Forecasting the California carbon credit allowance market
Vanamo, Pyry (2019-05-02)
Forecasting the California carbon credit allowance market
Vanamo, Pyry
(02.05.2019)
Julkaisu on tekijänoikeussäännösten alainen. Teosta voi lukea ja tulostaa henkilökohtaista käyttöä varten. Käyttö kaupallisiin tarkoituksiin on kielletty.
suljettu
Julkaisun pysyvä osoite on:
https://urn.fi/URN:NBN:fi-fe2019052116459
https://urn.fi/URN:NBN:fi-fe2019052116459
Tiivistelmä
Climate change has been a recurring topic in global and national politics. Policymakers all around the world are waking up to tackle the forthcoming threat of global warming. One of the ways to reduce the annual greenhouse gas emissions in an area is to create a dedicated emission trading scheme (ETS). As the public and private entities grow more and more climate-wary, new unidentified business opportunities are created for companies operating in related fields. That is why it is important to start researching these emission trading schemes to compose a fundamental analysis which can be used as a starting point for profit profitable actions.
This thesis aims to research the California state’s ETS, known as the California cap and trade. The goal is to analyze the California cap and trade market to build up a knowledge base relating to the scheme’s history, current state and future. In addition, factors affecting the prices in the carbon credit market are delineated. Using this knowledge and data gathered from the field, a statistical Autoregressive Integrated Moving-Average forecast model (ARIMA) is constructed to predict future price values for California carbon credit allowances. The main part of this thesis focuses on market analysis and forecast model creation and optimization.
In this thesis a market analysis of the California cap and trade -scheme is made and used as a basis for the ARIMA forecast model. The model is made with programming language R and it uses multiple datasets, such as historical price data of the allowance market, as a basis for the fit. First ARIMA model is created without any external variables to see if the technique is fit for the cause. The ultimate model chosen for the forecast itself uses a combination of the historical prices of the scheme and various external regressors.
Results show that ARIMA technique is useful for forecasting the California carbon credit market. The market analysis proves that the scheme is working and it’s future is guaranteed. The technical forecasting model shows good enough fit for the predictions to be used for business purposes in the future. Down the road, the model can be edited and improved by adding Machine Learning elements into it.
This thesis aims to research the California state’s ETS, known as the California cap and trade. The goal is to analyze the California cap and trade market to build up a knowledge base relating to the scheme’s history, current state and future. In addition, factors affecting the prices in the carbon credit market are delineated. Using this knowledge and data gathered from the field, a statistical Autoregressive Integrated Moving-Average forecast model (ARIMA) is constructed to predict future price values for California carbon credit allowances. The main part of this thesis focuses on market analysis and forecast model creation and optimization.
In this thesis a market analysis of the California cap and trade -scheme is made and used as a basis for the ARIMA forecast model. The model is made with programming language R and it uses multiple datasets, such as historical price data of the allowance market, as a basis for the fit. First ARIMA model is created without any external variables to see if the technique is fit for the cause. The ultimate model chosen for the forecast itself uses a combination of the historical prices of the scheme and various external regressors.
Results show that ARIMA technique is useful for forecasting the California carbon credit market. The market analysis proves that the scheme is working and it’s future is guaranteed. The technical forecasting model shows good enough fit for the predictions to be used for business purposes in the future. Down the road, the model can be edited and improved by adding Machine Learning elements into it.