How did Cinderella feel when midnight struck? It is a question that many commodity investors could probably answer, having endured a torrid time during the past half-decade. In the nine years after China acceded to the World Trade Organization (2002-2010), the asset class was the belle of the ball, delivering annual compounded returns of roughly 10% despite the global financial crisis. Since the beginning of 2011, however, commodities, as measured by the Thomson Reuters/Jefferies (TR/J) CRB Index, have declined by 12.5% per year, leading investors to wonder if Prince Charming will ever come calling to save them.
Commodities, of course, are not a monolithic asset class that moves together like a marching band. There are idiosyncratic supply and demand factors that shape the microeconomics of each market, and it is reasonable to expect a degree of differentiation between price trends in wheat and those of copper, for example. That said, it is obvious that some investors do regard commodities as a broad asset class and invest accordingly. The prevalence of commodity index products, medium-term notes and exchange-traded funds are a testament to the interest in commodities as a means of storing or increasing wealth. At the same time, it is clear that macro factors exert influence over commodity prices as a whole; witness the broad-based price declines at the height of the global financial crisis.
Here, we will focus on the macro rather than the micro as it pertains to commodity markets and attempt to distill some insights into what key macro factors are telling us about likely trends in commodity pricing moving forward. For analytical purposes, and to cast the widest net, we will focus on the TR/J CRB Index, which comprises 19 commodities and is less geared to petroleum-based products.
The first port of call is to identify a sustainable relationship between macro factors and trends in commodity prices. Unfortunately, it can be quite difficult to quantify causal relationships for some of them over meaningful periods of time. For example, with inflation, there is a tautological relationship that can be difficult to disentangle: Commodity prices heavily influence headline inflation rates and breakevens, so even if a mathematical study demonstrates a strong linkage between the two, it is far from clear that we can observe the degree to which inflation expectations influence commodity pricing.
With this caveat, we identified a list of market or economic factors to test against commodity price changes: The U.S. Dollar Index, liquidity as measured by real 10-year yields, credit spreads, Chinese demand as measured by import growth and world trade volumes.
We regressed year-over-year changes in these factors with the equivalent change in the CRB index since 2002 to identify those drivers that most heavily influence changes in commodity price trends. This date was chosen to reflect China’s accession to the WTO and with it the explosive growth of the Chinese economy and its voracious appetite for a number of industrial commodities.
Once we can identify which factors are the strongest drivers of commodity pricing, we can then derive a coherent view about the outlook moving forward (see “Commodity drivers,” below).
Encouragingly, all of the coefficients are the correct sign: a higher dollar, real yields, and credit spreads imply lower commodity prices while stronger Chinese imports and global trade volumes forecast higher commodity prices. The DXY, real yields and global trade are all highly significant, Chinese imports are moderately so, and credit spreads contribute little more than statistical noise to the model. As a sense check, we also ran the regression back into the 1990s. The results were broadly similar, both regarding the signs and the significance of the explanatory variables.
While there are clearly other factors that influence broad commodity prices, we can nevertheless distill some key insights from our regression study:
- The U.S. dollar is clearly a significant driver of commodity prices. This is not surprising as commodity prices are denominated and priced in dollars. Nevertheless, it is gratifying to see the so-called invoice currency effect confirmed in our study.
- Commodities are also clearly sensitive to the level of real interest rates. Again, this is not a surprising result; real rates are a useful proxy for global monetary and liquidity conditions, and we would expect easy conditions to generate inflation in a broad swathe of prices, including commodities.
- Global growth trends are also a key determinant of commodity price trends. Intuitively, this makes sense; stronger growth implies greater demand for a range of finished and investment goods, many of which require commodity inputs.
- While our study did identify China as exerting an influence on commodities, the relationship was weaker than we may have expected. To some extent, this may have been a definitional issue, with the Chinese influence manifesting itself in other ways or upon a narrower basket of commodities than those in the TR/J CRB Index.
- Despite the current focus on the credit market exposures of some commodity producers, particularly in the energy sector, we could not identify any meaningful long-term relationship between trends in the credit market and those in commodities.
Armed with this knowledge, what can we say about prices moving forwards, particularly when we account for factors not incorporated in our study? Our model suggests that we need to address the outlook for the U.S. dollar, real rates, global growth and China in formulating a forecast. We will also consider developments in asset flows, the super cycle and the cost of production as well.
The U.S. dollar has already evolved in a way that suggests somewhat more favorable conditions for commodity prices moving forward. After registering its strongest year-over-year growth in 2015 since the 1980s, the greenback has retreated somewhat and now appears more likely to remain around current levels on a trade-weighted basis. That the Federal Reserve has become acutely interested in both global conditions and U.S. dollar strength does not necessarily preclude any further rise in the U.S. dollar; it does suggest, however, that the kind of U.S. dollar gains that propelled commodities into free-fall may be met with a policy response.
Similarly, real rates have also fallen quite sharply since last summer’s market turmoil. While it is unlikely that they will plumb the depths observed at the height of the Fed QE era, by the same token, the extraordinary monetary measures taken elsewhere in the world are exerting a powerful downward influence on global real rates via both low nominal rates and, eventually, higher inflation.
The outlook for China is not encouraging. The economy has clearly entered a period of permanently lower growth, and the authorities continue to explore ways to transition toward a model geared towards domestic consumption rather than investment. The read on this is that we should experience a lower level of secular Chinese demand growth in the future than we’ve observed in the past. There has already been the talk of shuttering production in some sectors with excess capacity like steel (see “Steeling capacity,” below). While this may eventually push global supply/demand figures back toward equilibrium, in the intermediate term, it is difficult to escape the conclusion that Chinese developments are likely to remain bearish for many commodities.
Similarly, the prospects for global growth are not particularly encouraging. Economies in many parts of the world are struggling to gain traction despite accommodative monetary policies, and there are concerns that the cycle in the United States is mature and that a recession is due. The last two Presidents ushered in their terms with recessions; will the next one follow suit? To a degree, of course, this weakness is already in the price of many commodity markets, and the growth rate of trade for some bellwether countries has been negative for some time. One could argue that there is room for an upside surprise further down the line, but at this juncture it doesn’t look imminent.
If we combine these signals into our regression model, we get a somewhat surprising result (see “Time for a rebound?,” below). Indeed, the model projects year-over-year gains of 2%, in contrast to the current decline of 24%. Even if the drivers of commodity prices do not improve from here, historical relationships would suggest a flat-line in commodity prices at the very worst.
What of other factors not captured in the model? Asset flows have been negative for several years. Barclays estimates that the AUM of commodity index, MTN and ETF products rose from just over
$50 billion in 2005 to nearly $450 billion in early 2011 before collapsing back to $170 billion early this year. Much of this flow follows price trends and points to a serial correlation of returns. That said, the underlying returns of more sophisticated commodity strategies have also suffered: Carry strategies, for example, have returned very little for six years now. On the other hand, the performance of carry has recently picked up, and the bulk of liquidation flows from the asset class has already occurred. Thus, while asset flows are unlikely to provide near-term support for prices, over the somewhat longer term they may become more positive (see “Carry basket,” below).
What of the commodity super cycle? Although the gains of the last decade were not unprecedented, they nevertheless pushed the bounds of historical extremes. After prior episodes of strong real price appreciation in broad commodity baskets, peaks were followed by lengthy 20-year downturns. It’s difficult to escape the conclusion that the current cycle has indeed peaked, and thus that negative returns may be a feature of the foreseeable future.
On a more intermediate cyclical basis, it looks as if some key industrial commodities are moving slowly towards supply/demand balance. Crude oil is an obvious example, with some commentators suggesting that it could be in balance within a few quarters. More generally, it seems as if many key commodities have reached the point of fundamental stability at low levels.
Finally, it’s worth noting that the low cost of oil has reduced the production cost of many key commodities, and thus the producer breakeven thresholds. In a sense, this is a vicious circle turned virtuous and implies that commodity producers may be more tolerant (if not happier) of low prices than investors whose only concern is capital appreciation (see .
The weakness of the U.S. dollar and a global reduction in real interest rates both suggest an improvement in the underlying macro environment for commodities that has yet to manifest itself adequately in the price of broad commodity indices. Against this weighs the ongoing sluggishness of the global economic cycle and, perhaps more importantly, an evident peak in the secular growth rate of underlying demand led by China. These observations are consistent with the notion that following the large buildup of commodity production capacity, prices reached a long-term cyclical peak five years ago.
The upshot for investors is that there is a strong argument for closing short- or underweight positions in the commodity space given that the fundamental rationale for the weakness of 2014-15 is no longer in play. Agile risk takers may even wish to consider taking tactical long positions particularly in those markets where the microeconomic supply/demand balance is most favorable. Over the longer run, however, the heady gains of 2002-11 are unlikely to repeat for a generation. It appears that Cinderella had better get used to the pumpkin.