As a result, CREG decided to redesign the market structure. To evaluate the economic efficiency of the system, we created a model imitating XM's dispatch model, which mirrors the electricity market in Colombia. This analysis project consists of a structural model of energy dispatch to empirically evaluate the benefits or effects of the transition of the Colombian energy market in 2009. In order to analyze the impact of the reform, a dynamic model was proposed that reflects an accurate assessment of welfare detriments, and takes into account plant start-up costs.
To evaluate the economic efficiency of the system, Quantil, together with international consultants, created a model imitating XM's dispatch model, which reflects the electricity market in Colombia. This analysis project consists of a structural model of energy dispatch to empirically evaluate the benefits or effects of the transition of the Colombian energy market in 2009. The data used in this analysis comes from CREG and XM (the system operator). In order to analyze the impact of the reform, a dynamic model was proposed that reflects an accurate assessment of welfare detriments, and takes into account plant start-up costs.
For the evaluation, two models were estimated. The first is a production decision model that estimates the amount of energy produced as a function of the difference between the selling price and the production cost as a percentage of the cost, for the past, present and future. Not to leave out other relevant information, the first model also takes into account the opportunity cost of water. The model was calibrated with data from before 2009 and then used to simulate the production that would have resulted without the regulation changes. The graph indicates that on average the adjusted (simulated) supply curve is similar to the observed (actual) supply curve.
On the other hand, the second model estimates prices as a function of demand, taking into account the natural phenomena of El Niño and La Niña. This econometric model is a reduced model since it ignores the strategic behavior of the agents. The model simulates the price after the reform using pre-reform dynamics. It examines the relationship of price and aggregate production before 2009, and includes an adjustment to prices to reflect the real difference after 2009 as demand increased markedly. To achieve a more fit model, we increased the variance using an autoregressive process with the residuals, a Monte Carlo simulation and finally we added the error to draw out the price predictions. The process was repeated 100 times and the average was taken. From these two models, we created simulations of what production and cost would be after 2009 with and without market reform.
When analyzing the results with descriptive statistics, it was noted that after the reform, production efficiency increased as production costs were reduced. However, when including in the analysis the real prices and costs denominated by the energy producers, we noticed that these are higher than the simulated prices, suggesting that the benefits obtained by the reform did not reach the energy consumers who noticed an increase in prices. These facts suggest that the benefits of the energy market reform were captured by producers through market power.
In order to achieve the clearest results we took into account several factors that could influence the market power of producers. Finally, the results were inconclusive regarding the impact of the reform on overall welfare. It is concluded that production efficiency increased, but allocative efficiency decreased, as the results suggest that producers appropriated the benefits of the reform through market power.
In Projects, you can learn about real applications of our services, use cases and examples of them.
