The commodities crash

No doubt you will have noted the collapse in commodity prices that accelerated on Friday night, even as equity markets remained flat. Just in case you are unaware, here are a couple of quick charts. First the CRB:

And its more volatile cousin, the CCI:

The crash is across the commodity spectrum but is especially severe in metals (as a quick aside, I’d say Australia’s terms of trade have definitely peaked for this cycle).

A number of reasons have been put forward for the rout by any number of analysts. The most popular is that nice easy sound bite, easing global growth means the less demand and greater surplus of supply. This quote from Bloomberg is typical:

“We are seeing commodity prices correcting, so they are more compatible with the global economy,” said Christin Tuxen, a senior analyst with Danske Bank A/S in Copenhagen. “When we have fears over the economic cycle as we have now and a higher probability of contraction, it hits industrial metals and commodities.‘‘

And no doubt this is true to an extent. The analysis was most oft coupled with the explanation that last week’s concerned assessment of growth prospects by the Federal Reserve  has turned sentiment. But another question we might ask is, if it was growth worries alone that suddenly smacked commodities, then why did the equity market hold up? Moreover, there was no news on Chinese growth, the key source of demand, to trigger the bust.

In my view, therefore, it’s not enough to say that diminishing growth prospects triggered the rout.

To me, it’s was not what the FOMC meeting said, but what it did that made the difference. Its decision to proceed with Operation Twist, rather than ramp up another round of asset purchases that expands its balance sheet (ie, no more money printing) deeply disappointed commodity markets. Why so?

With no more money printing, the $US is now liberated from the simplistic monetarist view that more dollars means cheaper dollars. And global markets, which run on a simplistic monetarist paradigm, are responding accordingly, with a $US rally.

That means all things ‘undollar’ must fall. That is, all things that are priced in the $US are suddenly getting cheaper, and that includes, above all else, commodities. What I am basically describing here is the monetary deflation of commodities. We might also call it a reversal of the ‘financialisation’ effect.

So, is there no fundamental supply and demand paradigm at work here too? Yes, there is. The Economist has a nice article this week which picks up on a theme I’ve discussed before too, the crowding out of developed economies:

Broader measures of raw-material costs have jumped as well. The Economist’s index of non-oil commodity prices has trebled in the past decade. The recent surge has reversed a downward trend that had lasted a century. Industrial raw-material prices fell by around 80% in real terms between 1845, when The Economist began collecting data, and their low point in 2002 (see chart 3). But much of the ground lost over 150 years has been recovered in the space of just a decade.

This has raised the incomes of commodity-rich countries such as Brazil and Australia as well as parts of Africa. It has also caused even sober analysts to speak of a “new paradigm” in commodity markets. Even though GDP has slowed to a near-standstill in many parts of the rich world, the price of crude oil is close to $100 a barrel—as high in real terms as after the oil shocks of the 1970s.

What accounts for this turnaround? The price spikes over the past century were linked to interruptions in supply, notably during the first world war. But recent price rises have been too broad-based and long-lasting to be adequately explained by frost or bad harvests. Nor is it obvious that producers are hoarding supplies. At Tubarão everybody is straining to get the ore onto the ships more quickly. Valuable time is lost in docking and in starting the loading.

Optimists bet on human ingenuity to spring the Malthusian trap, as it has done so often before

There is a more straightforward explanation for the scarcity: the surge in commodity prices is simply the result of exploding demand and sluggish supply. The demand side has been boosted by industrial development unprecedented in its size, speed and breadth, led by China but not confined to it. Growth in emerging markets is both rapid and resource-intensive. The IMF estimates that in a middle-income country a 1% rise in GDP increases demand for energy by the same percentage. Rich economies are far less energy-hungry: the oil intensity of OECD countries has steadily fallen in recent years.

China’s appetite for raw materials is particularly voracious because of the country’s size and its high investment rate. Though it accounts for only about one-eighth of global output, China uses up between a third and half of the world’s annual production of iron ore, aluminium, lead and other non-precious metals (see chart 5). Most of the energy for Chinese industry comes from coal—a dirty fuel that contributes to China’s poor air quality. Its consumption of oil roughly tallies with the economy’s size but is likely to grow faster than GDP as China gets richer and buys more cars.

Supply has struggled to keep pace with this burgeoning demand. The world’s iron-ore production has doubled over the past decade but prices have risen 13-fold. The metal content of copper ore has been falling since the mid-1990s as existing mines are depleted. This mismatch between demand and supply is an age-old problem in commodity markets. It takes years to find and develop new mines and oil reservoirs and to build the infrastructure (rigs, pipelines, railways, ports) to bring the commodities to market. Supply responds slowly to price increases and delay often leads to excessive investment which then depresses prices.

So now I am blaming supply and demand! How inconsistent. Well…no. Both financialisation and the supply/demand imbalances are at work. Let me explain.

A year or so ago, Der Speigel carried a brilliant story on the rise of copper which included this era-defining chart from Bloomsbury Mineral Economics:

It shows that there is undoubtedly a demand boom. But it also shows an extraordinary explosion in price, beyond historical precedent. To me, it is absolutely no coincidence that the break with traditional fundamentals (and the beginning of financialisation) occurred in 2007 when the $US first began to substantially weaken as the housing bubble began to burst and US monetary policy headed south.

There is, of course, another element to the financialisation process; the accompanying transformation of commodity markets themselves into speculator casinos. The crash in metals has surely been pushed along by some nasty margin calls, not to mention the reversal of such schemes as Chinese ‘copper as collateral‘, as the new Fed position filtered through trader’s minds. But that is not my focus today.

My main point is that analysis of commodity markets tends to finger either the supply/demand imbalances or financialisation exclusively for sky high commodity prices. I think they do so for political reasons, adding another layer of complexity. But there is no reason for them to do so. Classical economic theory clearly explains the twin and simultaneous effect of financialisation and supply/demand imbalances (one of the few things it does get right!). To my mind, the following chart is the most important in contemporary commodity markets. As I have written before:

Like housing supply in Australia, commodity supply is inelastic. That is, it cannot respond quickly to a sudden surge in demand. The chart offered below (and mentioned above) shows the effects on any given market if supply cannot respond quickly. Don’t be scared of it, it is easier than it looks:

Q0 and P0 represent the initial equilibrium situation in any market. Initial demand is provided by D0, whereas supply is shown as either SR (restricted) or SU (unrestricted).

Following an increase in demand, such as a surge of emerging markets looking to engineer an historically swift catch-up in living standards through mass urbanisation, the demand curve shifts outwards from D0 to D1. When commodity supply is restricted, prices rise sharply from P0 to PR. By contrast, when supply is unrestricted, prices rise more gradually from P0 to PU.

The situation works the same way in reverse. For example, if there was a sharp fall in demand following a contraction like that of the GFC or an inflationary bust causing commodity demand to fall from D1 to D0, then prices fall much further when commodity supply is constrained.

The graph illustrates that demand shocks combined with inelastic supply do not result in a one way bet of upwards price movements. Rather, such economic settings produce volatility, with steeper price rises and spectacular collapses.

Why you might ask? It’s pretty simple. This is a mathematical representation of human panic. When a market is perceived to be unable to increase supply easily then speculators move in. In strategic markets like oil, governments begin to fret about security of supply. They stockpile. More speculators enter the market. So on, and so forth.

So long as the perception that supply is constrained remains, the frenzy continues until it exhausts itself in a new crash.

In other words, in markets that can be represented as supply constrained, you get rolling bubbles and busts. Or, put another way, at the heart of every bubble is a grain of truth, but that does not mean that it is not a bubble.

Right now we are seeing the reversal of commodity prices pumped by financialisation and by growing concerns over growth. But China is so far growing well and many analysts still seem to think Western economies can dodge a Western recession. This leads me to conclude two things. One, the collapse in metals markets is overdone in the short term (although the CME’s move to raise position limits after the close on Friday is a blowfly in ointment). Two, if a Western recession comes (still my base case) and Chinese growth is dented, the ultimate downside for the broader commodities complex (grains, energy) will make the last couple of days look like a Sunday picnic.

Houses and Holes
Latest posts by Houses and Holes (see all)


  1. There is, of course, another element to the financialisation process; the accompanying transformation of commodity markets themselves into speculator casinos.

    the best example of this was base metals post March 2009. Metals had overshot on the downside when the GFC hit. But then they rallied, and rallied, and rallied. Supply/demand balances never justified the price hikes e.g. we saw stockpiles and prices rising together. Amidst this we had every man and his dog talking up copper and telling us how there was going to be a supply deficit (never happened). This is what happens when you have exchanges dominated by speculators — big swings up and down that are either detached from fundamentals or amplify them.

    On the other hand iron ore and coal are not currently experiencing these swings — they are largely traded by contract between producer and consumer and therefore better reflect the underlying supply/demand balance.

      • I think the you’ve misread Dr Nick’s point. I would point out as well that since the spot market has become more important to iron ore pricing, the price swings have increased.

      • I wasn’t disagreeing with H&H actually. Maybe cause I so often do it is taken as a given 🙂

        re: the spot market. If you contract something daily, say, versus quarterly or annually you’d expect more volatility but since the participants are largely the producers and consumers of the product we’re still not seeing the same volatility as we do on exchanges.

        (exchanges may end up influencing iron ore and coal via what happens to the steel price.)

  2. The difference between housing and commodities is that once you have dug up the copper or whatever and used it in something, it isn’t supply for the next period as well. Supply capacity can be expanded (curve shift out) but it can also fall as mines are closed etc.

    But housing (and even more so commercial property) lasts for decades. The undershoot in prices in a country with “flexible” supply will be greater than if supply was sluggish on the upswing.

    People barracking for us to become Texas should be careful what they wish for.

      • No, the undershoot as in where prices end up after the boom ends. Much much worse if the supply curve shifted out during the boom period.

        Not sure what you means by addicted to asset inflation, but I suspect you are putting words in my mouth.

    • Some huge QE effort won’t help commodities if it’s based in Europe. Actually, I’d argue that will be bad for growth as it deflates the euro. That means a defacto rising $US, choked off export demand for the US etc.

      Only the Fed can revive the monetary inflation of commodities. Which, when things get bad enough to roll over the Congressional nutters, it will most likely do.

      Which is why I expect gold to turn ahead of the other metals.

      • J curve in gold is what I’ve been waiting for the last couple of months. I’d wait and see it bottom and turn before jumping back in though. A lot of negative sentiment about gold across all the trend following blogs. Fundamentals haven’t changed re debt and currency risk and nobody is putting interest rates up. This is not financial advice as I’m sitting on a loss on some gold bought at the top, therefore not as as smart as I though I was.

      • Regarding gold, what you expect is what has happened historically given it’s monetary value. I expect the same to happen this time, but I would not discount further action against gold in the short term. If CB’s keep buying gold should gain more, but I would not bet on any outcome given that you can’t trust the organisations running our monetary system. India the biggest buyer of gold is now starting gold more as a monetary item rather than jewelry so that is a big shift in perception.

  3. Diogenes the Cynic

    Nice article HnH. You have been talking about this potential tipping point for some time so bravo!

    Sell BHP, Rio et al and Australia, (lower exports = recession for Aus) so sell Aus domestic related stocks as well. Dr Nick’s point just means that the reaction for iron ore and coal will be a bit slower than copper and base metals, that is true but stockmarket looks through such things. On a long term basis unless a new source of energy is developed then coal is probably still a reasonable bet.

    Already done, this was a fairly obvious situation developing. The Chinese copper story was a classic alert of speculation just before the peak.

    Invest in US Treasuries…and?

  4. Thanks H&Hs

    Answers many questions I have been thinking about recently.

    Re coal and iron ore; I do not understand why China is paying the prices for these commodities that it has been especially since 2008. Supply and demand arguments are fine but what determines demand? How are the decisions made to create demand and pay in effect whatever is asked? Perhaps this is not speculative demand for the commodity but speculative demand for the asset building?

    The more you dig the more its the same story

    • From the point of view of an iron ore producer demand is from steel mills. Purchasers of contracts are receiving physical iron ore (as opposed to exchanges where speculators can by paper so the demand is actually a demand for paper rather than for the physical commodity).

      So the price reflects actual physical use of the product. I know there has been discussions about increased inventories at Chinese ports etc. but there are limits to how much stuff you can store. The bottom line is that steel production remains strong.

      So from there you can argue that overcapacity in steel represents speculative demand for things that use steel I suppose from which it follows that if steel mills change their collective view about demand for their product we could see a drop in demand for met coal and iron ore. But even then, because these are contracts between producer and consumer we aren’t going to see the wild swings on the downside. For example prices for some base metals in late 08/early 09 were well below cost and therefore unsustainably low (these contracts can be shorted — naked shorting). In a price negotiation that involves a producer and a consumer of the commodity you aren’t going to see that.

    • Some would say it’s all part of a plan. Peg your currency. Generate masses of reserves. Construct for your future. Create the meme. Make the sector dependent on your demand. Pay the market price. Hoard. Stockpile.

      Then…pull the plug. Commodity prices crash. Resource company stock prices crash. Step in and take controlling equity stakes. Future resource requirements secured.

      Now I wouldn’t necessarily say that…but some would.

      • +1 Many do say that. But we humans are very good at prescribing intent to observed events even when there is none.

      • Marc Faber did write a book describing how this would occur. Can’t recall the title but it was published a year or two before the current upswing in commodity prices …~2002/2003?

        He used quaint names for China and the USA but it was clear that he saw China as being the ultimate commodities insider trader.

        I find it hard to get past his Dr Strangelove accent and his lack of understanding of the monetary system but he did seem to nail this one pretty well.

      • If communist bureaucrats were that clever, we wouldn’t have this crisis today. The world would have become communist long time ago.

        Seriously, since the uncovering of massive corruption and fraud in the railways ministry and the tragic fast rail accident, it has become clearer and clearer that the commodities demand in China was driven by something quite unsustainable. While predicting the arrival date of a serious slowdown in China is difficult, the fact that it will come was beyond doubt. As usual, it will always arrive at an inconvenient time.

        Hence my suspicion that our good RBA is keeping rates high to save us for precisely that day. The only unfortunate thing was that the government of the day cannot stomach people like Professor McCubbin, who in my humble opinion is a fine example of the “neither fear no favour” creed.

        But talks of debt meltdown in China will eventually prove to be wrong. The government still owns such much of the prime business assets, they can wipe clean all their debt, plus retaining a nice kitty, if they are willing to pursue privatization. That leads to a whole different category of questions that is outside the interest of the readers of this web site.

  5. Nice article H&H.

    If you (and readers) didn’t already know, there is substantial research about the long run trends in commodity prices relative to prices of manufactured goods. One convincing theory is the Prebisch-Singer Thesis which says that in the long run, commodity prices will fall relative to the prices of manufactured goods.

    This is usually bad news for commodity exporting countries, which have been typically developing nations and Australia. If you don’t diversify your export mix your relative economic position is likely to deteriorate in the long run, but with some massive temporary spikes (which also trigger new supply).

    Taken together with Say’s Law, the P-S Thesis suggests that high commodity prices are, by definition, temporary.

    “Over the last 140 years, real commodity prices have declined by about 1 percent
    per year, but this has not been a smooth process, with prices sometimes changing
    by as much as 50 percent in a single year.”

    • I totally agree Cameron, and one point even though China has agreed to the new pricing mechanism, they are very angry about it. To that end it will remain in place until they can get other suppliers which is long underway.

      If the global economy continues to contract then commodities will also contract. BHP/Rio publically acknowledge this is happening now.

    • Cameron, a lot of what I have read about trends in commodity prices ignores commodity cash costs. Real prices for commodities probably need to be normalized to real extraction costs in order to achieve a better analysis. For example if a model points to a particular long run price which is below current production costs then either production costs have to fall or the market will not be supplied at the forecast price. This is where a lot of analysis broke down prior to the GFC. Pundits commented about commodity prices without considering the exponential growth in costs.

      One convincing theory is the Prebisch-Singer Thesis which says that in the long run, commodity prices will fall relative to the prices of manufactured goods.

      Given that it is commodities that are the inputs I would have thought it should be the other way around: prices of manufactured goods have to rise to accommodate commodity prices (particularly given the high price floor set by rising costs).

      It seems to me in any discussion you have to weigh up who has the bargaining power. It is commodity producers that have finite scarce resources. Extraction costs rise putting a higher and higher floor under prices. Barring a supply glut IMO manufacturers have no choice but to pay up and try to pass on costs.

      • “Given that it is commodities that are the inputs I would have thought it should be the other way around: prices of manufactured goods have to rise to accommodate commodity prices (particularly given the high price floor set by rising costs).”

        In real terms it is the same. If the costs of manufactured goods rise, this is a reduction in the value of the currecy. The thesis concerns just the relative prices. A sustained increase in commodity prices would lead to an even greater increase in manufactured goods prices (and cosumer prices).

        My personal view is that the long-run relative prices between all goods in the economy (commodities, land, manufactured goods etc) are determined by the technology frontier. I come to this conclusion because a high price for manufactured goods necessarily leads to higher costs of producing commodites – one good is not an input to the other, they are both inputs to each other (as in this years manufactured mining equipment is the capital to provide next year’s minerals).

        When looked at prices as a relationship determined by our production technology, we can see that price breakouts, such as high commodity price inflation, cannot persist with low consumer price inflation, since the relationship between these two prices is fairly well set.

        Anyway, more on that another time perhaps.

  6. “if a Western recession comes (still my base case) and Chinese growth is dented, the ultimate downside for the broader commodities complex (grains, energy) will make the last couple of days look like a Sunday picnic.”

    What would happen to our economy which has been held up by the twin pillars: credit/housing bubble and the commodity income bubble? Right now, the deflation of the first bubble is offset by the continuing second bubble. Otherwise, the massive change in savings rate (deflating credit bubble) would have resulted in a deep depression already.

    It makes one wonder whether RBA is deliberately restraining our spending to prevent a fall from even greater heights, and saving the bullets for the terrible days ahead?

    The RBA does not deserve the vitriol directed at it. It was mostly the Treasury that did the damage. Right back in 2002/2003, it was the Treasury that insisted on deeper rate cuts and ignited the housing bubble. The Treasury came up with the crazy housing bubble reflation plan during the GFC, too.

  7. The Supply/Demand graphic in a discussion about commodities crashing. Hilarious. Where’s the influence of half a years supply on the short side on your graphic? What about warehousing output limits?
    Was that just a comedy sideline at the end?

    What about the massive volume increase we saw last week? How do you equilibrate that to such sudden changes in supply and demand? Or was there not really sudden changes in supply and demand and it was all paper?

    Isn’t it really true that the price of commodities is driven more by Wall St than by miners and consumers?

    How long can you hold on to an obvious fallacy: “the following chart is the most important in contemporary commodity markets”

  8. The ghost of Julian Simon will be smiling….

    This is just “commodities” reverting to their trend line, that is all.