Karl Smith’s Less-Than-Soothing Fed Exit Strategy
I truly enjoy reading Karl Smith’s posts over at Modeled Behavior. Karl might be second only to Steve Landsburg in terms of thinking about standard issues in novel ways. (For example, check out this post where Karl consciously walks headfirst into the liquidity trap, and becomes more powerful than Paul Krugman can possibly imagine.)
So it’s with all due respect when I point with astonishment at Karl’s handling of worries over central bank insolvency:
A lot of smart folks seem to be worried about [central banks losing money] but I don’t completely understand why.
…
Lets suppose that either the Fed or the ECB takes a flyer on a bunch of debt. My immediate reaction is – so what?There is the issue that you can’t contract the total quantity of reserves down to an arbitrarily low level if you don’t have an asset to sell in exchange for them.
However, if you are paying interest on reserves then you don’t have to do this. You continue to print reserves in order to meet the interest payments. If the interest on reserves is higher than the rate of growth of checkable deposits then you are going to have the issue that excess reserves will grow without bound.
However, first its not immediately clear why that’s a problem and second its not likely that growth rate on checkable deposits will be less than the interest rate on reserves indefinitely.
Over the long long run the supply of money is going to grow roughly at the size of the nominal economy which in turn is going to be greater than interest on reserves.
So, eventually you will back down out of this process. However, even if for some reason you didn’t what exactly are you afraid of going wrong? Its not clear to me.
If nothing else, now I know how Joe Romm must feel when he reads typical libertarian arguments denying the danger of carbon emissions. (The inflation/climate change analogy is actually really interesting if you map out the standard way progressives and hard-money libertarians come down and the types of arguments they use.)
Yeah, I’d put myself into the “hard-money libertarians” category, and I used to be a believer in the whole climate change schtick until I started reading their websites closely and checking out how they could never give a straight answer to Steve McIntyre. Then climategate came along and I sat and read all the emails. I can’t say I was totally outraged by the emails, but there was clear evidence of scientific malpractice. What did outrage me was the number of people who claimed to have read the climategate emails and then strangely found “nothing to see here”, even when you point out particular issues they just shrug and ignore you. The massive aura of willful ignorance just cranked by suspicion meter up to 11. You don’t need to know the science to know it’s a scam, you just need to look at the people.
Anyhow, climate is difficult to measure. They have perhaps 150 years of surface temperature data and the daily variation is generally around 20 degrees, but then there’s variation depending on where you are on Earth, and also time of year and also variation from year to year and then there’s 30 year cycles and other things. More than that, stations have moved, the local environment has changed in places, measurement techniques have changed, and there have been distinctly non-random changes in the number of thermometers.
So they take all that noisy variability, and then they apply “homogenization” which supposedly cleans some of it up by injecting these adjustments (but when hauled up on particular cases such as Darwin, many of these adjustments simply have never been properly documented and cannot be either reproduced or explained). After the adjustments there’s an averaging process (which itself is complicated) and you get a figure of the order of somewhere between a half a degree per century and a whole degree per century. That’s at least 20 times smaller than typical daily variation and much smaller than the adjustments being inserted into the data. Needle in the haystack territory.
Now, when measuring inflation you run into similar measurement problems. Prices go up and down every day, all over the place, and that’s perfectly normal. However, we have perhaps a few hundred years of good quality price data, going right back to the world standard one-ounce silver dollar. On the hundred-year timescale, there’s an unavoidable trend of inflation that stands out in every price of goods that have remained vaguely similar for 100 years. Admittedly high technology items tend to get cheaper, and that’s a relatively recent situation, so over short timescales inflation is difficult to measure, and if you want an accurate measurement then it very much depends on your choice of items.
I think it’s fair to argue that there is no sensible comparison between my life and my great grandfather’s life in terms of the things I buy and sell compared to what my great grandfather did. You can compare the basics (e.g. land is the same, and is more expensive, food is the same and also more expensive but overall easier to get hold of) but what about computers, Internet, etc? Never the less, the dollar has clearly devalued during that time, the trend is unmistakable.
I’ll also point out that even gnarly gold coin advocates will accept that a few percent inflation is not the end of the world. As long as inflation doesn’t get out of hand, I don’t see a big problem, not that I’m saying it’s a good thing, but well there are much worse problems to worry about. The real problem of fiat money is the massive market influence that governments can leverage off it, and the resulting crazy bubbles that we have to ride through.
Now you want to look at that Craven guy with the silly hat telling everyone about Pascal’s Wager, as if he just discovered it. He waves around this half a degree per century drift in temperature as it if was panic stations, whoop, whoop, code red, the end of the world is nigh. Sheesh, please don’t go lumping me in with trash like that. I have some small amount of self respect you know.
“Over the long long run the supply of money is going to grow roughly at the size of the nominal economy which in turn is going to be greater than interest on reserves.”
I don’t get this. How can the money supply grow at the same rate as the “nominal economy”, when the money supply and real goods and services are incommensurable units? There’s no common denominator that can enable us to say something like “money supply grew 5% more than nominal economy” or “money supply and nominal economy grew at equal rates.”
Well I am sure Karl would say the common denominator is dollars. E.g. money supply went from $1 trillion to $1.05 trillion and nominal GDP went from $14 trillion to $14.7 trillion.
Yes, I suspect that too, but doesn’t Smith know that GDP is a function of prices, and that since prices are a function of the supply of dollars, then the very variable we’re trying to figure out how much should increase/decrease, dollars, is already implied in GDP? It’s circular logic.
It would be like determining where the thermostat in your house should be set, by adding up the average temperatures in each room of the house, and then observing what percentages they increased or decreased compared to yesterday, and then saying something like “the average temperatures in the rooms increased an average of 3 degrees compared to yesterday, so that means we have to set the thermostat 3 degrees higher as well so that they don’t cool down.”
MF wrote:
Yes, I suspect that too, but doesn’t Smith know that GDP is a function of prices, and that since prices are a function of the supply of dollars, then the very variable we’re trying to figure out how much should increase/decrease, dollars, is already implied in GDP?
I think that’s his whole point. If M1 in the long run grows at X%, then NGDP should too.
But, the problem with that thinking, is that X could be 1 or 1 billion. It’s not like the Zimbabweans were OK, because their nominal GDP kept pace with quantity of money.