RDP 2009-05: Macroeconomic Volatility and Terms of Trade Shocks 5. The Transmission of Terms of Trade Shocks

5.1 Terms of Trade Shocks and Expenditure Volatility

In this section we extend our analysis to consider how structural features of the economy affect the transmission of terms of trade shocks to the different components of expenditure. We start by estimating Equation (2) with no interaction terms, using various expenditure components of output as the dependent variables. Table 7 shows the result of this exercise.

Table 7: Panel Regression Results – Output Volatility
Fixed-effects estimation, five-year blocks, the first ending in 1975, the last in 2005
Regression
Dependent variable
Household consumption Gross fixed capital formation Government consumption Gross national expenditure Exports Imports
[7.1] [7.2] [7.3] [7.4] [7.5] [7.6]
Terms of trade variables
σ Terms of trade t 0.22*** 0.02 0.15* 0.16*** 0.21*** 0.14***
Control variables
Openness t−1 0.45*** 0.01 0.19 −0.05 0.20* −0.02
Credit t−1 0.05 0.16* −0.15 0.16** 0.01 0.05
Floating exchange rate t−1 0.05 0.08 0.15 −0.03 0.17** 0.05
Strict monetary policy t−1 −0.45*** −0.22* 0.22* −0.28** −0.19* −0.25*
Number of countries/observations 71/411 71/411 71/411 71/411 71/411 71/411
R2 within 0.24 0.10 0.22 0.22 0.18 0.25
Notes: ***, **, and * indicate that coefficients are significant at the 1, 5 and 10 per cent levels, respectively, using robust standard errors. All regressions include country- and time-fixed effects.

Focusing first on the domestic components, we find that an increase in the volatility of terms of trade shocks has a statistically significant positive effect on the volatility of (private and public) consumption, as well as the volatility of overall gross national expenditure (GNE).

While terms of trade movements are commonly thought to have a noticeable effect on investment, we find a small positive but statistically insignificant effect. One possible explanation for this finding is that the impact of terms of trade shocks on investment will depend upon the expected persistence of those shocks (Kent and Cashin 2003). Longer-lasting shocks, which affect expected returns to domestic production, are more likely to alter investment intentions; a purely temporary shock will affect current domestic income, but not future returns to domestic production. Since our sample contains a mixture of permanent and temporary shocks, it is quite plausible that, in aggregate, the volatility of these shocks will have little impact on investment volatility. The small impact of terms of trade volatility on investment volatility could also reflect the fact that our measure of gross fixed capital investment also includes dwelling and government investment, which are possibly less sensitive to terms of trade shocks.

Table 7 also suggests that terms of trade volatility has a positive and significant impact on the volatility of imports and exports. This is as expected since volatility in the relative prices of exports and imports is likely to reflect changes in global demand and supply, and influence domestic production. Among the control variables, the results also suggest that strict monetary policy reduces the volatility of most categories of domestic spending and exports. In particular, the negative relationship between strict monetary policy and consumption volatility could imply that better-anchored inflation expectations promote inter-temporal consumption smoothing or that strict monetary policy stabilises consumption to the extent that it results in less volatile nominal incomes. This term could also be picking up the impact of lower nominal interest rates, which are likely to have eased borrowing constraints on households (Kent, Ossolinski and Willard 2007) – at least in the second half of the sample (which concludes in 2005, before the onset of the Global Financial Crisis).

5.2 Stabilising Expenditure Volatility

Table 8 shows the results when we interact our institutional variables with terms of trade volatility in regressions including the various expenditure components of GDP. As in the output volatility regressions above, we find that the coefficient on the terms of trade term (by itself) remains positive (and significant) in most of these regressions.

Table 8: Panel Regression Results – Volatility of Output Components
Fixed-effects estimation, five-year blocks, the first ending in 1975, the last in 2005
Dependent variable: Consumption
σ Terms of trade t 0.61*** 0.24*** 0.23*** 0.60***
σ Terms of trade t * Credit t−1 −0.11**     −0.10*
σ Terms of trade t * Floating exchange rate t−1   −0.09   −0.07
σ Terms of trade t * Strict monetary policy t−1     −0.14 0.02
Wald tests (p-values)
H0: terms of trade coefficients (jointly) = 0 0.00 0.00 0.00 0.00
H0: institutional coefficients (jointly) = 0 0.07 0.41 0.00 0.00
Dependent variable: Gross fixed capital formation
σ Terms of trade t 0.13 0.05 0.04 0.07
σ Terms of trade t * Credit t−1 −0.03     −0.00
σ Terms of trade t * Floating exchange rate t−1   −0.13   −0.12
σ Terms of trade t * Strict monetary policy t−1     −0.17 −0.15
Wald tests (p-values)
H0: terms of trade coefficients (jointly) = 0 0.80 0.28 0.28 0.17
H0: institutional coefficients (jointly) = 0 0.11 0.15 0.08 0.05
Dependent variable: Government consumption expenditure
σ Terms of trade t 0.32 0.17** 0.17** 0.27
σ Terms of trade t * Credit t−1 −0.05     −0.02
σ Terms of trade t * Floating exchange rate t−1   −0.09   −0.08
σ Terms of trade t * Strict monetary policy t−1     −0.18* −0.15
Wald tests (p-values)
H0: terms of trade coefficients (jointly) = 0 0.07 0.06 0.03 0.09
H0: institutional coefficients (jointly) = 0 0.65 0.12 0.11 0.18
Dependent variable: Gross national expenditure
σ Terms of trade t 0.41** 0.17*** 0.18** 0.37**
σ Terms of trade t * Credit t−1 −0.07     −0.05
σ Terms of trade t * Floating exchange rate t−1   −0.05   −0.03
σ Terms of trade t * Strict monetary policy t−1     −0.17 −0.11
Wald tests (p-values)
H0: terms of trade coefficients (jointly) = 0 0.00 0.00 0.00 0.00
H0: institutional coefficients (jointly) = 0 0.05 0.77 0.02 0.02
Dependent variable: Exports
σ Terms of trade t 0.28 0.20*** 0.22*** 0.26
σ Terms of trade t * Credit t−1 −0.02     −0.01
σ Terms of trade t * Floating exchange rate t−1   0.03   0.04
σ Terms of trade t * Strict monetary policy t−1     −0.07 −0.07
Wald tests (p-values)
H0: terms of trade coefficients (jointly) = 0 0.00 0.00 0.00 0.01
H0: institutional coefficients (jointly) = 0 0.91 0.09 0.08 0.26
Dependent variable: Imports
σ Terms of trade t 0.11 0.16*** 0.16*** 0.02
σ Terms of trade t * Credit t−1 0.01     0.05
σ Terms of trade t * Floating exchange rate t−1   −0.09   −0.08
σ Terms of trade t * Strict monetary policy t−1     −0.21** −0.24**
Wald tests (p-values)
H0: terms of trade coefficients (jointly) = 0 0.00 0.00 0.00 0.00
H0: institutional coefficients (jointly) = 0 0.47 0.18 0.01 0.02
Notes: ***, **, and * indicate that coefficients are significant at the 1, 5 and 10 per cent levels, respectively, using robust standard errors. Control variables not shown for sake of brevity. All regressions include country- and time-fixed effects

Our results suggest that financial market development reduces the impact of terms of trade volatility on consumption volatility, and that this effect is statistically significant. The obvious interpretation of this result is that greater access to financial markets allows households to smooth their consumption in response to income volatility resulting from terms of trade shocks. When households are unable to save or borrow to smooth income fluctuations, consumption growth is more volatile. This is an interesting finding because previous studies (such as Beck et al 2006) have tended to downplay the extent to which financial market development stabilises output volatility in the presence of terms of trade shocks. It is also possible that the financial market development term – proxied by the ratio of credit to GDP – is also capturing the effects of access to business credit, or more generally, a broader set of economic reforms.

Among the other components of domestic spending, the coefficients on the strict monetary policy and flexible exchange rate interaction variables are usually negative, but rarely statistically significant. One exception is the strict monetary policy interaction term in the government consumption volatility regression, which is significant. Also, the strict monetary policy interaction term is almost significant at the 10 per cent level in the GNE and investment regressions, while the floating exchange rate interaction term is marginally significant in the investment volatility equation (with a p-value of 0.13). Overall, these results provide weak evidence that adopting a monetary policy regime that is strict on inflation and a flexible exchange rate regime can help to stabilise the domestic components of demand in the face of terms of trade shocks.

Turning to the external variables, our results suggest that strict monetary policy helps to reduce the volatility of imports growth in the presence of terms of trade shocks. The same is true of a floating exchange rate regime, though this result is on the margin of being significant (with a p-value of 0.14). Given that demand for imports is likely to be closely tied to domestic demand, this result provides evidence for our suggestion that these institutions help to mitigate the effect of terms of trade shocks on domestic economic conditions. None of the institutional variables has a statistically significant impact on the volatility of export growth in the presence of terms of trade shocks, although a floating exchange rate in particular is likely to stabilise income flows in response to terms of trade shocks.