Commodity Prices and Unit Root Tests

Abstract

Price theory suggests that commodity prices should be stationary series. Yet, tests for unit roots rather frequently imply that these prices are not stationary. This seeming inconsistency is investigated by applying alternative specifications of unit root tests to prices of corn, soybeans, barrows and gilts, and milk. The preponderance of evidence suggests that nominal prices do not have unit roots, but the results are sensitive to the specification of the test equation. Accounting for a structural change that shifts the mean appears to be an important issue in unit root tests. Copyright 2007, Oxford University Press.

Sharp Breaks or Smooth Shifts? an Investigation of the Evolution of Primary Commodity Prices

“…Recent examples of work in this area include Kellard and Wohar (2006), Balagtas andHolt (2009), andHarvey et al (2010). Others such as Wang and Tomek (2007) have examined the issue conceptually and empirically; they ar-gue that, due to storability and biological production constraints, prices for agricultural commodities should not follow a unit root process. They also discuss specification issues in testing for unit roots in commodity prices and present empirical evidence that several common price series appear to be trend stationary once mean breaks are accounted for.…”

“…As Wang and Tomek (2007) observe, the production processes for agricultural commodities in the presence of costly storage means that price changes are serially correlated. The autocorrelation is a result of: (1) the biological constraints that prevent an instantaneous supply response to changes in market conditions; and (2) the changes in inventories that help smooth consumption and hence price changes over time.…”

Price Discovery in Thinly Traded Futures Markets: How Thin is Too Thin?

“…As shown by Cavanagh et al (1995) and Elliott (1998), common unit root tests have low statistical power when the roots are near but not exactly one. Moreover, Wang and Tomek (2007) demonstrate that the outcome of unit root tests applied to commodity spot prices is often sensitive to the specification of the test equation. 3 As a robustness check we also conduct the Engle and Granger (1987) procedure and the Johansen (1991Johansen ( , 1995 trace test in our cointegration analysis.…”

Fiscal sustainability in the presence of structural breaks: Does overconfidence on resource exports hurt government’s ability to finance debt? Evidence from Nigeria

“…One obvious advantage of this method is that it does not require pre-testing the integration order of the variables. This eliminates the problem of low power associated with conventional unit root tests such as, for example, ADF and PP tests (Wang & Tomek, 2007). Thus, the approach is applicable regardless of whether the regressors in Equation (6) are purely I(0), purely I(1), or are mutually cointegrated.…”