Always Cry Over Spilt Milk

Includes: SPY
by: Jeremy Grantham

Part I

(An Admission Of A Past Mistake On Resources)


Six months ago, I wrote a 30-page, 24-exhibit paper, which I finally realized would have a strong appeal to six or seven strategists and no one else. So, here is the summary - written last year, but postponed to run after other more timely subjects - which is long enough for admitting errors.


  • The only thing that really matters in asset allocation is sidestepping some of the pain when the rare, great bubbles break. At other times, traditional diversification will usually be good enough.
  • The defining events at GMO were avoiding some or much of the pain associated with the three great bubbles of the last 100 years: the U.S. housing and finance bubble of 2008, the U.S. tech bubble of 2000, and the Japanese equity bubble of 1989. Exhibit 1 shows the last of these: the best example of an extreme outlier that nevertheless turned out to be an epic bubble rather than a paradigm shift.

    As of 9/30/12
    Source: Datastream, MSCI, GMO
  • Giant bubbles are easy to spot statistically, but hard to call from a career risk perspective. It is easy to be early, and being early may lose you your job, your clients, and your credibility.
  • The fear is always, "Are these new high prices permanent? Is it a paradigm shift?"
  • Every major bull event is called a paradigm shift, but they almost never exist. Almost never. But not never, ever.
  • In 1999, we presented 28 major bubbles of the past and were able to call the score: Mean Reversion, 28; Paradigm Shift, Nil!
  • To rub it in, near the 2000 peak I challenged audiences - perhaps 2000 professionals in total - to find a bubble that had not fully broken. There was not a single offer.
  • I appreciated the irony, therefore, when in 2005, I offered oil as the first important paradigm shift.
  • Oil, by 1979, had spiked to over $100 a barrel in today's currency. Against the previous 100-year trend of $16 that was over a 1 in 1 billion long shot (see Exhibit 2).

    As of 9/30/15
    Source: GMO, Global Financial Data, EIA
  • I said in my 2011 paper on resources: "When I see odds of over 1 in 1 million," - I inadvertently understated - "I don't believe the data unless it is ours, and then I don't believe the trend."
  • This 1 in 1 billion 1979 event was caused by the newly effective cartel, OPEC. The response of oil prices could not have been caused by any mixture of pre-existing factors. It was therefore a paradigm shift. But cartels can end. This produces a special case of paradigm shift: likely to last as long as the new factor, in this case OPEC, stays effective.
  • After 1999, the costs of finding new oil started to rise very rapidly. By today, in my opinion, the price needed to support the development of new oil reserves has risen to at least $65 a barrel. This rise from $16 (in today's currency) constitutes a second paradigm shift in oil. Less is found each year in smaller fields, and is more difficult and costly to extract.
  • If we were running out of low-cost oil, I asked myself in 2011, why should we not run out of other finite resources? That everything finite runs out is known by everyone except madmen and economists (said Kenneth Boulding). Also, China uses a much larger fraction of total global minerals than it does of oil.
  • So, predisposed to see signs of finite resources running out, I saw in 2011 a set of remarkable resource statistics. The most extreme among them, iron ore, was measured as a 1 in 2.2 million possibility. This data made me believe that China's growth, together with increases in world population was causing us to be facing "peak everything."
  • And my report to that effect got many multiples of hits than a typical quarterly letter, so, I must have been right!
  • But alas, I was fooled - along with all of the CEOs of the miners - by China. The four-sigma event in mineral prices did not occur because those resources were running out. Not yet. Although, there are suggestions in the data (see Exhibit 3) that we are running low on some low-cost resources: even in the current truly sensational glut, the 34 equal-weighted index is only two-thirds of the way back to the old trend. I would guess, though, that in explaining the price behavior of commodities for the last 15 years, I would only allocate a weighting of 20% or so to elements of paradigm shift.

    As of 12/31/15
    Source: Global Financial Data, World Bank, GMO calculations.

    Note: A technical improvement has been made to the original index: rice and potash in, pepper out, increasing the number of commodities from 33 to 34.
  • Nor was the bubble in minerals caused predominantly by massive speculation, momentum, and skullduggery, in my opinion; although those components, as was suggested at the time by Frank Veneroso (a financial consultant), had considerably more impact than I had then thought. I would still give these factors only a 20% weight in explaining our current glut and future outlook.
  • No, China was the dominant 60% input: the sheer magnitude and the long 30-year duration of its growth surge; the remarkable late acceleration in growth rate; and, finally, the abrupt cessation of growth. Taken together, these developments constituted the four-sigma event.
  • I had warned of trouble if China slowed its growth for demand for minerals quickly. They did far worse than that. They absolutely stopped growing at all (see Exhibit 4). I completely missed the possibility of such an extreme outcome.

    Source: Worldsteel, China Customs
    Steel production is used as a proxy for iron consumption. For the China figures, steel exports are subtracted in order to reflect only iron used to manufacture steel for the domestic market.
  • By the time China's growth in demand for metals (and coal) stopped dead, the miners had spent an unprecedented $1.25 trillion in expansion to keep up, helped in their efforts by several earlier years of record profits.
  • Due to long time lags (of up to seven years), new capacity will be coming on-line for two or three more years despite the current glut.
  • If China's GDP growth were to average a reasonably strong 4% a year for the next 10 years, its GDP will rise by 48%.
  • If the country simultaneously succeeds in lowering its share of capital spending to GDP to 32% in 10 years from today's 47% - which is its intention - then the need for growth in capital spending-type resources will be about nil (47% x 100 = 47 goes to 32% x 148 = 47).
  • There is, therefore, unlikely to be a quick or dramatic recovery in demand for metals. Stock prices, on the other hand, from these beaten-down levels (of late 2015), are much harder to predict.
  • My suggestion of a positive intermediate-term (10-year) outlook for mining resources therefore seems to have been a major error, especially for iron ore and bauxite.
  • How could I make such an error? I had thought that a 2.2 million to 1 possibility for iron ore must imply a new trend. Although this may usually be a valid assumption for such extreme outliers, it is obviously not universally so. For example, the land under the Emperor's Palace really did have a market value equal to the state of California. Yet, it spent the next 26 years declining, with all Japanese land values, back to normal: a classic looking paradigm shift that, like iron ore, turned out to be a mean-reverting bubble after all. In bubble history, it appears no fixed rule works 100% of the time.
  • I could, and should, have made good money playing against the bubble in minerals, as we did playing against the three other major asset bubbles of the last 20 years. This event in metals indeed turned out to be not the second important paradigm shift, as I thought, but the fourth great bubble of the last 100 years!
  • The best I can say in my defense is that I gave a good warning of the possible risks in my original report: "If China stumbles or if the weather is better than expected, a probability I would put at, say, 80%, then commodity prices will decline a lot. But if both events occur together, it will very probably break the market en masse. Not unlike the financial collapse." And this is indeed what happened. I had thought it probable that short-term prices would decline, possible that both negatives would occur together, but highly unlikely that they would both be as extreme, as turned out to be the case.
  • Food, which also featured in the 2011 report, is quite different. It faces long-term intractable problems that I believe will threaten political stability around the world, deep into the future. (A belief shared, apparently, by the U.S. Intelligence Community.1) Food is in a comparable situation to metals, though, in that it faces a short-term glut. But for very different reasons.
  • Almost four years of terrible growing weather and ensuing shortages and buying panics had pushed grain prices so high that an unprecedented quantity of extra land was brought into use, some of which had not been used for decades, if ever (see Exhibit 5).

    Source: UN Food and Agriculture Organization
  • The weather then changed to be very favorable for grain for almost three years (although, not so for almonds, etc., in California).
  • I had warned that an abrupt change to favorable weather following the increased land use would cause us to "drown in grain, rather than rain." This was a good warning of the risk, but does not mean that I expected such an abrupt shift, only that it was possible.
  • In spite of the current glut, the lack of a safety margin in grain continues, however, as the longer-term negatives remain:

    – Steady declines in productivity gains
    – Water, erosion, and pest problems
    – Increased meat eating (which is grain-intensive)
    – Continued global population growth at over 1.25% a year
  • Oil is indeed the real McCoy, a true paradigm shift. But it, too, faces a short-term glut caused by U.S. fracking.

    – U.S. fracking oil is a small resource, under one and one-half years of global consumption. It will soon run off and show the underlying implacable rising costs of finding ever-diminishing pools of new oil.
    – Existing oil wells deplete faster than they used to, because enhanced technologies squeeze more into the early years. Over 5 million barrels a day out of a global total of 95 a year now needs to be replaced every year. Half a new Saudi Arabia!
    – Today's draconian cutbacks in exploration almost guarantee another sharp price spike in the next two to four years.
    – But beyond a five-year recovery for oil prices and oil stocks, there lurks a third paradigm shift: the terra incognita of electric, self-driving vehicles; cheap electric storage; climate change; and carbon taxes. Taken together, this shift to alternative fuels is likely to cause oil's final paradigm shift!

    Investment advice2
  • Oil stocks should do well over the next five years, perhaps regaining much of their losses. But after 5 years, the prospects are more questionable, and beyond 10 years, much worse.
  • Mineral resource stocks are unlikely to regain their losses, but from current very low prices, they will probably outperform based on historical parallels following similar major crashes. But the next two years will be very risky because of the continued new excess capacity coming on-line, so I strongly recommend averaging in over one and one-half to two years.
  • These mining stocks, though, have several distinct portfolio advantages:

    – They are typically hated, for good reason, and therefore sell more cheaply than the market and have, consequently, a slight long-term performance edge.
    – They are the only group that, over 10-year holding periods, has a negative correlation with the balance of the portfolio (and a very low correlation over 5 years). (See Exhibit 6.)
    – They have proven to beat the market handsomely when unexpected inflation occurs. This is a critical portfolio advantage.

    As of 3/31/16
    Source: S&P, MSCI, CRSP, GMO
  • Farmland is likely to outperform most other assets. It is still my first choice for long-term investing.
  • Forestry should perform above aggregate portfolio averages and be less volatile than equities.

    Lessons learned
  1. There are very, very few paradigm shifts compared to natural investment bubbles that fully break. I, more than most investors, should have known. Even some of the rare, extraordinary outliers can prove to be mean-reverting bubbles eventually.

  2. If you are going to make errors, at least cover your tail! I did give a clear warning of the risks.

    Conclusion, or, sticking it up the reader's nose
  • I do not believe that the current distress in commodities is caused by the workings of some super-cycle, some decadal cosmic force, as some observers have been tempted to say.
  • I believe we are seeing three quite different and coincidental forces in three distinct commodity groups.

    1. Grains, dragged down by a vigorous response to prior bad weather and panic scarcity that resulted in 9% more arable land being planted, coupled with a change to favorable weather for three seasons.

    2. Metals and coal, buried by an unparalleled, abrupt change in Chinese demand from double-digit to negative, from a base that was 25-50% of global demand for each material.

    3. Oil, glutted by the surge of U.S. fracking and the strange gamble by Saudi Arabia.3 Each of these three price collapses would have occurred without the other two. (The usual interactions, such as oil prices or the cost of farming and mining, were very minor in the context of the major influences mentioned.)

Part II: Updates

Equity Markets

The tone of the market commentators back in January, when I was writing my last quarterly letter, seemed much too pessimistic on global stock markets, particularly the U.S. market, and I said so. This relative optimism was an unusual position for me and the snap-back in these markets has validated, to a modest degree, my thinking at the time. I still believe the following: 1) that we did not then, and do not today, have the necessary conditions to say that today's world has a bubble in any of the most important asset classes; 2) that we are unlikely, given the beliefs and practices of the U.S. Fed, to end this cycle without a bubble in the U.S. equity market or, perish the thought, in a repeat of the U.S. housing bubble; 3) the threshold for a bubble level for the U.S. market is about 2300 on the S&P 500, about 10% above current levels, and would normally require a substantially more bullish tone on the part of both individual and institutional investors; 4) it continues to seem unlikely to me that this current equity cycle will top out before the election, and perhaps it will last considerably longer; and 5) the U.S. housing market, although well below 2006 highs, is nonetheless approaching a one and one-half sigma level based on its previous history. Given the intensity of the pain we felt so recently, we might expect that such a bubble would be psychologically impossible, but the data in Exhibit 1 speaks for itself. This is a classic echo bubble, i.e., driven partly by the feeling that the substantially higher prices in 2006 (with its three-sigma bubble) somehow justify today's merely one and one-half sigma prices. Prices have been rising rapidly recently, and at this rate, will reach one and three-quarters sigma this summer. Thus, unlikely as it may sound, in 12-24 months, U.S. house prices - much more dangerous than inflated stock prices, in my opinion - might beat the U.S. equity market in the race to cause the next financial crisis.

As of 4/30/16
Source: National Association of Realtors, U.S. Census Bureau, GMO

In the meantime, however, we continue to have the typical equity overpricing that has characterized the great majority of time since the Greenspan regime introduced the policy of generally pushing down on short-term rates. At current prices, it is very close to impossible for the mass of pension funds or other institutions to realize longer-term targets of 5% a year, let alone 7%. A 60% stock/40% bond portfolio would be lucky to deliver 3%, even if we were to lock in current high P/Es and above-normal margins. Prudent managers will just have to grin and bear it. The worst argument is always that extra risk has to be taken because the need to deliver higher returns is desperate. The market does not care what your targets are! And in any case, there is a very wobbly relationship between risk and return, as the 50-year underperformance of high-risk equities to lower-risk equities attests. (And this underperformance applies to a varied definition of risk: Table 1 shows the gap between the top and bottom 20% in the U.S. equity market for volatility, beta, and fundamental quality - stability of high return and low debt.) My point over the last two years has been to emphasize how long and painful the grinning and bearing can be, and I firmly believe investors should be prepared for considerably more pain (of overpriced assets becoming yet more overpriced) without throwing in the towel 1999-style at the worst time possible.

As of 05/04/16
Source: Beta returns were generated from Fama/French portfolios. Quality/junk portfolios and volatility portfolios were generated using GMO's methodology. All portfolios are market cap-weighted.


As with stock prices, I am encouraged by the rise in oil prices since my unusual bullishness of last quarter. My belief remains that a multi-year clearing price for oil would be the cost of finding a material amount of new oil. This appears to be about $65 a barrel today, and costs are drifting steadily higher as the cheapest old oil is pumped. My guess is that the price of oil will indeed be as high as $100 a barrel again within five years, and perversely, I feel encouraged by the growing host of longer-term pessimists. Much as I would love, as an environmentalist, to have oil for transportation almost disappear in 10 years, it simply isn't going to happen. However, there will be dramatic technological improvements in non-oil based transportation well before 10 years is out, and after 10 years... well, the oil companies will wish they had taken Ted Levitt's advice in 1960 in his then groundbreaking "Marketing Myopia" article and defined themselves as energy and chemical companies and not defended their narrower definition of "oil companies."

New technologies for energy and transportation

The new discoveries and engineering insights in these fields keep coming. The Grantham Foundation's attraction to venture capital, described last quarter, plays conveniently into this, and we are investing 20% of our total capital in "mission-driven" projects, mainly in venture capital. One investment is aiming at doubling the power-to-weight ratio of lithium ion batteries, and another at a 200-second charge time for some types of lithium ion. Vehicle development we see includes ultra-light and ultrastreamlined electric-powered people movers, suitable for commuting and shopping in developed countries, and as a first cheap, yet state-of-the-art, vehicle everywhere, very fast and with a long range. (Better make hay soon, Tesla (NASDAQ:TSLA)!) But, you never know. And venture investing is particularly full of disappointments. So, just in case, I am one of the 400,000 preorders for the $35,000 Tesla Model 3, likely to be the cheapest electric car per mile of range at least for a while. The mission component on some of our investments is so on target that even if they fail, they represent attempts that will have deserved some of our non-profit making grants. As it is, we hope that at least one of the several game-changing projects driving these CO2-reducing technologies forward will simultaneously make our foundation a fortune!

Global warming accelerates: "So much for the pause"

Because 1998 was an outlier warm year due to a large El Niño effect in the Pacific, many subsequent years, including 2013, had lower global temperatures and led some to believe, or claim to believe, that global warming had ceased. But it turned out to be, after all, just another series like that of the S&P 500 in real terms, with a little steady signal often obscured by a very great deal of noise. As it turned out, the below-trend 2013 was followed immediately by a modest new record in 2014. And then came the real test as a new, powerful El Niño started to build up in 2015. Ten of the twelve months of 2015 set new all-time records, an unheard of event, and 2015 in total became a monster - not only the warmest year in recent millennia, but by a record increment. Yet, the early months of 2016, still under the influence of what had become one of the most powerful El Niño effects, showed temperature increases that were even more remarkable.

This current El Niño has accelerated underlying warming caused by increasing CO2 - as all El Niños do - but this time, the combined effect has been far ahead of scientific forecasts that, in general, remain dangerously conservative. January 2016 was the hottest January ever on the NASA series, and by a new record amount. It was a full 0.22 degrees Celsius above the previous high for January. Then, February became the new shocker, washing away that record by being 0.33 degrees Celsius above the previous February record. Most recently, March was once again the warmest ever March, although not quite by a record amount (see Exhibit 2). The exhibit makes the scary point clear: global temperature is not just increasing, but accelerating. The average increase from 1900 to 1958 was about 0.007 degrees Celsius per year. From 1958 to 2015, it doubled to 0.015 degrees Celsius per year, and from February 1998 to February 2016, it rose by an average of 0.025 degrees Celsius per year! Time is truly running out.

As of 4/30/16
Source: NASA Goddard Institute for Space Studies

Sadly, it has become obvious that the recent talk in Paris of limiting warming to 1.5 degrees Celsius is toast, as it were. And the dreaded 2 degrees Celsius is highly unlikely to be the limit of our warming. If you line up the previous El Niño outlier of 1998 with this March 2016 El Niño (as we might do in lining up bull market highs), it gives an idea of when 2 degrees Celsius might first be broached in a future El Niño effect: just 17 years! Meanwhile, the most obvious effect to watch for in destabilizing weather patterns is an increase in record-breaking, intense rainfall, such as occurred last month in Houston. Three inches an hour4 fell, and kept falling hour after hour, delivering four months' average rain in under 24 hours (unprecedented without a major hurricane), flooding major parts of the city under several feet of water.

Let me just make the point here that those who still think climate problems are off topic and not a major economic and financial issue are dead wrong. Dealing with the increasing damage from climate extremes and, just as important, the growing economic potential in activities to overcome, it will increasingly dominate entrepreneurial efforts in future decades. As investors, we should try to be prepared for this.


1 Office of the Director of National Intelligence, "Global Food Security Assessment," October 2015.

2 As always, my personal opinions.

3 See "The Oil Glut, Saudi Decisions, and the Uniqueness of U.S. Fracking," GMO Quarterly Letter, 4Q 2014.

4 One inch per hour is usually considered a "deluge."

Disclaimer: The views expressed are the views of Jeremy Grantham through the period ending May 2016, and are subject to change at any time based on market and other conditions. This is not an offer or solicitation for the purchase or sale of any security and should not be construed as such. References to specific securities and issuers are for illustrative purposes only and are not intended to be, and should not be interpreted as, recommendations to purchase or sell such securities.