Forecasting short-term carbon emission futures price volatility: information for hedging carbon emission futures risk

  • 465 Views
  • 82 Downloads

Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License

This paper aimed to illustrate how short-term carbon futures speculators might use short-term carbon emission futures data to predict and forecast carbon prices. The paper became apposite given ubiquitous research focussing on long-term carbon futures data, which has left out short-term carbon emission futures speculators with information. Therefore, this paper demonstrated that short-term speculators in carbon futures could indeed use short-term time series data on carbon futures to make a reliable prediction and forecasting of carbon emissions futures price volatility within a short term and thus decide on investment opportunity. The sample data results showed that short-term data could produce a dependable in-sample futures prediction since the in-sample prediction fell within the 95% confidence interval. The demonstration also showed that short-term carbon futures data could assist speculators to conduct a reliable short-term out of sample forecast of carbon futures prices within the closer period. The paper offers practical assistance to carbon futures speculators and is equally important for academic studies for business and economic students on discussions and research bordering on carbon emissions, carbon trading, environmental economics and sustainable development. More carbon short-term forecasting is encouraged – such research should compare short-term forecasting of carbon futures amongst different carbon markets.

view full abstract hide full abstract
    • Fig. 1. The spikes of carbon futures in April 2015 to July 2017
    • Fig. 2. Lines of in-sample prediction September 2015 – July 2016 and the actual lines
    • Fig.3. Out-of-sample forecast and actual lines
    • Table 1. Linear regression between of carbon futures with time trend as main regressor. Model 5: OLS, using observations September 2015 – July 2017 (T = 23). Dependent variable: carbon price
    • Table 2. In-sample prediction Sept 2015 – July 2016 For 95% confidence intervals, t(10, 0.025) = 2.228
    • Table 3. Out-of-sample forecast July 2017 – July 2018 For 95% confidence intervals, t(10, 0.025) = 2.228