Skip to main content

Analyzing the Impact of Electricity Market Structure Changes and Mergers: The Importance of Forward Commitments


We investigate how the effects of market structure changes and mergers in restructured electricity markets depend on the level of forward contracting. We develop a Cournot model of Alberta’s wholesale electricity market that incorporates firms’ forward positions. Using data from 2013 to 2014, we calibrate the forward positions of the five largest firms and simulate the effects of different market structure changes, including a hypothetical merger with asset divestitures. We examine the sensitivity of the simulated effects to assumptions with regard to the firms’ forward commitments. The wholesale market effects of these transactions depend critically on firms’ forward commitments. These results demonstrate the importance of establishing a clear understanding of the size and nature of forward commitments in forecasting the effects of mergers and other market structure changes in wholesale electricity markets.

This is a preview of subscription content, access via your institution.

Fig. 1
Fig. 2
Fig. 3
Fig. 4


  1. 1.

    This concern could also arise in other commodities markets that can be concentrated, including natural gas and markets for certain metals. See for example Slade and Thille (2006) and Van Eijkel et al. (2016).

  2. 2.

    In the United States, the electricity industry has been subject to a recent wave of consolidation through mergers (EIA 2013). For example, the recent merger of Exelon and Constellation was valued at $8 billion dollars. Wolak (2011) demonstrated that this merger raised significant antitrust concerns and emphasized the importance of the merging firms’ forward contracts.

  3. 3.

    Under virtual divestitures, the buyer purchases from the owner of a power plant the right to sell power for a portion of the asset’s capacity and for a pre-specified period of time (e.g., 20 years in Alberta). The sale of the capacity is virtual because no physical production capacity changes hands. Rather, the buyer of the virtual capacity can sell power into the wholesale market, while the physical operation of the asset remains with the owner. The buyer of the virtual asset earns the wholesale price on production and pays the owner of the asset for the variable and fixed costs of maintaining the asset (Ausubel and Cramton 2010; MSA 2012).

  4. 4.

    See for example Ausubel and Cramton (2010) and Frutos and Fabra (2012) for a discussion of the extent of virtual divestitures in different jurisdictions and the associated economic issues.

  5. 5.

    Alternatively, numerous jurisdictions provide capacity payments that compensate generators for the capacity that they provide to the power system, in addition to revenues they earn in energy markets. This alternative market design aims to correct for market imperfections that arise in energy-only markets; see Joskow (2008) for details.

  6. 6.

    See MSA(2012, pp. 30–33) and Appendix 2 for a discussion of forward contracting in Alberta.

  7. 7.

    Commonly traded forward products include flat products that provide a fixed quantity in every hour of the contract period (e.g, a month) and peak and off-peak products that provide a fixed quantity for peak or off-peak hours (MSA 2010).

  8. 8.

    This flexible modeling approach comes at a cost: Our calibrated forward positions lack the statistical properties that are necessary to carry out statistical inference.

  9. 9.

    Others have utilized econometric techniques to estimate firms’ forward positions (Hortacsu and Puller 2008; Allcott 2012). However, these approaches often limit the model complexity and require closed-form solutions. Recent literature begins to fill the gap between parameters estimated with the use of mathematical programming and econometrics. Reguant (2014) establishes an innovative econometric approach to model non-convexities in firms’ marginal cost to model bidding behavior in electricity markets. Su and Judd (2012) demonstrate that an econometric approach that is commonly used to estimate parameters in structural models yields the same parameter estimates as those that are obtained by solving a mathematical program that is similar to ours.

  10. 10.

    We do not explicitly define the distribution of the stochastic process that generates the deviations of our model’s estimated production decisions from observed output (e.g., due to idiosyncratic cost shocks). We utilize advances in mathematical programming methods to calibrate forward positions that rely on standard nonlinear optimization techniques (Ferris et al. 2005). A detailed characterization of the distribution of our calibrated forward positions is beyond the scope of the current analysis.

  11. 11.

    The system of constraints that are defined in (7)–(10) represents a “square” complementarity program. In the numerical program we verify that the each firm’s profit functions are globally pseudoconcave and locally strictly concave at the solution to the MPCC. This ensures that the Cournot–Nash equilibrium exists and is unique (Kolstad and Mathiesen 1991).

  12. 12.

    To alleviate concerns that our calibrated forward positions are sensitive to the starting values selected, we also utilize a solver called KNITRO which has a multi-start algorithm which randomly selects starting values for all endogenous variables and then launches the Levenberg–Marquardt algorithm to search for the solution to the MPCC from this starting point (Xian et al. 2004). For each month in our sample, we use the multi-start algorithm to solve the MPCC 100 times. We find systematic convergence to the same forward positions regardless of the starting values.

  13. 13.

    The large difference between the minimum and maximum observed market-clearing wholesale spot prices arise because of the heterogeneous portfolio of generation assets, market power execution, and periodic scarcity events triggering a price equal to the price-cap of $999.99. Wholesale spot electricity prices exhibit greater dispersion than natural gas prices because they reflect natural gas prices as well as these other factors.

  14. 14.

    See Appendix 1 for additional details on the data sources that are used to estimate unit-level marginal costs.

  15. 15.

    All wind facilities and the majority of cogeneration units offer all of their production at a price of zero into the wholesale market, although several cogeneration units produce electricity beyond their on-site needs and submit non-zero bids. For these units, because they operate on natural gas we use unit-specific heat rates and natural gas prices to impute their underlying marginal cost as we would with a traditional natural gas facility.

  16. 16.

    This implies that we apply the observed output of hydro for each hour and investigate the non-hydro units needed to satisfy the remaining demand. The potential biases from this assumption are likely to be small in Alberta due to the limited Hydro generation. Hydro makes up 6–7% of generation capacity and 2–3% of annual generation output in Alberta.

  17. 17.

    Wolfram (1999) and Bushnell et al. (2008) use the forced outage factor (fof) to reflect the probability of a complete unit forced outage. We use Canadian generation unit data on fof’s that are derated to account for partial unit outages (CEA 2012).

  18. 18.

    TransCanada’s R-squared increases to 0.94 with the removal of a 30-MW high marginal cost biomass unit that fits poorly in our polynomial approximation.

  19. 19.

    Limited imports from Montana are included in the estimation of the BC import supply function.

  20. 20.

    The linear specification is chosen for computational ease. Alternative specifications such as log-linear increase the computational complexity of the bilevel Cournot optimization substantially. A log-linear specification increases the price-elasticity of imports (and hence, the elasticity of residual demand). While this reduces the level of implied forward contracts, the key qualitative conclusions of the analysis persist. See Sect. 8.3 for a related discussion.

  21. 21.

    The cities considered in AB, BC, and SK are Calgary, Edmonton, Vancouver, and Saskatoon, respectively. The results of the analysis are robust to the consideration of alternative large cities in each province and higher degree polynomials.

  22. 22.

    This is particularly relevant in the current context because imports from BC and SK arise primarily from hydro and fossil fuel generators, respectively. If the imports arose from wind production, the strong geographical correlation that is observed in wind output would raise concerns over the validity of our IVs.

  23. 23.

    When solving the perfectly competitive equilibrium, this problem takes the form of a Mixed Complementarity program and is solved using Newton’s Method and the PATH algorithm in GAMS.

  24. 24.

    We also compared the observed and model estimated firm-specific Lerner Indices \(\left( \frac{P_t-C^\prime _{jt}(q_{jt})}{P_t}\right)\) for all \(j = 1, 2,\ldots , 5\). We find that our model captures the heterogeneity in firm-specific behavior well. These results are available upon request.

  25. 25.

    Similar to Bushnell et al. (2008), we find that the Cournot model with no forward commitments fits the data poorly; for almost all hours in our sample, the Cournot model with zero forward commitments yields prices that exceed the price-cap of $1,000. This supports prior literature that emphasizes the importance of accounting for forward commitments.

  26. 26.

    The marginal cost of coal units includes dynamic (non-convex) costs that are associated with ramping and turning on and off inflexible coal units. Mansur (2008) uses a reduced-form approach to illustrate the importance of accounting for these start-up costs and ramping constraints, especially in off-peak periods when coal units set prices. Due to computational complexity, we are unable to account for these non-convex costs.

  27. 27.

    While the fit of the Cournot model for certain variables is statistically different in peak and off-peak hours, the average difference between observed and estimated outcomes is smaller for Cournot than the perfectly competitive model.

  28. 28.

    The largest deviation between the Cournot and observed outcomes is in the Fringe’s output. This is the result of two factors: First, we overestimate prices on average and particularly in off-peak hours when the fringe is not capacity constrained. Second, our approach to exogenously defining hourly capacity availability results in expected fringe capacities that exceed observed capacities over our sample period by approximately 5%.

  29. 29.

    See Werden et al. (2004) for a related discussion.

  30. 30.

    The calibrated forward positions remained largely unchanged with the exclusion of the hours with the highest 5% of demand. The 5th and 95th percentile differences in forward positions from our baseline estimates were −14.46 and 22.32 MWs (−1.58 and 1.51%).

  31. 31.

    TransAlta may prefer to produce at a loss for several hours rather than shut down its coal assets. Our model would attribute this behavior to having a high forward contract position. Reguant (2014) emphasizes the importance of future research that incorporates these dynamic non-convexities when modeling firms’ marginal cost functions.

  32. 32.

    The calibrated forward positions exhibit lower variation than the overall NGX market. The average monthly percentage change in total forward traded MWs on the NGX is 19.2%. More detailed information on the variation in the calibrated forward positions are available upon request.

  33. 33.

    Willems et al. (2009) uses both a Supply Function Equilibrium (SFE) and Cournot model to model firm behavior. The authors find forward contract positions of around 25% of installed capacity for the SFE model and 50% for the Cournot Model.

  34. 34.

    The results for the off-peak hours are qualitatively similar to peak hours. These results are available upon request.

  35. 35.

    We utilize prior literature that analyzes how firms’ forward commitments change in the presence of a merger to guide our initial comparative statics (e.g., see Brown and Eckert 2017).

  36. 36.

    The combined market share of the top 4 firms was 54.7% in 2013 and 54.1% in 2014. Further, the Herfindahl–Hirschman Index (HHI) was 939.5 in 2013 and 933.4 in 2014 (MSA 2014).

  37. 37.

    A comparison of bidding by BR3 and BR4 in 2013 and 2014 suggests there was a sizable change in the offer behavior of these units. However, other explanations for this change in behaviour exist, including the observation that maintenance outages for these units increased after the expiration, which further motivates the use of this counterfactual analysis.

  38. 38.

    An exogenous change to a firm’s generation portfolio due to an asset-level PPA contract expiration could affect its firm-level contracting decision. Section 8.1 considers changes to firm-level forward contracts as a result of the PPA expiration.

  39. 39.

    Section 8.2 analyzes the changes in profits and total costs in the divestiture and no-divestiture cases.

  40. 40.

    In all cases, all of the firm’s profits increase. ATCO’s (ENMAX’s) total costs decrease (increase) because it gains (loses) low-cost base-load coal assets. The increase in market power execution causes a small increase in total production costs. Detailed results are available upon request.

  41. 41.

    Further reductions in AE’s forward quantities can result in the merger with divestiture increasing price. For example, reducing the merged firm’s forward MWs by 40%, which is less than the 50% assumed in Wolak (2011), results in an average price of $89.51.

  42. 42.

    Note however that the degree of inelasticity of our residual demand curve is consistent with the results for U.S. markets, with the exception of California, which exhibits a much higher reliance on imports. See for example Bushnell et al. (2008).

  43. 43.

    This result will not necessarily hold if assets are divested to a large dominant firm that has large amount of forward commitments. However, if the dominant firm is able to adjust its forward contracts after acquiring the assets, the asset divestiture to a dominant firm may result in more market power compared to asset divestiture to the competitive fringe.

  44. 44.

    Brown and Olmstead (2017) use 2008–2014 data from Alberta’s electricity market and a Monte-Carlo approach that selects unit-specific non-zero coal bids into the wholesale market in low demand hours to represent each coal unit’s marginal costs. The authors demonstrate that this approach results in marginal cost estimates that closely reflect estimates that are obtained with the use of coal unit heat rates and PRB coal prices.


  1. Aid, R., Chemla, G., Porchet, A., & Touzi, N. (2011). Hedging and vertical integration in electricity markets. Management Science, 57(8), 1438–1452.

    Article  Google Scholar 

  2. Alberta Energy (2014). Coal statistics. Available at:

  3. Allaz, B., & Vila, J. (1993). Cournot competition, forward markets, and efficiency. Journal of Economic Theory, 59(1), 1–16.

    Article  Google Scholar 

  4. Allcott, H. (2012). The smart grid, entry, and imperfect competition in electricity markets, NBER Working Paper 18071.

  5. Anderson, E., Hu, X., & Winchester, D. (2007). Forward contracts in electricity markets: The Australian experience. Energy Policy, 35(5), 3089–3103.

    Article  Google Scholar 

  6. Ausubel, L., & Cramton, D. (2010). Virtual power plant auctions. Utilities Policy, 18, 201–208.

    Article  Google Scholar 

  7. Baldick, R., Grant, R., & Kahn, E. (2004). Theory and application of linear supply function equilibrium in electricity markets. Journal of Regulatory Economics, 25(2), 143–167.

    Article  Google Scholar 

  8. Borenstein, S., & Bushnell, J. (1999). An empirical analysis of the potential for market power in California’s electricity industry. Journal of Industrial Economics, 47(3), 285–323.

    Article  Google Scholar 

  9. Borenstein, S., Bushnell, J., & Knittel, C. (1999). Market power in electricity markets: Beyond concentration measures. The Energy Journal, 20(4), 65–88.

    Article  Google Scholar 

  10. Borenstein, S., Bushnell, J., & Wolak, F. (2002). Measuring market inefficiencies in California’s deregulated wholesale electricity market. American Economic Review, 92(5), 1376–1405.

    Article  Google Scholar 

  11. Brandts, J., Pezanis-Christou, P., & Schram, A. (2008). Competition with forward contracts: A laboratory analysis motivated by electricity market design. The Economic Journal, 118, 192–214.

    Article  Google Scholar 

  12. Brown, D., & Eckert, A. (2017). Electricity market mergers with endogenous forward contracting. Journal of Regulatory Economics, 51(3), 269–310.

    Article  Google Scholar 

  13. Brown, D. & Olmstead, D. (2017). Measuring market power and the efficiency of Alberta’s restructured electricity market: An energy-only market design. Canadian Journal of Economics. (in press).

  14. Bushnell, J. (2005). Looking for trouble: Competition policy in the U.S. electricity industry. In J. Griffin & S. Puller (Eds.), Electricity deregulation: Choices and challenge (pp. 256–296). Chicago: University of Chicago Press.

    Chapter  Google Scholar 

  15. Bushnell, J. (2007). Oligopoly equilibria in electricity contract markets. Journal of Regulatory Economics, 32(3), 225–245.

    Article  Google Scholar 

  16. Bushnell, J., Mansur, E., & Saravia, C. (2008). Vertical arrangements, market structure, and competition: An analysis of restructured electricity markets. American Economic Review, 98(1), 237–266.

    Article  Google Scholar 

  17. Bushnell, J., Mansur, E., & Wolak, F. (2014). Vertical forward commitments. Working Paper.

  18. CASA. (2004). A study on the efficiency of Alberta’s electrical supply system. Project # CASA-EEEC-02-04. Clean air strategic alliance.

  19. CEA. (2012). 2007 generation equipment status: Annual report. Canadian electricity association.

  20. Crew, M., & Kleindorfer, K. (2002). Regulatory economics: Twenty years of progress? Journal of Regulatory Economics, 21(1), 5–22.

    Article  Google Scholar 

  21. Czyzyk, J., Mesnier, M., & More, J. (1998). The NEOS server. IEEE Journal of Computational Science and Engineering, 5(3), 68–75.

    Article  Google Scholar 

  22. EIA. (2013). Large mergers drive changes in ownership of electric generation in 2012. U.S. Energy Information Administration.

  23. EIA. (2016). Updated capital cost estimates for utility scale electricity generating plants. U.S. Energy Information Administration.

  24. Federal Energy Regulatory Commission. (2012). Analysis of horizontal market power under the Federal Power Act. In Order Reaffiriming Commission Policy and Terminating Proceeding. 138 FERC 61,109. Docket No. RM11-14-000.

  25. Ferreira, J. (2014). Capacity precommitment, price competition and forward markets. Economics Letters, 122(2), 362–364.

    Article  Google Scholar 

  26. Ferris, M., Dirkse, S., & Meeraus, A. (2005). Mathematical programs with equilibrium constraints: Automatic reformulation and solution via constrained optimization. In T. Kehoe, T. Srinivasan, & J. Whalley (Eds.), Frontiers in applied general equilibrium modeling (pp. 67–94). Cambridge: Cambridge University Press.

    Chapter  Google Scholar 

  27. Ferris, M., Dirkse, S., Jagla, J., & Meeraus, A. (2009). An extended mathematical programming framework. Computers & Chemical Engineering, 33(12), 1973–1982.

    Article  Google Scholar 

  28. Frutos, M., & Fabra, N. (2012). How to allocate forward contracts: The case of electricity markets. European Economic Review, 56(3), 451–469.

    Article  Google Scholar 

  29. Green, R. (1999). The electricity market in England and Wales. Journal of Industrial Economics, 47(1), 107–124.

    Article  Google Scholar 

  30. Holmberg, P. (2011). Strategic forward contracting in the wholesale electricity market. The Energy Journal, 32(1), 169–202.

    Article  Google Scholar 

  31. Holmberg, P., & Willems, B. (2015). Relaxing competition through speculation: Committing to a negative supply slope. Journal of Economic Theory, 159, 236–266.

    Article  Google Scholar 

  32. Hughes, J., & Kao, J. (1997). Strategic forward contracting and observability. International Journal of Industrial Organization, 16(1), 121–133.

    Article  Google Scholar 

  33. Hortacsu, A., & Puller, S. (2008). Understanding strategic bidding in multi-unit auctions: A case study of the texas electricity spot market. RAND Journal of Economics, 39(1), 86–114.

    Article  Google Scholar 

  34. Joskow, P. (2008). Capacity payments in imperfect electricity markets: Need and design. Utility Policy, 16(3), 159–170.

    Article  Google Scholar 

  35. Klemperer, P., & Meyer, M. (1989). Supply function equilibria in oligopoly under uncertainty. Econometrica, 57(6), 1243–1277.

    Article  Google Scholar 

  36. Kolstad, C., & Mathiesen, L. (1991). Computing Cournot–Nash equilibria. Operations Research, 39(5), 739–748.

    Article  Google Scholar 

  37. Le Coq, C., & Orzen, H. (2006). Do forward markets enhance competition? Experimental evidence. Journal of Economic Behavior and Organization, 61(3), 415–431.

    Article  Google Scholar 

  38. Liski, M., & Montero, J. (2006). Forward trading and collusion in oligopoly. Journal of Economic Theory, 131(1), 212–230.

    Article  Google Scholar 

  39. Mahenc, P., & Salanie, F. (2004). Softening competition through forward trading. Journal of Economic Theory, 116(2), 282–293.

    Article  Google Scholar 

  40. Mansur, E. (2007). Upstream competition and vertical integration in electricity markets. Journal of Law and Economics, 50(1), 125–156.

    Article  Google Scholar 

  41. Mansur, E. (2008). Measuring welfare in restructured electricity markets. Review of Economics and Statistics, 90(2), 369–386.

    Article  Google Scholar 

  42. McAfee, R., & Williams, M. (1992). Horizontal mergers and antitrust policy. Journal of Industrial Economics, 40(2), 181–187.

    Article  Google Scholar 

  43. Miller, N. (2013). Forward contracting and the welfare effects of mergers. U.S. Department of Justice Antitrust Division Economic Analysis Group Discussion Paper 13-1.

  44. Moselle, B., Newbery, D., & Harris, D. (2006). Factors affecting geographical market definition and merger control for the dutch electricity sector. Report by The Brattle Group.

  45. MSA. (2010). An introduction to Alberta’s financial electricity market. Alberta Market Surveillance Administrator.

  46. MSA. (2012). State of the market report 2012: An assessment of structure, conduct, and performance of Alberta’s wholesale electricity market. Alberta Market Surveillance Administrator.

  47. MSA. (2013). Market share offer control 2013. Alberta Market Surveillance Administrator

  48. MSA. (2014). Market share offer control 2014. Alberta Market Surveillance Administrator.

  49. MSA. (2015). Retail market update 2015. Alberta Market Surveillance Administrator.

  50. Murphy, F., & Smeers, Y. (2010). On the impact of forward markets on investments in oligopolistic markets with reference to electricity. Operations Research, 58(3), 515–528.

    Article  Google Scholar 

  51. Perry, M., & Porter, R. (1985). Oligopoly and the incentive for horizontal merger. American Economic Review, 75(1), 219–227.

    Google Scholar 

  52. Puller, S. (2007). Pricing and firm conduct in California’s deregulated electricity market. The Review of Economics and Statistics, 89(1), 75–87.

    Article  Google Scholar 

  53. Reguant, M. (2014). Complementary bidding mechanisms and startup costs in electricity markets. Review of Economic Studies, 81(4), 1708–1742.

    Article  Google Scholar 

  54. Schubert, J. (2015). The impact of forward contracting on tacit collusion: Experimental evidence. Journal of Economic Behavior and Organization, 119, 109–123.

    Article  Google Scholar 

  55. Slade, M., & Thille, H. (2006). Commodity spot prices: An exploratory assessment of market structure and forward-trading effects. Economica, 73(290), 229–256.

    Article  Google Scholar 

  56. Su, C., & Judd, K. (2012). Constrained optimization approaches to estimation of structural models. Econometrica, 80(5), 2213–2230.

    Article  Google Scholar 

  57. Van Eijkel, R., Kuper, G., & Moraga-Gonzalez, J. (2016). Do firms sell forward for strategic reasons? An application to the wholesale market for natural gas. International Journal of Industrial Organization, 49, 1–35.

    Article  Google Scholar 

  58. Van Koten, S., & Ortmann, A. (2013). Structural versus behavioral remedies in the deregulation of electricity markets: An experimental investigation motivated by policy concerns. European Economic Review, 64, 256–265.

    Article  Google Scholar 

  59. Werden, G., Froeb, L., & Scheffman, D. (2004). A daubert discipline for merger simulation. Antitrust, 18, 89–95.

    Google Scholar 

  60. Willems, B., Rumiantseva, I., & Weigt, H. (2009). Cournot versus supply functions: What does the data tell us? Energy Economics, 31(1), 38–47.

    Article  Google Scholar 

  61. Willig, R. (2011). Market power testimony. In the matter of the merger of exelon corporation and constellation energy group. Case No. 9271. Maryland Public Service Commission.

  62. Wolak, F. (2000). An empirical analysis of the impact of hedge contracts on bidding behavior in a competitive electricity market. International Economic Journal, 14(2), 1–39.

    Article  Google Scholar 

  63. Wolak, F. (2007). Quantifying the supply-side benefits from forward contracting in wholesale electricity markets. Journal of Applied Econometrics, 22, 1179–1209.

    Article  Google Scholar 

  64. Wolak, F. (2011). A price effect analysis of the proposed exelon and constellation merger by Frank Wolak. Case No. 9271. Maryland Public Service Commission.

  65. Wolak, F. (2014). Regulating competition in wholesale electricity supply. In N. Rose (Ed.), Economic regulation and its reform: What have we learned. Chicago: University of Chicago Press.

    Google Scholar 

  66. Wolfram, C. (1999). Measuring duopoly power in the british electricity spot market. American Economic Review, 89(4), 805–826.

    Article  Google Scholar 

  67. Xian, W., Yuzeng, L., & Shaohua, Z. (2004). Oligopolistic equilibrium analysis for electricity markets: A nonlinear complementarity approach. IEEE Transactions on Power Systems, 19(3), 1348–1355.

    Article  Google Scholar 

Download references


The authors would like to thank Lawrence J. White, two anonymous referees, and seminar participants at the University of Alberta, the University of Lethbridge, and the Canadian Economics Association Annual Conference for helpful comments and suggestions.

Author information



Corresponding author

Correspondence to David P. Brown.


Appendix 1: A Marginal Cost Functions

Data on natural gas unit heat rates were obtained from Alberta’s Market Surveillance Administrator (MSA), the Alberta Utility Commission, and the AESO. Coal unit heat rates were obtained from CASA (2004). Data on technology-specific variable O&M costs were obtained from the Energy Information Administration (EIA 2016). Hourly natural gas prices were obtained from Alberta’s Natural Gas Exchange. We use weekly coal prices from Wyoming’s Powder River Basin (PRB) from the Energy Information Administration to estimate the marginal cost of coal units in Alberta.Footnote 44 PRB coal is sub-bituminuous coal which is the primary coal used for electricity generation in Alberta (Alberta Energy 2014). Further, PRB coal and Alberta’s sub-bituminuous coal have similar heat-rate contents. We adjusted the PRB coal prices and USD-based variable O&M costs from USD to CAD using Bank of Canada exchange rates. The environmental compliance costs are analogous to those detailed in Brown and Olmstead (2017). We obtained asset-level generation capacity from the MSA. Derated forced outage statistics were obtained from the North American Electric Reliability Corporation’s Generation Availability Data System Reports.

Appendix 2: Retailing and Forward Contracting

Overview of Retailing in Alberta

Residential consumers, farms, and small commercial and industrial consumers can choose whether to purchase their electricity from the regulated default providers (the RRO) or from a competitive retailer. Large commercial and industrial sites do not have access to a regulated default option (the RRO). During our sample period, according to monthly retail statistics from the MSA, on average 59% of residential sites were on the RRO; for farm sites this percentage is 75%, while for small commercial and industrial sites it is 45%. Consumers who are on the RRO face a retail price that is fixed for the month. By regulation, the RRO rate must be based on forward electricity prices during the preceding months.

It is expected that the majority of sites (other than large industrial and commercial) that are not on the RRO have accessed a fixed rate contract through a competitive retailer. While little public data on the uptake of different competitive retail products exists, MSA (2015) indicated that 64% of residential sites that had left the RRO were on dual fuel (electricity and natural gas) contracts with fixed electricity prices. In addition, some consumers will be on single fuel electricity contracts with fixed rates. The length of fixed rate contracts ranges from 1 to 5 years.

Estimating ENMAX’s Retail Volume

Monthly data on retail electricity volumes and site counts, by customer class (residential, farm, small commercial and industrial, and large commercial and industrial) and by region are given on the MSA website. These data also divide volumes and site counts in each customer class according to whether the electricity was purchased through a default or competitive contract. While ENMAX’s sales through the RRO or default contracts are precisely identified, these data do not separate sales through competitive contracts by firm. Market shares by firm for each region and customer class from January 2012 to March 2015 are available in the MSA (2015). Market shares are by site counts for all classes except large industrial and commercial, for which market shares are reported both by site count and volume. These market shares are used to approximate ENMAX’s share of retail sales on competitive forward contracts.

To divide residential, farm, small commercial, and small residential sales according to peak and off-peak hours, we employ data from ENMAX’s monthly RRO energy charge filings. This is available through the Alberta Utilities Commission website, which reports the monthly MWs that were procured under flat and peak forward contracts based on ENMAX’s internal forecasts of peak and off-peak demand.

Over 2013–2014, ENMAX’s retail sales per hour to residential, small commercial, and small industrial customers come to 838 MWhs (561.2 MWhs) in peak (off-peak) hours. While ENMAX’s large commercial and industrial sales are not easily broken down into peak and off-peak hours, the MSA data yield an average hourly sales to large commercial and industrial customers of 1390.3 MWhs. However, it should be noted that more customers in this category are expected to purchase electricity through wholesale price pass-through contracts.

To illustrate the fit of our estimated forward positions from the Cournot model with ENMAX’s fixed-price commitments, Fig. 5 plots ENMAX’s estimated forward MWs with the use of our Cournot model, along with retail sales to farm, small commercial, and residential customers plus 30% of large industrial and commercial sales. In general, while the latter exhibits less fluctuation, the longer-term movements and level of the two series are similar, which provides further support for our forward market positions.

Fig. 5

ENMAX estimated forward position and retail commitment, peak hours including 30% of large industrial/commercial

Table 16 Hourly import supply function IV estimation

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Brown, D.P., Eckert, A. Analyzing the Impact of Electricity Market Structure Changes and Mergers: The Importance of Forward Commitments. Rev Ind Organ 52, 101–137 (2018).

Download citation


  • Electricity
  • Mergers
  • Forward contracts
  • Market power

JEL Classification

  • D43
  • L40
  • L51
  • L94
  • Q40