The Chinese Fiscal Stimulus, correction

Damn, this gets so complicated when you try to do it rigorously. In my earlier post, my mind had scrambled together 3 or 4 different models, thrown in some extra stuff to try to make the result realistic, and ended up with a high-cholesterol omelet. Now that I have separated the yolk from the white and the white from the cheese, and taken out the bacon that was supposed to make it realistic, the result is much tighter theoretically but difficult to explain in a single blog post. The bottom line is that, according to standard Keynesian-style models with (usually assumed but obviously not realistic in China’s case) perfect capital mobility and perfect asset substitutability, the Chinese stimulus has an unambiguously positive (well, non-negative, anyhow) effect on non-dollar countries and an ambiguous effect on dollar countries (particularly the US).

The two critical elements of the reasoning are

1. that the stimulus causes Chinese to import more (possibly indirectly) from the US (likely, since China is not self-sufficient, and some of the additional things the Chinese will want to buy will be things that are made in the US – certain kinds of business services, for example).
   
   
2. that the stimulus causes the value of the dollar to rise against floating currencies such as the euro (usually considered likely for the reason I gave in my previous post – that the financing of the stimulus will put pressure on credit markets, thus raising dollar interest rates and making dollars more attractive relative to the floating currencies)

Contrary to what I said in my earlier post, the first of these effects should dominate the direct effect of rising interest rates rather than the other way around, and if it weren’t for the floating currencies, the stimulus would be at worst neutral, and more likely positive, for the US. However, the indirect effect of rising interest rates – namely the strengthening of the dollar – is negative for the US, since it makes US goods less attractive relative to goods from floating-rate countries. The direction of the net effect on the US is ambiguous. One should note also, though, that the stimulus will have an unambiguously positive (well, again, non-negative) impact on the floating rate countries, because (among other things) it will weaken their currencies and make their goods more attractive.

I started out writing a post that tried to explain all this in more rigorous detail, but it got hopelessly long and complicated. I might put the explanation in the next post, but I’m trying to keep this one reasonably user-friendly

When you introduce imperfect capital mobility and imperfect asset substitutability, the story becomes even more complicated, and I haven’t even thought through the implications. Imperfect capital mobility and imperfect asset substitutability essentially represent the idea that China’s monetary policy doesn’t have to depend on its exchange rate policy. In other words, China can affect the exchange rate by, in effect, swapping huge quantities of Chinese bonds for huge quantities of US bonds. Institutionally speaking, China issues renminbi bonds, uses the proceeds to buy dollars, and invests the dollars in dollar bonds (really in Treasury notes, mostly). Ordinarily the effect of such intervention would be offset by the actions of private investors, but China can do it on such a massive scale that it overwhelms the actions of private investors. Moreover, China can restrict the actions of investors.

To the extent that imperfect asset substitutability is important here, I got things completely backwards (sort of) when I said

...given China’s de facto currency peg, [the stimulus] will have to be financed by reducing purchases of foreign assets (particularly US government securities).

Wrong! Actually, if China operates fiscal policy, monetary policy, and exchange rate policy independently from one another, it will have to buy more US government securities, not less. The fiscal stimulus will raise Chinese interest rates (and also probably make investment activities in China more profitable in general), thus making renminbi more attractive to investors. The Chinese authorities will have to offset the incipient increase in the value of renminbi by absorbing all the new dollars which investors want to exchange for renmibni that they can use for investment in China. Then the authorities will invest these dollars largely in US Treasury securities.

But here’s where the “sort of” part comes in: the increased demand for US Treasury securities, while indeed it doesn’t increase overall stress on US asset markets, doesn’t actually reduce such stress either. The whole point of buying the dollars to be invested in US Treasury securities was to offset what private investors were doing: taking dollars out of US assets and putting them in Chinese assets. All things considered, the effect China’s attempt to maintain its exchange rate will indeed be bad, because it will shift demand out of US private assets into US public assets, when the big problem in the US now is that everybody wants Treasury bills and nobody wants anything else. We want people to keep their risk capital in the US, not move it to China. So, the reasons are much more complicated than I thought, but we can still expect that the Chinese stimulus might have a detrimental effect in the US.