If strong economic growth is not the explanation for the large increases since 2001 in prices of virtually all mineral and agricultural commodities, then what is? One wouldn’t want to try to reduce commodity markets to a single factor, nor to claim proof of any theory by a single data point. Nevertheless, the developments of the last six months provided added support for a theory I have long favored: real interest rates are an important determinant of real commodity prices. High interest rates reduce the demand for storable commodities, or increase the supply, through a variety of channels:
- by increasing the incentive for extraction today rather than tomorrow (think of the rates at which oil is pumped, copper mined, forests logged, or livestock herds culled)
- by decreasing firms’ desire to carry inventories (think of oil inventories held in tanks)
- by encouraging speculators to shift out of spot commodity contracts (think gold), and into treasury bills.
All three mechanisms work to reduce the market price of commodities, as happened when real interest rates where high in the early 1980s. A decrease in real interest rates has the opposite effect, lowering the cost of carrying inventories, and raising commodity prices, as happened in the 1970s, and again during 2001-2004. It’s the original “carry trade.”
The theoretical model can be summarized as follows:
A monetary expansion temporarily lowers the real interest rate (whether via a fall in the nominal interest rate, a rise in expected inflation, or both – as now). Real commodity prices rise. How far? Until commodities are widely considered “overvalued” — so overvalued that there is an expectation of future depreciation (together with the other costs of carrying inventories: storage costs plus any risk premium) that is sufficient to offset the lower interest rate (and other advantages of holding inventories, namely the “convenience yield”). Only then do firms feel they have high enough inventories despite the low carrying cost. In the long run, the general price level adjusts to the change in the money supply. As a result, the real money supply, real interest rate, and real commodity price eventually return to where they were. The theory is the same as Rudiger Dornbusch’s famous theory of exchange rate overshooting, with the price of commodities substituted for the price of foreign exchange.There was already some empirical evidence to support the theory: Monetary policy news and real interest rates, along with other factors, do appear to be significant determinants of real commodity prices historically. (For a simpler illustration, see graph below).
But the events since August 2007 provide a further data point. As economic growth has slowed sharply, both in the US and globally, the Fed has reduced interest rates, both nominal and real. Firms and investors have responded by shifting into commodities, not out. This is why commodity prices have resumed their upward march over the last six months, rather than reversing it. 
[...] ETF Trends - Keeping a grip on exchange traded funds (ETFs) wrote an interesting post today onHere’s a quick excerpt If strong economic growth is not the explanation for the large increases since 2001 in prices of virtually all mineral and agricultural commodities, then what is? One wouldn’t want to try to reduce commodity markets to a single factor, nor to claim proof of any theory by a single data point. Nevertheless, the developments of the last six months provided added support for a theory I have long favored: real interest rates are an important determinant of real commodity prices. High intere [...]
[...] Frankel expains why commodities’ prices have been [...]
[...] think the Fed missed an opportunity here. A 25 or a 50 basis point cut would have sent commodity prices crashing. Even the mildly hawkish surprise of “only” a 75 basis point cut may have some effects [...]
That regression line looks like pretty weak evidence. If you took out the three points to the northwest you wouldnt have much of a relationship at all. Either there are threshold effects for extremely negative real interest rates and/or something else is going on. I would not like to hang my hat on a regression where the whole argument seemed to rest on three observations. What was going on with those three? And more importantly what explains the huge scatter around the line for the rest of the observations?
If you took out those three points and the line you drew in, how many people would draw the relationship as you have? Not too many I suspect, and I bet the regression package you use would confirm that it is a pretty useless model.
THat isnt to say that your theoretical argument is wrong - but there is a lot more going on than your argument can explain
Interesting look at the relationship.
High prices for commodities and bonds both aided by excess liquidity. Maybe y/y global broad minus narrow money explains both.
[...] … I think the Fed missed an opportunity here. A 25 or a 50 basis point cut would have sent commodity prices crashing. Even the mildly hawkish surprise of “only” a 75 basis point cut may have some effects [...]
You see a very good correlation between total US debt growth since 2002, and the historical low in commodity prices between 1999-2002. Commodity prices had been falling in real terms for decades, while the US government had flooded the market with deficit spending in the wake of 9/11, while the Fed kept real interest rates low.
The informal Bretton Woods II agreement between oil exporters and Asian manufacturers either through their quasi-US dollar pegs or through concerted intervention simply meant that they were artificially keeping their currencies weak against an already weak US dollar that subsequently weakened further. Hence why almost all asset prices rose globally from Shanghai to sugar.
The 25% kicker we got in commodities since the Fed eases in the third week in January was merely a desperate attempt by investors into the last asset class left standing as a hedge against falling financial asset prices elsewhere and as a hedge against a weak US dollar and future inflation expectations going forward.
Those gains are now retracing as traders realize that firms struggling with growth and profitability will be hurt by those higher input prices at a time when they have very little pricing power. But it is still worth remembering that $12 wheat in 2008 is equal to $1.69 per bushel in 1958 inflation-adjusted US dollars. Commodity prices relative to say houses are still historically very low!
Happy Easter!
Bill.
The commodities bubble has a much more simple dynamic than this. Liquidity flows to the rising asset, just as with the dot.com boom and the housing boom. Commodities got a bump with the collapse of real estate, then grew with the force-feeding of liquidity by the Fed.
The Fed’s aggressiveness, obviously, has produced no real results (other than inflation pressure) and this is why.
I would like to believe your assertion. However, if you drop the three data points below -4.0% real rate(or -2.0% — given the large gap), the relationship between commodity prices and real rates looks uncorrelated. To be fair, if you also all trim away the data points associated with the three highest real returns, the relationship still looks uncorrelated.
How do you explain the late 1970s and early 1980s when there were high priced commodities and the interest rate was over 10%?
A more likely explanation for the increase in commodity prices is the movement from global liquidity from an asset bubble in equities, to real estate, which has now settled in commodities. This would explain the recent run up in gold to over $1000 an oz. Cuts in interest rates have accompanied the deflation of the asset bubble, but this does not mean they are a direct result of low rates. Correlation does not necessarily prove causation.
The Yield Curve and the Mineral Bubble
The Hidden Parameter in Interest Rates
Executive summary:
There is a strong link between the evolution of the market price of minerals and the shape of the yield curve
The idea is that given the shape of the yield curve the marginal cost of extraction of minerals becomes irrelevant to their market price as miners stop maximizing their output under constraint of the marginal cost of extraction.
Profit maximization would have them trying to retain their minerals rather than extract them.
The Hidden Parameter:
The price of minerals has grown unabated since the Federal Reserve has started increasing short term rates above 2.5%.
All of the minerals have grown together, which cannot be explained by the growth of the marginal price of extraction alone: no price increases have caused the price of any mineral to stop its growth as a result of an increased investment in exploration.
This correlated increase in the price of minerals must be caused by a global parameter.
Harvard Economics Professor Jeffrey Frankel made the hypothesis that a decrease in real interest rates (”real” rates exclude inflation) increases the demand for storable commodities.
However during the increase of short term rates from 1% under the chairmanship of Alan Greenspan, inflation didn’t grow as fast as short term rates still minerals kept growing.
Moreover any storage outside the ground would not be economically viable as the cost of storage would be added to the interest rates.
My hypothesis is different: financial decisions are about choices: if you consider the minerals as short term assets, you come to the conclusion that miners, confronted with an inverted yield curve would, as a group, prefer to hoard their minerals in the ground where storage cost is almost free rather than sell them and invest the proceeds in long term assets.
How else would we understand that the cost of oil was multiplied by 5 over such a short period with a low depletion of the proven reserves?
How else would we understand that all the minerals saw their cost rising at the same period with a next to perfect correlation?
How else would you explain the fact that the rise of minerals started shortly after the rise from 2.5% of short term interest rates by the Federal Reserve?
At the same time the yield curve was under the influence of the famous Greenspan Conundrum, which caused the inversion of the yield curve.
Should the yield curve on the U.S. dollar return to normal, as it did on Monday, the miners would stop hoarding their reserves in the ground and the prices should go down in the direction of their marginal cost of extraction.
That price must be much lower than the expectation of all market participants.
Because of hedge funds holding, we should see some overshooting, in particular with minerals with low industrial use: gold and silver.
The correlation between the shape of the yield curve and the price of commodities is only one of the many overlooked signals that are embedded in the yield curve.
Among others, it can give a precise timing of a possible systemic collapse through a Keynes’ Liquidity Trap.
My model of the yield curve never gave a signal of a systemic collapse during the recent credit crisis: it has always forecasted that the Federal Reserve had sufficient room to rescue the market and the economy.
Shalom Hamou
Independant Yield Curve Special Advisor
shalem.ashalem@gmail.com
[...] of the spurt in commodity prices. If they are due to loose monetary policy, as suggested by Frankel [1] (manifesting itself in negative real interest rates) and more recently by Jim [2], then high [...]
[...] sent commodity prices crashing. … commodity prices had been falling in real terms for decades, …http://content.ksg.harvard.edu/blog/jeff_frankels_weblog/2008/03/17/falling-interest-rates-explain-r…Commodity Prices And Currency MovementsFind out which currencies are most affected by fluctuations [...]