Understanding the Link Between Aggregated Industrial Production and the Carbon Price - Green Energy and Efficiency: An Economic Perspective - page 118

Note that the investigation of macroeconomic risk factors 9 c to the EU

ETS by [ 7 ], as well as the impulse response function analysis in the Factor-Aug-
mented Vector AutoRegressive (FAVAR) framework conducted by [ 8 ] could also
fall in this category.

speci

2. The macroeconomic approach: in this category, we will nd the early work

“ ”

by [ 2 , 3 ], as well as a series of new studies.

First, Chevallier [ 9 ] provides several nonlinearity tests for the univariate time
series of industrial production and carbon prices, which can satisfactorily be tted

with Self-Exciting Threshold AutoRegressive (SETAR) models. In addition, a
multivariate econometric strategy featuring industrial production as the logistic
transition function in a Smooth Transition AutoRegressive (STAR) model includ-
ing both variables brings fruitful results. On the one hand, contemporaneous
changes in the industrial production index impact negatively carbon price changes
(i.e., the decrease in industrial production precedes the decrease in carbon prices).
On the other hand, changes in the industrial production index lagged one period
impact positively carbon price changes (i.e., the uptake in economic activity

encourages the carbon price to go up).
Second, Chevallier [ 10 ] uses again the EU 27 industrial production index
computed by Eurostat as a proxy of economic activity in the perimeter of EU ETS
sectors. This choice is assessed based on a preliminary forecasting exercise
including various candidates (monetary, industrial, and nancial variables): it could

be shown that the industrial production index minimizes all criteria. Then, the
originality of the article lies in the two-regime threshold cointegration exercise
between EU industrial production and the carbon price. The threshold Vector Error-
Correction Model (VECM) estimates reveal that the EU industrial production index
impacts positively the EUA futures price: the carbon-macroeconomy relationship
goes from the EU industrial production index (lagged one period) to the carbon
futures price. On the contrary, the EUA futures price has no statistically signi cant

effect on the EU industrial production index. In short, the industrial production
index governs most of the adjustment from the short-run to the long-run equilibrium
of the model. Should any short-term deviations occur, the industrial production
index acts as a feedback force to restore the long-run equilibrium relationship.
Hence, it can be concluded that industrial production leads the nonlinear mean-
reverting behavior of the carbon price, but not vice versa.
Third, Chevallier [ 11 ] con

rms that the presence of nonlinearities may con-
tribute to explain why early regression studies did not capture well the carbon-
macroeconomy relationship. In a two-regime Markov-switching model between
industrial production and carbon prices, the author shows that industrial production
has two types of effects on the carbon price: positive during the expansion regime,
and negative during the recession regime. Macroeconomic activity is likely to affect

9 i.e., dividend yields, junk bond yields, T-bill rates and market portfolio excess returns in the
Fama-French literature.