Download presentation

Presentation is loading. Please wait.

Published byBailey Cowlishaw Modified about 1 year ago

1

2
New reforms have been passed to modernize the Swiss electricity sector. Legal (transmission and generation are separate legal entities) and functional (separate accounts and management for transmission and generation activities) unbundling of the utilities is required. The intended result of unbundling the services into separate functions will allow for greater efficiency through stronger and more transparent competition in the generation and sale activities of this sector

3
Unbundling reduces the chances of companies benefiting from vertical integration Vertical Integration can be cost effective › Substantial need for coordination across stages › High transaction costs are associated with using intermediary markets

4
A multi-stage cost function using two panel data econometric models were used: › A Generalized Least Squares (GLS) model with random intercept › A Random-Coefficients (RC) model

5
The cost function includes › Two outputs: electricity generation and distribution › Two inputs › Three output characteristic variables: customer density, load factor of the network and the capacity utilization factor at the generation level › A Linear time trend

6
C=C(Q1, Q2, PC, PL, CD, EC, LF, CU,T) › C= Total Costs › Q1= Electricity Generated › Q2= Electricity Distributed › PL=Input Factor Prices for Labor Services › PC= Input Factor Prices for Capital Services › CD= Customer Density › EC= Share of the Distributed Electricity to End Consumers › LF= Load Factor of the Network › T= Linear Time Trend Controlling for Technical Change.

7
Quadratic functional form › Unobserved heterogeneity › Readily adaptable to panel data econometric models › Relevant option for estimating scope economies › Accommodates zero values for outputs Allows for the inclusion of specialized companies Straight forward identification of economies of vertical integration. › Disadvantage: the linear homogeneity of the cost function in input prices cannot be imposed by parametric restrictions without sacrificing the flexibility of the functional form. Solution: Normalization of prices- dividing the costs and all factor prices by one common factor price

8
Another option was to use a logarithmic function › However, with the vast number of zero’s logarithmic functions could result in large errors in the estimation of scope › Could mislead conclusion about economies of vertical integration

9
The quadratic form is a flexible functional form with a second order Taylor approximation of any arbitrary function around a local approximation point. C it = 0 + Q1 Q1 it+ Q2 Q2 it + 1/2 Q1Q1 (Q1 it ) 2 + 1/2 Q2Q2 (Q2 it ) 2 + 1/2 Q1Q2 Q1 it Q2 it + PL PL it + CD CD it + BC BC it + LF LF it + CU CU it + T T t › Subscript “i” represents the company › Subscript “t” represents the year

10
The data set contains financial and technical information from 74 companies Covers a 9 year period from 36 companies are integrated with electricity distribution and hydropower generation 9 companies are only electricity distributors 29 companies are hydropower generators This data set covers 42% of total electricity distribution and 40% of total hydropower generation in Switzerland

11
Output and Input price coefficients are highly significant and have the expected positive sign. Customer Density coefficient is negative › an increase in the customer density decreases costs End Consumers coefficient is positive Load Factor and Capacity Utilization Factor are negative › A higher peak load profile leads to higher costs, ceteris paribus Time trend coefficient › GLS suggests a cost decrease in the electricity sector › RC shows the coefficient as not statistically significant RC model shows the standard deviations are all statistically significant for the outputs and intercept › Significant variation in the output coefficients and the intercept across companies › There is a considerable unobserved firm-specific heterogeneity

12

13
Global economies of scale describe the cost behavior caused by proportional changes in the entire production. Economies of scale exist when ES>1 Diseconomies of scale exist when ES<1

14

15
EVIs are present when costs can be reduced by joint production of multiple outputs across different production stages The degree of EVI across two stages can be defined as the ratio of excess costs of separate production to the costs of joining production in one company

16

17
Both GLS and RC estimates suggest the existence of EVI and scale economies across a major part of the sample. The majority of companies in the analysis have less than 100,000 customers › These results apply mainly to small and middle-sized companies RC model provides lower estimates of both economies › Gives a relatively lower weight to differences regarding fixed costs No conclusive pattern suggesting one-sided bias Table 4 confirms the presence of product-specific economies of scale

18
Additional costs could result from unbundling vertically integrated companies › Due to the significant economies of vertical integration and economies of scale that exist in a majority of the electricity companies Results support the EU policy directive that allows for companies with fewer than 100,000 customers to be exempt from any functional unbundling requirement. Unbundling processes reduce the possibilities to exploit the advantages of vertical integration

Similar presentations

© 2017 SlidePlayer.com Inc.

All rights reserved.

Ads by Google