RDP 2013-10: Stochastic Terms of Trade Volatility in Small Open Economies 1. Introduction

The terms of trade of many commodity-producing small open economies are subject to large shocks that can be an important source of macroeconomic fluctuations. Alongside times of high volatility, however, these economies also experience periods in which their terms of trade are comparatively stable. The effect of shocks to the level of the terms of trade has been widely studied. But little is known about the impact of changes in the volatility of terms of trade shocks. This is the focus of our paper.

To illustrate what we mean by changes in the volatility of terms of trade shocks, consider Figure 1. This shows the quarterly growth rate of the terms of trade for a selection of commodity-producing small open economies. At various times, each economy has experienced an increase or decrease in its terms of trade of more than 10 per cent, while fluctuations of 5 per cent or more are common. On average, these economies' terms of trade are highly volatile.

Figure 1 also reveals that terms of trade volatility varies over time. The economies in the figure have all endured episodes of extremely high terms of trade volatility, including during the 1970s for Australia and New Zealand, 1980s for Brazil, Mexico and South Africa, and 2000s for Australia and Canada. But these economies have also experienced lengthy periods in which their terms of trade were comparatively stable. Examples include the late 1990s for Australia, New Zealand and Mexico. We refer to these shifts in the absolute magnitude of changes in the terms of trade as volatility shocks.

To examine the economic implications of changes in terms of trade volatility, we first estimate the empirical process of the terms of trade for the six economies featured in Figure 1. We use the estimated time series of terms of trade volatility produced in this exercise to identify the effect of volatility shocks on output, consumption, investment, the current account and prices in a vector autoregression (VAR). We then set up and augment a small open economy real business cycle model to incorporate stochastic terms of trade volatility. We demonstrate that this model can broadly replicate the empirical responses produced by the VAR and use it to explore the theoretical causes and sectoral impacts of these responses. Finally, we quantify the importance of terms of trade volatility shocks as a source of macroeconomic fluctuations.

Our empirical results suggest that, by itself, an increase in terms of trade volatility depresses domestic demand and leads to an improvement in the current account, leaving the response of aggregate output ambiguous. Our model successfully replicates these patterns. It also suggests that increased terms of trade volatility causes a shift in the composition of output from non-tradeables to tradeables and a substitution in factor inputs from capital to labour.

The effects of terms of trade volatility shocks are generally small. But, interacted with shocks to the level of the terms of trade, variance decompositions suggest that these shocks have an economically meaningful impact. For a typical small open economy we find that shocks to volatility account for around one quarter of the total impact of terms of trade shocks on the standard deviations of output, consumption and investment.