How well are long-run commodity price series characterized by trend components?

Abstract

Various studies have investigated trends in commodity prices in the context of the Prebisch Singer Hypothesis. This paper applies new evidence on significance tests for trends in differenced and correlated stationary processes, to individual price series. It is also investigated how well trends are suited for price forecasts and what these findings imply for developing countries. Few commodity price series are well characterized by a trend and trends are useful for forecasting in even fewer cases. Commodity specialization is still problematic for LDCs but the main issue is a lack of predictability rather than a secular decline. Copyright © 2005 John Wiley & Sons, Ltd.

A reexamination of trends in primary commodity prices

“…In the event where a country experiences both current and capital account deficits, one may observe a decline in international monetary reserves and currency instability. Given that for several commodities (coffee, cocoa, tea, maize, sugar, lamb, jute, hides, silver and zinc) we find no evidence of a trend, one may infer from these results that foreign exchange constraints facing developing countries can be relaxed (Newbold et al, 2005). However, Newbold et al (2005) point out that the risk of revenue shortfall during the period of time when the repayments are anticipated can be quite high.…”

“…Given that for several commodities (coffee, cocoa, tea, maize, sugar, lamb, jute, hides, silver and zinc) we find no evidence of a trend, one may infer from these results that foreign exchange constraints facing developing countries can be relaxed (Newbold et al, 2005). However, Newbold et al (2005) point out that the risk of revenue shortfall during the period of time when the repayments are anticipated can be quite high. Given that we find very little evidence of a persistent long run trend, one cannot draw conclusions regarding future terms of trade.…”

“…The original work on the Prebisch-Singer hypothesis has assumed that the underlying data series is trend stationary (Prebisch (1950), Singer (1950)). While Sapsford (1985), Grilli and Yang (1988), Helg (1991) and Ardeni and Wright (1992) among others have advocated for commodity prices to follow a trend stationary model, Cuddington and Urzua (1989), Cuddington (1992), Bleaney and Greenaway (1993) and Newbold et al (2005) recognized that commodity prices may be difference stationary. The evidence on the Prebisch-Singer hypothesis has been mixed.…”

Underlying Trends and International Price Transmission of Agricultural Commodities

“…In the event where a country experiences both current and capital account deficits, one may observe a decline in international monetary reserves and currency instability. Given that for several commodities (palm oil, coconut oil, tea, and beef) we find no evidence of a trend in recent years, one may infer from these results that foreign exchange constraints facing developing countries can be relaxed (Newbold et al 2005). However, Newbold et al point out that the risk of revenue shortfall during the period of time when the repayments are anticipated can be quite high.…”

“…Past studies that have analyzed trends assumed that the underlying data series is trend-stationary. While Sapsford (1985), Grilli and Yang (1988), Helg (1991), and Ardeni and Wright (1992) among others have advocated for commodity prices to follow a trend stationary model, Cuddington and Urzua (1989), Cuddington (1992), Bleaney and Greenaway (1993), and Newbold et al (2005) recognized that commodity prices may be difference-stationary. The evidence on the trend function has been mixed.…”