Thoughts on the dollar, treasuries, and inflation

Although unusual from my recent ramblings... this post will be more coherent since I wrote it in response to a blog entry from Stefan Karlsson. I saved it because there is a chance that he might delete it.

I thought long-dated bond yields would go up despite a deflationary environment. My reasoning is based on the fact that in a deflationary environment, people generally would move towards liquid assets and relatively shun "duration" exposure so the yield curve would steepen unless the front-end interest rate was high (which is usually the case before a recession). Also, Brad Setser reported increased demand for treasuries by private investors in the 3rd and 4th quarter of 2008. (Setser, to the best of my knowledge did not say anything about the duration preferences of these investors.) I think that this was a one-time maneuver by many investors shifting a large portion of their portfolio into bonds (bond demand would not be financed by a stream of income by private investors) so a large portion of their capital is already exposed to treasury bonds. In addition, foreign demand for long-term treasuries I expected to decrease as government's acknowledge the risk of the "risk-free" long-dated asset (Sester did acknowledge this) and lower trade surpluses in many of the US' trading partners. Lastly, even though private saving would go up, I suppose it would not finance the large deficit as the private sector attempts to deleverages themselves and prefer more liquid shorter-duration exposure. In addition, earning power would decrease so less money would be available for saving which is different from the Japanese scenario. It would seem that the increase in supply for treasuries (especially long-dated) would outstrip an increase in demand if we excluded the printing-press and people suddenly buying bonds after they heard of Bernanke's QE policy (even if there was not sell-off caused by a bursting of the "bond bubble"). I thought Bernanke would result to using QE eventually, but I actually expected yields to go up during the bear market rally.

I did expect the dollar to rise against the euro because it is the world's reserve currency and the world is highly levered in the world's reserve currency so this would increase the demand for dollars during deleveraging. But once the deleveraging was over, the dollar would resume its fall. In addition, the eastern european exposure to the euro would cause significant stress to the currency regime and maybe lower euro interest rates in an attempt to "stimulate" the economy. Also, I erroneously thought that the US trade deficit would narrow (it did), but the latest trade data showed that exports declined more relative to the decline in non-petroleum imports.

Of course, in the future, there is a non-trivial chance that the velocity of money would dramatically accelerate. This would cause inflation or hyperinflation (as defined by an increase in prices), not simply increases in the money supply although that might be the catalyst to trigger such an increase. Of course, inflation and hyperinflation do have a reflexive component to it, and since people believe that the Fed will cause inflation, as it did by increasing the money supply, it drives gold prices up as people seek "real" exposure, not "nominal" exposure. Gold prices do not measure directly inflation as there was positive inflation in the 90's yet lower gold prices as people wanted exposure to assets (or "delusion" in the case of most dot-coms). It measures some derivative of the demand for "real" exposure relative to "nominal" exposure which would increase during times of both deflation and in some inflationary environments. In a deflationary environment, "nominal" low-duration exposure might not offer significant interest, so the opportunity cost for holding gold would be lower. This is one reason why gold could rise in a deflationary environment.

I do not think "reflation" will "work" (which I'll define as producing "desired results" not merely producing inflation.) Even if banks become well-capitalized when the government monetizes their assets, it would not spur "enough" lending because the demand for lending would decrease. While Bernanke can lower the interest on interest-bearing assets by intervening in the bond market by buying agencies and government debt, it would not necessarily lower the "spread" or risk premium for those type of debts and consumer loans. The spread, of course, needs to compensate for the risk in the loans the banks make. I expect a situation similar to that of Japanese banks if the banks are to become well-capitalized. It seems that Bernanke's goals for QE is to lower the interest rates for consumers to get loans rather than recapitalization of banks although the latter is means for the former. I do not think QE will achieve this end. QE might not necessarily cause inflation in a given nation if it is not invested in a the country (in Japan, it might be used for "carry trades" in other currencies) or if the excess money was withdrawn from the money supply after it recapitalized the banks (which happened in Japan according to the figure here).

I still like the deflationist thesis and "Mish" (and not Peter Schiff) made me consider the merits of Austrian empirical economics, although I do not like some Austrian normative economics. Regardless of "deflation" and "inflation," this is an environment characterized by hoarding not entrepreneurialship... who knows whether it will be cash when the velocity of money is low, or real assets.