Bill Miller Letters Archive

2006 Q1 – Market Commentary

When I saw the front page headline in the Financial Times on April 10, I was momentarily disoriented. “Commodity prices set to soar,” it read. My first thought was I read the date wrong, that it was April 1, and this was an April fool’s story. But it was April 10, not April the first.

Then I thought perhaps they had delivered me a copy of the FT from the spring of 1999, when commodities prices were all in the tank and oil sold for $10 a barrel. Reminding the reader that you had predicted the great commodity bull market in 1999 when prices were down and everyone was bearish would be good marketing. But the date was unmistakably 2006. And then I recalled that the FT’s sister publication, The Economist, had on its cover in the spring of 1999 the headline “Drowning in oil,” saying that despite the price of crude having fallen in half in two years, “$10 might actually be too optimistic. We may be heading for $5.” What it did not say, as oil was making its lows, was that the price was set to soar, and that you could make six times your money in the next seven years.

Is it any surprise now that oil is a six bagger, that copper has quadrupled in the past four years, that silver has tripled in three years, as has sugar, that orange juice has doubled in the past two years, and that after the biggest commodity rally in 50 years, it is NOW that prices are “set to soar” and that pension funds are falling all over themselves to allocate a portion of their assets to commodities? From the same FT:

Last month, J Sainsbury, the U.K. supermarket chain, said it planned to invest about 5% of its 3.2 billion pound ($5.5 bn, euro 4.6 bn) fund in commodities. . .

The article went on to note how this move is being driven by pension consultants armed with data showing how well commodities have performed, how over 50 years they have performed about as well as equities, and how this was part of a move to “find surer ways” to fund long-term obligations than by investing in main- stream equity indices, those surer ways consisting of “alternative assets” such as private equity, hedge funds, commodities, and so on.

These alternative assets are judged surer mostly because they have done so well the past five years, just as venture capital was so judged in 1999 and 2000 because of its stellar past returns, leading to a flood of money and a collapse in returns. (I am leaving out of this any discussion of the dispersion of returns; suffice it to say they are wide and not everyone will get the top quartile re- turns, though all will so believe a priority.)

This brought back memories of the spring of 2000, when on March 31 The New York Times business section headlined “The End of the Game,” about how Julian Robertson was closing his legendary hedge fund because the value investing game he practiced, of buying the cheapest companies and shorting the most ex- pensive, was over. Many well-known value investors such as Bob Sanborn of Harris and George Vander-Heiden at Fidelity, in addition to Robertson, had left their firms as a result of not being in the tech and tele- com game. That game having run five years was deemed permanent and a new paradigm. Of course, it was that game that was ending, and those value inves- tors who survived and who practiced a deep value strategy (which often involves commodities or cyclicals) were about to enter one of their best periods in market history.

It is not an accident that despite the consultants being armed with data going back 50 years about how adding commodities to a portfolio can improve risk-adjusted rates of return, there was zero interest among pension plans and investors generally in owning them until very recently. “Two years ago it would have been hard to find any funds investing in commodities,” says Stephen Woodcock, investment consultant.

Suffice it to say it wasn’t a steely eyed look at the data sets and time series that has galvanized the investment world into believing commodity prices are “set to soar,” and that they better move fast to get in before the big move. Part of the problem with the data is what the professors call sensitive dependence on initial conditions: it matters when you start the measurement period. If you start it in 1950 and end in 2000, it shows about 47 years of oscillation and three years straight up in the early to mid 1970’s. Since the last big move in the 70’s, for a couple of decades commodities did nothing, so despite all the data about noncorrelation with equities, nobody cared.

When prices were down, people could cite good fundamental reasons why they shouldn’t go up, just as they can now trot out all the reasons for why they should continue to rise: insatiable demand from China, India and the developing world, 20 years of underinvestment in production capabilities, sources of supply are either depleted or in geopolitically unstable parts of the world, skepticism about the sustainability of the price move, greater capital discipline on the part of the companies who produce commodities, massive new investment demand which is permanent since it is driven by indexation and thus will not leave even if prices decline (want to bet on that last one?). There are more but you get the picture.

The great commodities bull market in the 1970’s had many of the same features as this one, except that most of what drives this story is demand, while that one was driven by scarcity of supply, either due to the OPEC cartel, or due to Club of Rome type analyses about running out of all the major commodities. (If you don’t remember the Club of Rome you are of a different generation.) I recall one prominent forecast around 1979 that oil would go up 15% per year “forever.”

A feature of the data showing the performance of commodities over 50 years that appears lost on those now allocating assets illustrates the distinction made by Stephen Jay Gould in his book Full House between conflating the trends of a system with the trends within a system. From the early 1950’s to the early 1970’s commodities prices oscillated, and were lower in 1972 in nominal terms than they were in 1950. Prices then tripled over the next three years, then declined for a couple of years, then spiked up in 1979 and 1980, and traded sideways for nearly twenty years before declining into the tech and telecom boom. They have subsequently soared over the past 3 or 4 years.

The 50-year trend OF the system is indeed up at about the rate of equities—after this big move we have just experienced— with minimal correlation, but that obscures the trends IN the system, which show 20+ years of returns that oscillate around zero (depending on the index and the measurement period there are modest nominal returns that offer about a zero real rate), followed by very sharp moves up to a new equilibrium level that is about 3x the old level, then a few decades of stasis.

Since the rally we are experiencing is already bigger and longer lasting than the one that kicked off the 70’s, it takes a deter- mined optimist to say that now is time to be putting money in commodities.

The reason to own commodities may be that one believes they provide equity like returns with little correlation with equities. The time to own commodities is (or at least has been) when they are down, when everybody has lost money in them, and when they trade below the cost of production. That time is not now. The data showing the returns of commodities will look very different if you start measuring just after prices have tripled.

Every commodity we can get data on trades significantly above both the average and the marginal cost of production. Copper, for example, has an average cost of production of around 90

cents per pound, and a marginal cost of about $1.30 per pound. The marginal cost should approximate the equilibrium price over time. The current price is around $3.25 per pound. It is not a question of if copper prices are going down, it is a question of when.

The best argument that it’s different this time and we are in a new paradigm and you can therefore actually earn an excess return by buying commodities now is that demand will grow faster than supply on average for many years (or even for- ever!), due to all the demand driven/supply constrained arguments noted above. This demand driven cycle will lead to price increases that exceed nominal GDP, and therefore corporate profits on average.

If true, it will be the first time in economic history, since historically commodity prices have tended to fall in real terms, due to the improvements in technology for producing the items outpacing the growth in demand. But that doesn’t mean it won’t happen. Lots of things happen that never happened before.

The second best argument is that due to the amount of demand, both industrial and investment, the amount of new supply that needs to be put in place is so large that prices need to be this high to induce the amount of investment necessary to increase supply sufficiently to meet it. There have to be big profits, not small ones, and that means prices way above the marginal cost for a lot longer than we are used to seeing. That argument might be right, but all it says is prices look high, and are high, and they will stay this way longer than normal before you lose your money. In fact, given the negative roll caused by the contango in most commodities markets, investors in futures are already losing money if they are buying the front month and staying fully invested.

I am more sympathetic to the second argument than the first. It is undeniable that the demand picture is different this time. But the prices are different, too. What the commodity bulls implicitly assume is that higher prices will not curtail demand, that price elasticity is less than 1 indefinitely. They also assume that the global tightening that is underway will not affect prices, or that it will stop before that happens. I see little reason to adopt either assumption.

The markets say I am wrong, or at least the newspapers do, as do the current money flows and the advice of the consult- ants. Both the weekend (April 15) editions of The Wall Street Journal and the FT have articles about how to get in on the commodity boom, (“Going for the Gold” WSJ), and offering predictions about gains to be made (“Next goal for gold is $850 an ounce.”) Morgan Stanley has concluded that individuals should now have 4% in commodities, “up from 2% last year,” says the Journal. Not to be outdone, Citigroup is advising a 5% to 10% allocation. Even so sage a market observer as Peter Bernstein, writing in the April 15 FT, says that “commodity prices are likely to keep rising as a result of supply constraints for at least three or four years and possibly longer than that. There will be fluctuations, but the price trend will be up.”

Investing is all about probabilities, and just because there ap- pears to be a strong consensus prices are going to keep going up, doesn’t mean that is wrong, or right. The consensus does tend to be wrong at the turning points, being invariably bullish at the top and bearish at the bottom. Remember all the advice to go to cash AFTER the 1987 Crash, since it was clear a depression would follow. Or how “risky” the high yield bond market was in the summer of 2002 AFTER the Enron and Worldcom collapses led to record spreads? I can’t help but be skeptical of the advice to start or increase a position in commodities AFTER the biggest bull move in 50 years.

Fifteen Fed tightenings have taken us from negative real short rates to real rates headed toward 3% in the U.S. Japan is moving away from its zero interest rate policy, and global central banks are mopping up the excess liquidity created when deflation fears were rampant. The ten year bond is now over 5% for the first time in 4 years, and housing data is rolling over, which is likely to put some pressure on consumer spending, as well. The so-called global carry trade is fading in popularity, and some of its biggest beneficiaries such as New Zealand and Iceland have seen their currencies break down. Other areas where speculation ran high until recently, such as the Middle Eastern markets, have experienced dramatic corrections.

A big difference between today and the commodity bull markets of the 1970’s is that then the US Fed monetized the rise in prices leading to persistent inflation. Today central banks are withdrawing liquidity, not adding it.

So far there has been no impact on commodities, and except for a few scattered areas of the world, most assets and markets are continuing to rise. I think that will get more difficult to sustain, especially if liquidity becomes scarcer. The first to feel the pain could be where the gains have been the greatest, which would be emerging markets and commodities.

The US equity market has lagged those of the rest of the world by a wide margin for several years, and within our market the mega cap S&P names have lagged the small and mid caps, which are in the 7th year of relative outperformance, quite long in the tooth by historic standards. Part of the reason for the relative lack of interest in US stocks has been the relentless rise in short rates. Our central bank has been noticeably more hawkish than the rest of world, and money has flowed to where money was the easiest, outside the US. As we end our tightening cycle, and others remain engaged in theirs, our mar- ket should become relatively more attractive.

In a world where global liquidity may be diminishing, relatively illiquid assets are likely to begin to lose their allure. Liquidity will become more valuable. I think the most liquid market in the world, the US market, will become more attractive, and within that market, money will flow to the largest, most liquid names, which also happen to be the cheapest part of our market.

The excitement and enthusiasm surrounding commodities, and the belief that they will continue to rise, is not surprising. People want to buy today what they should have bought 5 or 6 years ago; call it the 5 year psychological cycle.

Today people want commodities, emerging market, non US assets, and small and mid-cap stocks. Those were all cheap 5 years ago and had you bought them then you would be sit- ting on enormous gains. But 5 or 6 years ago, everyone wanted tech and internet and telecom stocks, and venture capital and US mega caps. The time to buy them was in 1994 or 1995, when they were cheap. But in 1994 or 1995, people wanted banks and small and mid-caps, which should have been bought in 1990, and well, you get the picture.

In general, you can get a good sense of what to buy now by looking to see what the worst performing assets or groups were over the past five or six years. That is long term for most people, and long enough to convince them that the malaise is permanent and to have migrated their money else- where, such as to whatever has done best in the past 5 or 6 years.

Given the choice of buying Commodities with a capital C, or buying capital C—Citigroup—at current prices, I’ll take the latter. Check back in 5 years.

Bill Miller
April 24, 2006