May 16, 2006
Commodity price inflation
Commodity markets have been a little too exciting recently for my quiet tastes.
|Commodity||Jan 3-May 12||May 12-May 15|
Up until this week, it had certainly been a spectacular ride for many commodities this year, with copper and zinc, for example, nearly doubling in price since January. One popular interpretation is that this increase in commodity prices signals an incipient resurgence of inflation.
I know this may take all the fun out of the argument, but before we get into a discussion of whether this is or is not inflation, it might be helpful to define what we mean by "inflation." What I mean by "inflation" is a deterioration in the purchasing power of a dollar bill.
Now, isn't it clear that there's quite a bit of that precise phenomenon manifest in the above charts, in that a dollar bill will buy you a whole lot less copper than it did six months ago? Well, yes, but a bushel of oats or a pound of coffee will also buy you a whole lot less copper than they would six months ago as well. And when a bushel of oats buys you a whole lot less copper, I want to describe that as an increase in the relative price of copper rather than as a pure monetary inflation.
The reason I want to use my definition is that I'm particularly interested in whether the Federal Reserve may have been creating too many dollars and thereby contributing to the phenomenon of inflation as I have defined it. I don't find it natural to assume that an excess quantity of dollars being printed should logically result in needing to pay more coffee to get some copper.
It is true that there is a broad tendency for many of these items, particularly the precious metals and commodities used in manufacturing as opposed to agricultural goods, to have moved up together over the last year, suggesting that there is some common cause behind that price surge. Besides the monetary inflation story, another popular account (e.g., see Abnormal Returns or Tim Iacono) is that commodities generally are experiencing a speculative price bubble. Certainly the fact that so many of these prices dropped spectacularly together on Monday would cause one to take this hypothesis seriously, as for example both palladium and zinc dropped by about 10% in a single day.
One challenge for either the monetary inflation or the speculative bubble view of the commodity price movements is the fact that there is such diversity, with some commodities going up a great deal and others not at all. Indeed, these differences across commodities are actually much bigger in magnitude than the average movements common to them all. Although not entirely discounting the potential role of monetary or speculative factors, I'm therefore inclined to try to interpret much of the relative price movements as resulting from the same factors that have always made commodity prices much more cyclically sensitive than other prices. Specifically, the long lead times in production and short-run demand and supply inelasticity mean the price can be particularly sensitive to demand fluctuations. A broad increase in the level of economic activity can thus lead to a broad increase in the relative price of a particular group, though with huge differences across items reflecting the particular factors of supply and demand in each market.
What, then, was the news that caused the spectacular price drops on Monday? I could call attention to bearish indications about American consumers. Some might say, but the commodity price inflation of 2006 has clearly been fueled by global economic activity. Quite true, though if the American economy goes into recession, I'm quite certain that Chinese manufacturing will notice it.
Even so, a 10% price drop in a single day seems a bit much to try to lay all on fundamentals, leading me to wonder if recent market moves might at least in part have resulted from the actions of some speculators who will soon be parted from their wealth.
Posted by James Hamilton at May 16, 2006 02:49 PMdigg this | reddit
As the US economy moves into the future, the ability for the Federal Reserve to control commodity prices by slowing the economy might be reduced. If countries such as China or India have a larger effect on commodity prices, their consumption may charge ahead regardless of the U.S. Economy. I do realize that the U.S. economy is still huge, but that will change over time.
Copper used to be an economic indicator for the U.S., but it now may be more of a global indicator as other economies consumption grows relative to the US's. The WSJ had a piece on "Dr. Copper" yesterday.
Posted by: Mcwop at May 16, 2006 03:20 PM
One of the things that's always puzzled me about financial markets is, what determines how much the price usually changes on a day-to-day basis?
We've got these price changes of 10% and we're kind of shocked. That seems like a lot! But compared to what? Well, compared to how much it usually changes.
Usually financial prices change every day by a percent or so, maybe a couple of percent, maybe a little less. Oil's at 70, it's going up or down by a dollar or so, a bit more than a percent. You have a stock trading at 20, it changes by half a point, that's 2.5%. Changes of a percent or two are very normal.
But why is this? What is it about the nature of reality which causes financial prices to typically change by a percent every day?
I could imagine some alien race which has financial markets which change much less. Maybe every day their prices only change by 0.01%. If they ever have a day when a price changes by 0.1% they're all in a panic! The sky is falling! Or maybe there's some other aliens and their prices change by 50% every day. For them, a 10% day means the market was unchanged.
Are these absurd scenarios? It might seem so. We're so used to our markets, it's hard to imagine how prices could be vastly more stable or unstable.
But in a way, a percent per day seems like a lot to me. How could the value of something change by that much in just one day? Did we really gain that much more information about the world overnight, that we reevaluate everything by as much as one percent? Is my toothpaste worth a percent more today than yesterday? Is my chair worth a percent less? I don't think so. Most things seem pretty stable in terms of their value. But financial valuations are more likely to fluctuate.
Is there any economic theory that lets you plug in human psychology and/or natural law and out pops "1 percent per day" as a natural level of variation in markets?
Posted by: Hal at May 16, 2006 03:30 PM
Am I belaboring the obvious? Commodity prices are some varying mix of demand (China, India), inflationary pressures (until recently by the dollar and yen) and speculation. I thought good economists got paid to sit around and tease out how much of each of the above was responsible for x movement in x commodity.
Posted by: Qingdao at May 16, 2006 05:14 PM
Ummm, it doesn't seem to me that commodity prices by themselves indicate much of anything about inflation (with inflation measured by the purchasing power of a dollar for a GDP-representative basket of U.S. goods AND SERVICES).
From a global perspective, there's an unusual dichotomy between labor, which is abundant and cheap, and certain strategic raw materials necessary for industrialization (copper, zinc, crude oil and met. coal, for starters), which are scarce and expensive. At present, the two most populous nations on the planet are moving rapidly toward developed nation status and demand for key commodities has ratcheted up much faster than new supplies can be brought on line. The unsurprising result is that prices of the most critical raw materials have increased sharply.
Large daily price changes up and down are just noise and should NOT surprise anyone considering how large the more significant 6, 9 and 12 month price changes have been. Prices have gone up substantially in the past 6-12 months because the demand/supply balance has changed dynamically; in such a fast-changing environment prices also should shift rapidly as participants struggle to adapt to new realities that are hard to interpret.
The bell-cow of the herd is copper; keep your eye on that one, it'll tell the tale.
Posted by: Anarchus at May 16, 2006 06:09 PM
I believe commodity prices are relevant to CPI inflation. Commodity prices often rise before consumer prices because the FED/banking system creates money that is first used by producers. The fresh money is lent to entrepreneurs who spend it on the factors of production. Over time the money finds its way to consumers.
Posted by: algernon at May 16, 2006 06:39 PM
The following article from Scientific American might be of interest. Some excerpts follow.
“The researchers went on to examine per capita use of copper in the U.S. and other developed countries. While some theorists had predicted that metal use would decline as economies advanced beyond building metallic infrastructure, the teams' data showed that overall copper use in the U.S. climbed to a high of 238 kilograms per person by 1999. Declines in areas like manufacturing and railroads were more than offset by increases in areas like motor vehicles and domestic devices. In fact, residents of Canada, Mexico and the U.S. required an average of 170 kilograms of copper per person. Multiply that by overall population estimates of 10 billion people by 2100 and the world will require 1.7 billion metric tons of copper by that date--more than even the most generous estimate of available resources.
"Certainly every square meter of earth hasn't been dug up but there aren't many places that haven't been investigated pretty thoroughly," Graedel notes. "We're not going to suddenly discover a new continent."
“The same dynamic also holds true for other critical metals such as platinum, required for catalytic converters and other pollution control devices, and zinc, used to galvanize steel. While iron, aluminum and other more abundant metals could conceivably be used as substitutes, more research would need to be directed into such technology shifts, the team writes in the paper published online this week in the Proceedings of the National Academy of Sciences. And certainly as much currently processed copper, platinum and zinc should be conserved as possible; the world needs it. "Either the rest of the world can't live like the developed world or we need, as a society, to think more about the technology of providing these services with less intensive use of at least certain materials," Graedel explains. "We need to do a more diligent job of good housekeeping."
Certainly we will have our ups and downs as we push ever onward and finally upward. Speculators will try to be ahead of the curve…and lose their shirts.
do you not see a curious parallel with all the arguments concerning energy and oil? I prefer the long view. China, India, and the rest…are Johnny-come-latelys to the party. Not enough time, not enough resources.
Posted by: Stormy at May 17, 2006 07:32 AM
Thanks for making some good points and providing useful information, Stormy. Yes, I do see some parallels, which is one of the reasons I'm following all this with interest.
Posted by: JDH at May 17, 2006 08:07 AM
"While iron, aluminum and other more abundant metals could conceivably be used as substitutes, more research would need to be directed into such technology shifts"
Aluminum is an adequate and proven substitute for copper for home wiring. It was used widely several decades ago during a similar peak in copper prices, and abandoned when copper prices returned to earth. It's not as desirable as copper (in particular, it expands and contracts more than copper, which creates connection problems), but it's workable and proven - no research needed.
Posted by: Nick at May 17, 2006 01:48 PM
Arguing whether massive increases in price is inflationary is akin to arguing whether water is wet. Inflation is higher prices.
The current rise in commodity prices has nothing to do with free market demand. It is a rise caused by a cornering of a market by financial groups who have been allowed to buy up the supply of a commodity and ration that supply back to would-be consumers from a position of monopoly.
We have seen this behaviour all year through the futures market of various commodities. The natural gas debacle in early winter, oil and Goldman Sachs' $100 super spike, the metals markets, and the recent oil spike.
This is not free market behaviour. This is massive manipulation by a financial industry once again out of control. The last economic recession followed on similiar debacles including the internet swindles, the California energy swindle, the telco swindles, the commodity energy trading swindles, among others promoted and conducted by a financial system unleashed from the restrictions placed on them following their role in the great depression.
Perhaps it may be argued that prices recede back following one of these swindles and therefore they are not inflationary. But, if we spend twice what we normally would this period, but next period we spend what we normally would have, we are none the less worse off because the two period cost averages 1.5 times the prior period.
The arguments that commodity prices should be high because, logically, China and India have more people and therefore will eventually consume all that we can possibly produce .... is not a logical economic analysis. We, me and you, have no way of knowing what the future consumption patterns will be nor do we know what technological substitutions are on the horizon. What we have is the current supply and the current demand in the physical market. When inventories are growing or stable, as in metals, crude oil, natural gas, and pricing goes parabolic ( 70 -100 % increase in two months), the market mechanism is malfunctioning, whether by hysteria or by manipulation. Both have enormous precedent throughout our history.
Posted by: saywhat at May 17, 2006 09:01 PM
Saywhat: no, inflation is not the same as higher prices. The price of oil may increse while prices of computers, cell phones and fish fall. That, by definition, is not inflation. And do you really dismiss the arrival of 2 billion consumers?
Posted by: Qingdao at May 18, 2006 01:18 AM
Regarding your 1% variation question.
Perhaps its "Benfords Law" at work? Don't know if you know about it - but its a curious little law also known as the "First Digit Law"
The basics of it, is that the first digit of many natural measurements are "1". The second most commond first digit is "2" and so on.
There is much debate to why this is.....but it does seem to hold true for many things.
Posted by: Barry S at May 18, 2006 11:49 AM
Rather than signalling inflation higher commodity prices contribute to causing inflation. The concern would be that as the higher commodity prices feed through into final demand that there is an impact in the general price level.
If you were to separate commodity prices into those that were readily tradeable in markets as compared to those where only trade buyers and sellers operate you would have a better idea of speculative impacts.
Posted by: David Leitch at May 18, 2006 06:29 PM
Show me where the 2 billion consumers are and I'll let you know if I dismiss them or not.
That's one of the great problems. They don't consume, only produce and save. Personal consumption expenditures in China accounts for less than 15 % of GDP, and very little of it is American goods and services. How big is that trade deficit?
They have tapped into the consumption cycle of the US and are holding on with all their might. We'll see how much they consume, and especially of American goods, when the US consumption carcass has been sucked dry.
Posted by: saywhat at May 18, 2006 07:27 PM
By the way, I find this to be a fine weblog.
It seems, to me, that |saywhat| is speaking more truth than many like to think about. Though, there are, obviously, more than a few simple drivers in play v. commodity prices. I think a main one is the World's growing lack of confidence in the future utility of Fiat v. hard assets. As well, coming out of the deep trough in commodity prices, producers, on the whole, are ill-prepared to rapidly ramp production. The pure metals, still widely available from exchange stocks, are a grand hedge against multitudinous uncertainties. And, yes, of course, the worldwide cascade of Fiat emanating from central banks across the globe, is certainly adding fuel to many fires, including commodity prices.
The certainty, going forward, is the uncertainty about the probability of discontinuity.
Metals, in this environment, are cheap, and whether the future offers deflationary depression through to hyperinflated mania, they will remain dear.
Posted by: Mark E Hoffer at May 28, 2006 04:47 AM