Meet the Author

Andrea Pescatori |

Economist

Andrea Pescatori

Andrea Pescatori is a former research economist in the Research Department of the Federal Reserve Bank of Cleveland.

Meet the Author

Beth Mowry |

Research Assistant

Beth Mowry

Beth Mowry was formerly a research assistant in the Research Department of the Federal Reserve Bank of Cleveland. Her work focuses on labor markets and business cycles.

11.28.08

Economic Trends

Industrial Production, Commodity Prices, and the Baltic Dry Index

Beth Mowry and Andrea Pescatori

Industrial production rebounded in October, rising 1.26 percent after declining a downwardly revised 3.7 percent in September. The revision to September output was caused, in part, by a larger–than–anticipated estimate of the impact of hurricanes Gustav and Ike on the chemical industry. The September drop resulted in the largest month–over–month percent decline in the series since February 1946.

Splitting the series into its cyclical and trend components, we notice two things about the behavior of industrial production (IP). First, the trough currently observed in the cyclical component is of a similar magnitude to that of the 1973–74 recession. Second, since the 1970s, the trend in IP has been lagging the dates of recessions: The trough in the trend is reached at the very end of recessions. (A caveat: The filter used to construct the trend might have introduced an artificial phase shift on the order of 2–4 months for the most recent data.)

Industrial production reached its peak in January 2008, with an index reading of 112.6. It has since fallen about 4.7 percent to 107.3. It is worth noting that the 1974 and 2001 recessions were each preceded by an exceptional increase in IP and then followed by a trough. The current situation seems more similar to the 1980-81 scenario, when IP did not surge before turning downward.

Industrial production can also be divided by major market and industry groups. Market groups include, for example, final products, nonindustrial supplies, and materials. Construction lies within the nonindustrial supplies group. It is interesting to note that, while more volatile, the IP–construction series was well-synchronized with the total IP series up until recently. This seems to have changed now, as the trend in construction production is leading the decline in total IP, confirming the fact that the current decline in economic conditions started in the housing sector.

The manufacturing sector is the most important component of industrial production in most countries, emerging countries included. Its definition excludes construction and comprises “establishments engaged in the mechanical, physical, or chemical transformation of materials, substances, or components into new products.” The inputs used and transformed by manufacturing establishments are raw materials that are products of agriculture, forestry, fishing, mining, or quarrying as well as products of other manufacturing establishments. Those materials are usually purchased directly from producers or obtained through customary trade channels. Consequently, most of the commodities traded throughout the world are (directly or indirectly) the main input materials of the manufacturing sector.

The fact that those commodities have a worldwide market means that their prices are a good barometer for economic activity around the world, and, when those prices are available at a high frequency, they can be used as an indicator of future economic conditions. Recent data on commodity prices confirm that the slowdown in industrial production is a worldwide phenomenon.

The overall spot commodity price index (published by the Commodity Research Bureau) peaked between May and July 2008 and has greatly retreated since then. That index is clearly strongly affected by oil prices, which peaked in the first half of July. West Texas Intermediate (WTI) oil sold for more than $145 per barrel until July 14. However, other commodities peaked earlier than that. For example, metals peaked around April and May 2008, textiles and fibers peaked in March (although the steep decline in this index did not begin until after July), raw industrials peaked between March and May, and foodstuffs peaked in July. The industrial and precious metals spot indexes (published by Standard and Poor’s) peaked in March, while wheat and corn peaked in March and June, respectively.

Another indicator that is supposed to be a relatively accurate barometer of global trade volume and, in turn, global production, is the Baltic dry index. It is issued daily by the London–based Baltic exchange and is considered useful in part because it contains no speculative content. World economic activity is the most important determinant of the demand for transport service, and the Baltic dry index, loosely speaking, provides an assessment of the price of moving major raw materials by sea.

This “price” reflects the demand for shipping capacity with respect to the supply of dry bulk carriers, which is inelastic in the short run. Hence, the index indirectly measures global demand for the commodities shipped aboard dry bulk carriers, such as building materials, coal, crude oil, metallic ores, and grains. These materials function as raw material inputs to the production of intermediate or finished goods, such as concrete, electricity, steel, and food, which makes the index an economic indicator of future manufacturing activity and, more generally, worldwide industrial output.

The Baltic index, after skyrocketing to almost 12,000 in mid–November 2008, now sits around 840 (about a 93 percent drop!). Part of the run-up reflected oil–price patterns, given that bunker fuel is a significant part of shipping costs. However, bunker fuel prices can explain only a small fraction of the Baltic index’s volatility. Moreover, WTI oil prices were still above $146 on July 14, while the index was already receding from its peak. In fact, it peaked back in May (on a daily basis it peaked June 5 at 11,689) and by the beginning of July was already below 9000 (a 25 percent decline).

All in all, data on commodity prices and freight rates suggest that world industrial production (or economic activity) was still exceptionally buoyant during the winter of 2007. Furthermore, world production was not perfectly synchronized with U.S. industrial production, which saw its turning point in January 2008. This created a situation resembling the 1970s, where producer prices were skyrocketing due to increasing commodity prices, while industrial production was stagnating. However, unlike the 1970s, the initial shock behind the current slowdown (stemming from the financial industry and the housing market) has restrained the ability to lend, an effect more similar to the one caused by the disinflation shock engineered by Fed Chairman Volcker in the early 1980s. This effect, coupled with a better management of inflation expectations by central banks, has avoided a persistent world rise in inflation together with the downturn.

Hence, the lack of perfect business cycle synchronization, especially between developed and emerging economies (which, like China, now represent a large share of the world manufacturing sector), contributed to the spectacular rise in commodity prices up until the summer of 2008. However, this clearly could not last for long. The spillover from credit markets and the drop in U.S. imports had already hit world industrial production by the spring of 2008 (as reflected in some commodity prices and especially in freight rates). This also suggests that the high oil prices were not the outcome of pure speculation but were reflecting demand pressure originating in the manufacturing sector; at most, we can say that it took it a little bit too long before oil prices started to decrease.