat different situation than we were as of even a year ago. It is probably the case that both the yuan/renmimbi is generally undervalued, and the US dollar is generally overvalued.
Barkley — my strong impression is that China continues to tighten controls on inflows, while loosening controls on outflows. it wants to try to reduce its pace of reserve accumulation. to me that is indicative evidence of the direction of market pressures.
On capital flight. To clarify, I’m not saying capital controls work. The evidence is clear that they do not. My point is exactly that current conditions matter and the current conditions are a euphoria about China and expectations of further yuan appreciation. Inflation in China is low and the balance of payments is strong (unlike in 85, 89, or 93/94 when it experienced captial flight). So my point is under current conditions I doubt that an easing of capital controls would lead to a flood of money out of China. Conditions would have to change to rampant inflation, powerfully negative actual and ex ante real interest rates and concern about the balance of payments to enough Chinese money to flee to weaken the yuan.
Camille Roy: Yep, the Japan experience says you don’t appreciate your currency in the wake of a massive increase in industrial capacity that realistically needs to be exported to clear. But the geopolitical history says politics may force you to do it anyway…that’s the problem with China’s massive and, most important, prolonged, undervaluation. Now that they are long capacity they need to stay undervalued to avoid the Japanese experience, but the longer they stay undervalued the more excess capacity they induce…
Menzie: I think, but don’t have the resources to test, that part of the global output gap is directly related to Chinese and Indian undervaluation. With the cost of not just production, but also investment, tremendously lower in these countries why should investment in the West or Asia ex-China rise much? And my guess, again a guess, is that on a unit-output basis, the $ investment cost of a factory in China is much less than in say Germany. If correct, if OECD industralist X needs to expand capacity by 1m widgits to meet still strong US consumer demand, he spends $0.5bn building a 1m widgit factory in China rather than $1.0bn in China. In real terms its all the same so commodities prices rise because the factroy requires the same cement and steel input wherever it is. But land and labor and supporting industry are seriously displaced in Germany, depressing German consumption. Now I accept that this type of process is always happening and happened in the 19th century. My contention is the problem is that a) there is no adjustment mechanism to the undervauation driving the process (even in the 19th century the price-specie flow mech led to adjustments in the US European balance that seem to be absent in today’s fiat currency world) and b) that China is so large it can continue to displace investment activity in the rest of the world for quite a while.
Is it possible to plot the Big Mac index (price level) on the Y-axis? (would be nice if BMI of all countries in another color, and also China’s BMI highlighted).
would also be interesting to see the dots for asian currencies from 1996-98.
I would claim the opposite. From a chinese point of view, the RMB exchange rate may be at its equilibrium given the amount of unemployed in China (100-150 millions), moreover PPP, BEER, and FEER models are not accurate for devloping countries, as it can be seen in a working paper i wrote with two colleagues
“The RMB equilibrium exchange rate: an agnostic view”
http://www.ofce.sciences-po.fr/pdf/dtravail/WP2006-13.pdf