Ensemble methods in computational market dynamics
Kyabo, Gloria Sharon (2018)
Diplomityö
Kyabo, Gloria Sharon
2018
School of Engineering Science, Laskennallinen tekniikka
Kaikki oikeudet pidätetään.
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi-fe2018080233320
https://urn.fi/URN:NBN:fi-fe2018080233320
Tiivistelmä
The dynamics of financial and commodity markets play a vital role in determining the prices of goods or services traded in these markets. Whereas numerous market models have been developed to study these dynamics, more findings find weaknesses in the existing models. For that cause, this study applies the Variational Ensemble Kalman Filter to simulate electricity spot prices for the Finnish and Swedish electricity markets. The model applies an approach to derive nonlinear system dynamics from Kalman filtering called Kalman Dynamics to simulate electricity spot prices by using traders as ensemble members. These traders make bids in perception of what they expect the prices to be in regard to other traders’ bidding. We thus make them turn inward toward themselves and predict the behavior of future prices.We indicate that the dynamics in the electricity spot market influence the electricity spot prices making them to have an unstable movement.