Essays in financial transmission rights pricing

by 1964- Posner, Barry

This work examines issues in the pricing of financial transmission rights in the PJM market region. The US federal government is advocating the creation of large-scale, notfor-profit regional transmission organizations to increase the efficiency of the transmission of electricity. As a non-profit entity, PJM needs to allocate excess revenues collected as congestion rents, and the participants in the transmission markets need to be able to hedge their exposure to congestion rents. For these purposes, PJM has developed an instrument known as the financial transmission right (FTR). This research, utilizing a new data set assembled by the author, looks at two aspects of the FTR market. The first chapter examines the problem of forecasting congestion in a transmission grid. In the PJM FTR system firms bid in a competitive auction for FTRs that cover a period of one month. The auctions take place in the middle of the previous month; therefore firms have to forecast congestion rents for the period two to six weeks after the auction. The common methods of forecasting congestion are either time-series models or fullinformation engineering studies. In this research, the author develops a forecasting system that is more economically grounded than a simple time-series model, but requires less information than an engineering model. This method is based upon the arbitrage-cost methodology, whereby congesting is calculated as the difference of two non-observable variables: the transmission price difference that would exist in the total absence of transmission capacity between two nodes, and the ability of the existing transmission to reduced that price difference. If the ability to reduce the price difference is greater than the price difference, then the cost of electricity at each node will be the same, and congestion rent will be zero. If transmission capacity limits are binding on the flow of power, then a price difference persists and congestion rents exist. Three transmission paths in the Delmarva Peninsula were examined. The maximumlikelihood two-way Tobit model developed in Chapter One consistently predicts the expected responses to the independent variables that have employed, but the model as defined here does a poor job of predicting prices. This is likely due to the inability to iii include system outages (i.e., short-term changes in the structure of the transmission grid) as variables in the estimation model. The second chapter addresses the behavior of firms in the monthly auctions for FTRs. FTRs are a claim to congestion rent revenues along a certain path within the PJM grid, and are awarded in a uniform-price divisible-goods auction. Firms typically submit a schedule of bids for different amounts of FTR at different prices, akin to a demand curve. A firm bidding too high a price may cause the clearing price of the FTR to be higher than the realized value of the FTR, creating a loss from ownership of the FTR. A firm bidding too low means that it wins no FTRs, depriving itself of the ability to profit from ownership or to hedge against congestion. Several questions concerning firm behavior are addressed in this study. It is found that firms adjust their bids in response to new information that is obtained from past auctions: they raise or lower bids in accordance with changes in recent FTR prices and payoffs. Firms consistently bid below the value of the FTR (i.e., shade their bids.) This adds empirical evidence to the theoretically-posited notion that uniform-price auctions are not truth-telling, unlike the second-price auction for a non-divisible good. Firms employ greater bid-shading in response to increases in the volatility of both FTR clearing prices and realized FTR values. This validates the notion that firms are risk-averse. It is discovered that better-informed “insider” firms employ structurally different bidding strategies, but these differences do not lead to greater profits. However, profits do increase as firms gain more experience in these markets, lending credence to the notion that firms learn over time and that markets discipline poorly performing firms by either educating them or driving them out of the market. It is also found that firms that employ complicated bidding strategies enjoy greater profitability than firms which employ simple bidding strategies. A surprising corollary finding is that firm strategies do not converge to a common form, but that different firms continue to employ different strategies, and often move away from the seemingly dominant strategy. Firms can enter this market as either long-buyers or short-sellers, and it is discovered that long and short players display structurally divergent bidding strategies. This is perhaps unsurprising, given that long players can be either hedgers or speculators, but short players are overwhelmingly speculators. iv