Piecewise linear trends and cycles in primary commodity prices

We extend the methodology put forward in Yamada and Yoon (2014) to analyze the trend and cyclical behavior of relative primary commodity prices. These authors propose the use of the so-called L1-filter that renders piecewise linear trends of the underlying data. Our focus on the calibration of such filter and its implications for the empirical analysis of primary commodity prices, especially the interpretation given to the resulting trend. We also illustrate how suitably calibrated filters may be used to compute piecewise linear (super) cycles, whose turning points are easy to identify.