A modest proposal

by on September 8, 2006 at 8:23 am in Economics | Permalink

The "Hayekian" argument for international currency adjustments, as made by Alex and Brad DeLong, implies that the United States needs to shift labor into its export sector, and sooner rather than later... 

Brad does offer some policy recommendations, but he leaves one out.  If one sees the need for a big sectoral shift at home, yuan revaluation is hardly the most direct policy instrument.  China is neither our leading trade partner nor the leading foreign investor in the United States.  It would have to be the case that the dollar is significantly overvalued and that market prices, not just the pegged Asian exchange rates, are all wrong.  There would be a more direct solution: boost taxes on foreign investment in the United States.  The demand for dollar-denominated assets would fall, the value of the U.S. dollar would fall, and the demand for U.S. exports would rise.  (If we are counting only American gdp, note that this tax brings revenue to the American government, unlike yuan revaluation, which raises borrowing costs and puts a burden on Wal-Mart and on the American consumer.)  Voila!

This would put both Alex and Brad in the odd position of believing that we have not enough foreign labor, but too much foreign capital.

I find it hard to accept that conclusion.  And if we had the requisite betting markets, I find it hard to believe that they would (should?) reflect U.S. economic performance as improving, contingent on such a tax hike.

1 Edgardo September 8, 2006 at 10:29 am

As I told you two or three years ago, the large accumulation of “international reserves” by the People Bank of China must be regarded as an attempt to diversify the portfolio of the state-owned banking system. Given (1) the large inflow of domestic deposits into state banks (I don’t have data at hand but likely around 25% of GDP for more than ten years!!!), (2) the huge portfolio of risky domestic loans accumulated by these banks and (3) the lack of other domestic financial assets, PBC has been wise to invest part of that inflow abroad and they should continue doing it until they are able to deal with (2) and as long as (1) holds. Of course, exchange rate policy has an impact on trade flows but we know that everywhere exchange rates are determined by capital flows and exchange rate policy is largely conditioned on capital flows. In China, capital flows are controlled by government, but because of the need to diversify the banking system’s portfolio this control allows PBC to invest abroad a much larger amount than the FDI flow. If China’s investment in US bonds were taxed, PBC would invest in EU bonds and the tax would have minor effects on exchange rates.

2 brad setser September 8, 2006 at 12:02 pm

Tyler — whoah — “China is neither our leading trade partner nor the leading foreign investor in the US” —

really? The US may trade more with Canada/ and Mexico, but that is largely because they both import a lot more goods than China does. I’ll dig up the data, but i think US imports from China are pretty close to uS imports from Canada and Mexico, tho they may be slightly lower. moreover, they are certainly growing faster.

and as for China not being the biggest foreign investor in the US, on a stock basis that is true. But what data supports the argument that they are not the biggest current lender on a flow basis. the UK may buy more US debt, but most credible folks (including Martin Feldstein) think that reflects custodial bias. From June 04 to June 05, China bought around $180b of US debt (change in the stock of Chinese holdings in the Treasury survey). The data from June 05 to June 06 isn’t out, but a 70% dollar share would imply another $150b.

The oil exporters collectively may buy more, but they aren’t a unified actor. And as for the dollar’s overvaluation, i would note that all emerging markets, not just those in Asia, are now intervening heavily to prevent their currencies from appreciating. THe gulf countries peg to the dollar. Brazil is adding $4-5B a month to its reserves, russia more like $15b …

As for the first comment, Chinese reserve growth is no way substitutes for domestic bank lending. China will end up with more domestic loans and more reserves. the mechanism works as follows — when China intervenes, it buys $ for RMB. those RMB go into the banking system and are used to finance an expansion of domestic lending unless a) the PBoC with draws the RMB by issuing sterilization bills or b) the PBoC keeps the banks from lending with administration controls. John Makin of the AEI is good on this.

Tis hard to understand, I know, but an international reserve asset is useful if there is massive domestic capital flight. But it isn’t a useful asset for backing a banking system that has mostly RMB deposits and RMB loans. China’s banks need performing RMB assets not dollars. THat is why domestic bank recapitalization has been primarily done by issuing domestic RMB bonds — typically bonds from the Asset Management Companies. Guonang Ma (not sure of the spelling) of the BIS is a great source on this.


3 dj superflat September 8, 2006 at 12:58 pm

both the foriegn account deficit and the inequality debate are amazing — first, a graduate level seminar in economics, with multiple professors; second, assuming everyone is operating in good faith (which i assume), even the best economists can’t convince one another, let alone the world, that the data should be read one way or another. this, to me, means that there’s no irrefutable basis to argue there’s an impending crisis (because no one has a definitive model of what’s happening and why) or that drastic action is warranted (which reminds me of the global warming debate, etc.). people who stridently assert that the problem is obvious — along with its solution/results? — are plainly wrong.

4 RogerM September 8, 2006 at 2:56 pm

I agree with dj superflat! I learned a lot from reading the 19th century
writer Bastiat. I think he clears the air and makes a good case for not
worrying about trade deficits or currency exchange rates.

5 Edgardo September 8, 2006 at 3:55 pm

To Brad Setser: Sorry but you’re wrong about how the accumulation of “international reserves” is financed. PBC takes funds from state-owned banks through reserve requirements on all deposits and uses these funds to purchase US bonds. It’s as simple as that. The consolidated balance sheet of state-owned banks + PBC shows “international reserves” and domestic loans in the asset side and deposits and currency in the liability side. Contrary to other experiences where the accumulation of “international reserves” is (or used to be) financed with the inflationary tax or direct domestic borrowing by the central bank, in China the financing comes directly from state banks and does not create any inflationary pressure because of the high flow of domestic deposits. This policy started around 1995 and although today I do not follow closely what is going on in China it is my understanding that it has not changed (fyi: as an economist with the World Bank Mission in Beijing I was an adviser to PBC for financial reform between 1994 and 1996 and I used to prepare flow-of-funds accounts so the Chinese could understand how they were financing the spending of government and state enterprises as well as the accumulation of reserves).

6 brad setser September 8, 2006 at 4:23 pm

Edgardo —

China’s reserve accumulation far exceeds the funds that flow to the PBoC from mandatory reserve requirements on the (growing) deposit base of the banking system. the PRC has raised reserve requirements as well to help with sterilization, but that isn’t enough — the PBoC still has to sell massive amounts of sterilization bills. and as Jon Anderson (UBS) and Stephen Green (standard Charted) note, the banks all have excess liquidity (i.e. more reserves than required) so the higher reserve requirements aren’t really biting. what is biting is administration caps on new loans.

So yes, the consolidated state bank + PBoC balance sheet will show liabilities in RMB (from the PBoC to the banks, both sterilization bills and reserves) against the the PBoC’s fx reserves. The PBoC has an intrinsic currency mismatch right now.

But I don’t think that implies that the banks are holding fx assets as a hedge against their domestic bad loan portfolio. Nor should they — i can think of a set of correlated shocks that would hurt both their RMB portfolio and any unhedged fx position (a large revaluation for one). They are actually holding RMB assets, not fx assets, with the PBoC issuing the RMB assets and taking on the currency risk with its fx portfolio.

and given the scale of inflows into the PBoC, i think it is relatively clear that the causality works more the other way — the surge in fx at the PBoC leads to a surge in RMB issuance, and that RMB issuance has been a key reason for the overall strong growth in domestic liquidity — i.e. it is one reason why m2 growth has been quite high, now close to 20%. The PBoC constantly complains about the difficulties sterilizing inflows on this scale — see Yu Yongding’s various speeches.

Tyler — I responded to some of your points — perhaps not the actual points themselves so much as the the thrust – on my blog.

I also agree with Andrew S’s point in the inflow tax … it would force adjustment, but most of the adjustment would come via a slowdown in domestic activity. true, a slowdown in domestic demand growth/ import growth has to be part of the adjustment picture. but i would like to see the US sustain its current pace of export growth as well … and that likely implies further $ depreciation, particularly if global growth slows — as is likely — with less impetus from US demand growth.

7 Edgardo September 8, 2006 at 5:13 pm

To Brad Setser: You apply the standard approach to Western banking systems to a system that is totally controlled by government (through PBC). In China, effective reserve requirements are determined ex-post by whatever amount of funds PBC is able to mobilize from state banks (the reforms of 1993-94 succeeded in improving substantially PBC’s ability to mobilize these funds); it doesn’t matter how you call the specific regulation or mechanism used by PBC to do it. You should look at state banks + PBC as just one bank to realize that if you are forced both to lend huge amounts to state enterprises and to pay back depositors 100% of their claims in case of distress (this policy has been applied many times when bank branches run into liquidity and solvency problems), you should invest part of your funds into low risk assets and US bonds are such assets (exchange rate risk is a minor problem compared with the high risks of all other domestic assets).
Unfortunately the rethoric of Western monetary policy has been learnt by some Chinese officials and they repeat arguments about sterilization that make no sense when you have such a huge inflow of savings into state banks and you control these banks. From my early experience in Argentina and Chile (between 1965 and 1984) I learnt that what matters is how government spending is financed and how this spending is allocated to the accumulation of international reserves and all other uses. In my approach the relevant issue is how large a reserve of foreign assets the government wants to accumulate (indeed my definition of “international reserves” is often broader than the standard IMF definition) and how is going to finance the accumulation. You may remember the many studies on the demand for international reserves based on trade flows that turned out to be useless; they ignore that governments demand reserves to deal with a lot of risks many of which have nothing to do with trade flows. I hope this explanation may help you to understand better what is going on in China.

8 brad setser September 8, 2006 at 6:31 pm

Edgardo — The state banks still finance SOEs, but I don’t think that is all they do. They also have a fairly active business financing property developers (often with ties to the party, bien sur) and make consumer loans (auto loans, mortgages) and the like. to me, China’s financial system increasingly looks like that of many of the Asian tigers before their crises. Directed lending to the SOEs is still a problem, but I increasingly worry about connected lending …

Two key difference tho. Chinese banks don’t have lots of exchange rate risk. They take in RMB deposits and make RMB loans. And China has growing reserves and a current account surplus, so there isn’t much risk of a devaluation. I agree that demand for international reserves has little to do with trade flows, in the classic sense. months of imports is worthless. But China doesn’t build up its reserves now because it needs protection v. international capital movements either. It has far more than it needs for that purpose. Rather it is building up reserves to support its export sector …

Call it directed lending — but of a different kind.

I disagree strongly though with the argument that

“you should invest part of your funds into low risk assets and US bonds are such assets”

US bonds are not such an asset for banks in emerging economies that have RMB (local currency generally) deposits. $ bonds and RMB liabilities = an intrinsic mismatch, and a big one.

A 40% reval would wipe out any bank with that kind of mismatch. I suspect the PBOC’s capital losses on China’s dollar reserves will likely be quite significant — maybe not quite as large as the losses the state banks incurredlending to SOEs in the 90s, but large. and I wouldn’t want that risk in the banking system.

Fortunately, the banks don’t have that kind of mismatch — their assets constitute a mix of RMB sterilization bills, RMB loans to SOEs (some performing/ some not), AMC bonds (which largely have replaced dud SOE loans from the 90s) and a mix of new kinds of lending — to property firms (often with local gov connections), to consumers and the like.

In some broad sense tho it is true that the banks are vehicle mobilize Chinese savings to buy US debt. Through a combination of required reserves and sterilization bill purchases, the banks hold lots of central bank liabilities — and the growth in those liabilities (along with base money) offsets the increase in the foreign assets of the central bank (tho the fx risk lies with the government). Or, put a bit differently, I do think it is fair to argue that the part of China’s reserve growth that corresponds with China’s domestic savings surplus is ulimately financed out of domestic bank deposits.

But the causality isn’t one way in my view — the PBoC’s reserves are growing by more than China’s current account surplus. And the pace of deposit growth is in my view a function in part of base money growth from reserve growth …

9 brad setser September 8, 2006 at 6:39 pm

oops — meant to hit preview not post — drop the obviously at the end and “if I am write” is “if I am right”

10 Barkley Rosser September 8, 2006 at 8:50 pm


They have not been sudden, but the rise of the dollar in the early
80s exceeded 40% I believe, and we have seen similarly large swings
among the major currencies over the last couple of decades. However,
I would agree that we are not likely to see a huge and sudden
appreciation of the RMB/yuan. The Chinese will not let that happen,
even if there are some in the US who would like to see it happen.

11 monkyboy September 9, 2006 at 5:15 am

“The demand for dollar-denominated assets would fall, the value of the U.S. dollar would fall, and the demand for U.S. exports would rise.”

If the Chinese have pegged their currency to the dollar…how exactly would the value of the dollar fall?

No matter what we do…the value of the dollar is exactly what China decides it is.

12 STS September 9, 2006 at 2:24 pm


No one pretends there is a consensus about the likely consequences of present trade/investment imbalances or the best policy course. Climate change is much less controversial. In that debate essentially all of the opposition is paid advocacy.

Is the planet warming? Consensus: yes. Is human activity a major driver? Consensus: yes. Should we act to reduce C02 emissions? Consensus: yes. How do we accomplish that? All hell breaks loose — it’s a political and economic question then.

13 monkyboy September 9, 2006 at 3:42 pm


There is more to China than its central bank.

The high valuation of the dollar helps Chinese businesses compete against American companies for customers and investors.

Let’s be honest, America and China are in direct economic competition…making America carry the “weight” of a high valuation of the dollar is an advantage China isn’t going to give up without a reason.

14 Kris September 10, 2006 at 3:42 am

The US has been wasting money on the high scale for lots of years on questionable issues why would we want to chang?

15 brad setser September 11, 2006 at 10:03 am

While I am all in favor of gradual adjustment (especitally since keeping the adjustment gradual requires china to keep financing the US, even as it is losing money on its reserves — that is the choice China has made), i don’t see evidence it is happening.

Aug Chinese trade surplus: over $18b
bilateral RMB/ $ is more or less unchanged in real terms, with the appreciation (small) to date just offseting lower chinese inflation
And on a broad basis, the real value of the RMB has fallen this year .. b/c the $ has fallen v. the euro.

Gradualy is adjustment is needed alright. Let it start soon!

16 Anom September 11, 2006 at 3:43 pm

Aug Chinese trade surplus: over $18b * 12 mos
= 216 b / 150k jobs
= $144,000,000 * .2 reevaluation of yuan
= $28,800,000 / job


17 Barkley Rosser September 12, 2006 at 1:00 pm

In case it was not clear, I agree with brad s.
that the RMB/yuan is not being adjusted much.
Over a year ago the Chinese announced they would
engage in gradual adjustments. There were some
slight wiggles against the dollar, but very slight.
There was a lot of rumoring that they were doing
a slow crawl against a basket, like Singapore. So,
the discussion then became, what’s in the basket
of currencies and with what weights?

Now, it is clear they are really keeping it very
tight to the dollar with only the most imperceptible
change, so much so that indeed the RMB/yuan has gone
down against some important currencies. More change
than we have been seeing is definitely called for.

Comments on this entry are closed.

Previous post:

Next post: