Mystery-blogger knzn packs an avalanche of insights into a very brief discussion of why China's policy of puchasing dollars to peg its nominal exchange rate has not led to real appreciation via inflation in China. knzn is quite critical of Charles Schumer and Lindsey Graham, who've written an op-ed in the Wall Street Journal (via Mark Thoma) today in defense of a proposal that amounts to a faint and distant threat of tariffs on Chinese goods. I am broadly sympathetic to Schumer and Graham's proposal, despite considering myself a staunch free-trader, and I think their point largely stands, despite problems that knzn and Greg Mankiw highlight. I hope he will forgive me for abusing his insights in support of a position he may not much like. Here's knzn:

China today is a special case for several reasons. First, there is a rapid flow of workers into the industrial sector, and this flow is helping to prevent the inflation that might ordinarily attend an undervalued currency. Second, much of China's foreign exchange intervention is sterilized, which is to say, China is making attempts to slow down domestic sources of demand to compensate for the foreign demand occasioned by its weak currency. Third, China's government is effectively running a very large surplus, which also tends to slow down domestic demand. Fourth, arguably, China is following policies that encourage a high level of private saving, which also tends to slow domestic demand.

knzn whizzes through this like it was nothing, but there's a ton of interesting stuff here. knzn's first reason is conventional wisdom, fine as far as it goes, and as the vowelless one notes, "if China has a lot of workers available to do these jobs, whereas the US has only a few (relatively speaking), why shouldn't trade be set up to create jobs in China rather than the US?"

But knzn's second point left me standing naked on the shoreline in wonder. Countries that manage their exchange rates are supposed to sterilize their interventions. It's like Central Banking 101. If China's central bank prints Yuan to buy dollars, it's got to borrow some of those Yuan back to avoid flooding the market with newly printed Chinese bills, right? knzn is having none of it, and of course he's absolutely right to note that sterilization is just a fancy word for what central banks do when they want to restrain domestic demand to avoid inflation. What I considered central bank prudence, knzn considers cheating, "[S]terilized intervention, might reasonably be considered an unfair trade practice (and probably a foolish practice as well)." Central banks may "naturally" want to prevent their exchange rate policy from causing inflation, but doing so in effect pegs the real, rather than nominal exchange rate. While pegging nominal exchange rates need not impede trade balance adjustments (because those adjustments can occur via changes in the price level), pegging real exchange rates blocks trade adjustment entirely.

knzn's third point is also a shocker. He notes that China is "effectively" running a large surplus. But hold the phone. Isn't China running a significant fiscal deficit? Well, yes, if you consider only tax revenues. But, if you consider a consolidated income statement that includes China's central bank, China's deficit would be dwarfed by the piles of foreign currency flying into the vaults, net of sterilization costs (and even net of significant presumed valuation losses). So is knzn is right to refer to this as "a very large surplus, which also tends to slow down domestic demand"? I think he's only half right, in a very subtle way.

Imagine that China did no sterilization, and purchased dollars with freshly printed Yuan to maintain its peg. This would maximize the surplus achieved by China's government (including central bank). But it would also be a very loose, a stimulative monetary policy that might provoke inflation, as China's central bank would be increasing the money supply dramatically. An unsterilized peg then has two paradoxical and countervailing effects, amounting to a stimulus and tax at the same time. It taxes holders of domestic currency, by increasing the cost in local money of forign goods (an upward shift in the supply curve for imports). But it stimulates domestic demand for all goods. The net effect is almost certain inflation of the price of imported goods (relative to the price that would prevail without intervention), but a mix of real GDP growth (to the degree that there is "slack" in the economy ) and inflation in the domestic economy.

Now let's put all this together. If China were simply pegging its currency without sterilizing, it would stimulate demand for nonimported goods and services, but it might also stimulate inflation, permitting real exchange rate adjustment. But, by adjusting the degree of sterilization, it can seek a sweet spot that grows unsterilized money at a level chosen to maximize real GDP without provoking inflation, preventing real adjustment. The net effect is a synthetic tariff, raising the price of foreign goods while stimulating the domestic economy and providing revenue to the state. It's no wonder that China's trade partners object.

Steve Randy Waldman — Monday September 25, 2006 at 7:24pm permalink
Aaron Krowne (mail) (www):
How specifically is China implementing sterilization? It seems to be too precise to be occasioned through vague macro-policies like "making imported goods expensive" and "encouraging savings".
9.29.2006 1:50am
Steve Randy Waldman (mail) (www):
"Sterilizing" is a fancy word for a simple thing.

If a central bank wants to put a ceiling on the value of its currency with respect to some other currency, it can do so without problem. (We'll use China as an example, but this is a central banking thing generally, not a China thing.) Suppose China wishes to cap the value of its currency at roughly 12.6¢ (US) per Yuan. It can do so easily: It just offers to sell a Yuan to all comers for 12.6¢. This strategy is effective. The market price of a Yuan can never go higher than that 12.6¢, because anyone trying to sell for more would be undercut by the central bank.

But if there's a lot of demand for the Yuan, that means the central bank ends up selling a lot of them to enforce its price cap. In one sense that's no problem -- the central bank of China can print as many Yuan as it needs. But in another sense it is a problem, as maintaining the price cap effectively forces the central bank to increase the number of Yuan in circulation. Ordinarily, increasing the quantity of a currency in circulation decreases its value in terms of goods and services.

When a central bank "sterilizes" a currency intervention, it sucks back the new money it was forced to put into circulation by borrowing it back from the public. In this way the central bank gets to have its cake and eat it too: it stands ready to sell Yuan for 12.6¢, capping the value of a Yuan. But it offsets the increase in the money supply with a bond sale.

Note that selling bonds to the public is the ordinary way a central bank runs a "contractionary" monetary policy. When the US Fed "raises rates" to fight inflation, it often ends up having to sell bonds -- depressing bond prices, forcing up yields, diminishing the cash money supply -- to bring market rates up to the rate it has targeted. (These "open market interventions" are what the Fed Open Market Committee does.) This suppresses aggregate demand, by raising the cost of financing purchases.

"Sterilized exchange rate intervention" is the practice of capping a currency's nominal exchange rate while preventing the domestic inflation that would ordinarily provoke in the same way central banks always fight inflation. If successful, the net effect is to peg the real exchange rate.

"Real exchange rates" are a bit counterintuitive. Normally we think of inflation as decreasing the value of a currency. But in the context of a fixed exchange rate, inflation corresponds to currency strength! Let's talk it out: Suppose the US and China have a fixed nominal exchange rate, but there is zero inflation in the US, and 10% inflation in China. Both prices and wages are increasing in China at 10% per annum.

Suppose that today a US hamburger costs $1 and a Chinese worker makes 100 Yuan per day. 100 Yuan = $12.60, so the Chinese worker can buy 12 US burgers per day. Next year, a burger is still $1, but a Chinese worker earns 110 Yuan, due to inflation. The Chinese worker hasn't benefited in domestic purchasing power from his raise, because the price of Chinese goods has also risen 10%. But the price of US goods has not gone up at all, and the exchange rate has not changed, so the 10 Yuan raise is meaningful in terms of purchasing power of US goods! 110 Yuan is $13.86, so the Chinese worker can buy 13 hamburgers a day, when last year he could only buy 12!!! Inflation combined with a nominal peg has driven Chinese purchasing power of US goods up in the same way that an increase in the official value of the Yuan would have without inflation. Whenever quoted exchange rates don't adjust to exactly compensate for inflation differentials, there is some change in the "real exchange rate", that is the relative purchasing power of wages typically earned in one country for goods produced by another.

In theory, central banks ought to be able to cap nominal exchange rates, but real exchange rates should be market-determined, "endogenous" in the lingo. If a currency "ought" to rise according to "market fundamentals", a central bank's cap should force it to print more currency, increasing domestic inflation, and causing the real exchange rate to rise. In practice, central banks have been successful at manipulating real exchange rates via a combination of nominal exchange rate pegs, capital flow controls, capital market regulation and interventions, and sterilization.

And when central banks succeed in pegging real exchange rates, they are manipulating trade flows as surely as they might with explicit tariffs and subsidies.
9.29.2006 9:59am
Note: You don't have to register to comment! Just use the pull-down menu to comment as "Guest".