An Experimental Study of Complex-Offer Auctions: Payment Cost Minimization vs. Offer Cost Minimization

A Payment Cost Minimization (PCM) auction has been proposed as an alternative to the Offer Cost Minimization (OCM) auction to be used in wholesale electric power markets with the intention to lower the procurement cost of electricity. Efficiency concerns about this proposal have relied on the assumption of true production cost revelation. Using an experimental approach, I compare the two auctions, strictly controlling for the level of unilateral market power. A specific feature of these complex-offer auctions is that the sellers submit not only the quantities and the minimum prices at which they are willing to sell, but also the start-up fees that are designed to reimburse the fixed start-up costs of the generation plants. I find that both auctions result in start-up fees that are significantly higher than the start-up costs. Overall, the two auctions perform similarly in terms of procurement cost and efficiency. Surprisingly, I do not find a substantial difference between less market power and more market power designs. Both designs result in similar inefficiencies and equally higher procurement costs over the competitive prediction. The PCM auction tends to have lower price volatility than the OCM auction when the market power is minimal but this property vanishes in the designs with market power. These findings lead me to conclude that both the PCM and the OCM auctions do not belong to the class of truth revealing mechanisms and do not easily elicit competitive behavior.

Simple-Offer vs. Complex-Offer Auctions in Deregulated Electricity Markets

In my recent experimental research of wholesale electricity auctions, I discovered that the complex structure of the offers leaves a lot of room for strategic behavior, which consequently leads to anti- competitive and inefficient outcomes in the market. A specific feature of these complex-offer auctions is that the sellers submit not only the quantities and the minimum prices at which they are willing to sell, but also the start-up fees that are designed to reimburse the fixed start-up costs of the generation plants. In this paper, using the experimental method I compare the performance of two complex-offer auctions (COAs) against the performance of a simple-offer auction (SOA), in which the sellers have to recover all their generation costs --- fixed and variable ---through a uniform market-clearing price. I find that the SOA significantly reduces consumer prices and lowers price volatility. It mitigates anti-competitive effects that are present in the COAs and achieves allocative efficiency more quickly.

The Molecular Characterization of a Predicted Glycosyltransferase from Lycopersicon esculentum

The synthesis of the plant cell wall is very complex, and understanding how this process occurs will lead to many benefits for future research and industries dependent upon cell walls for their products. The recent discovery of the functions of AtMUR3 and AtGT18 in Arabidopsis thaliana as xyloglucan galactosyltransferases has led to the identification of many more putative glycosyltransferases in the Arabidopsis genome. Due to the structural differences between the xyloglucans of Arabidopsis and solanaceous plants, we decided to search for putative arabinosyltransferases in the Solanaceae. Solanaceous xyloglucan is substituted by one to two arabinosyl residues at the second xylose position, and sometimes contains an arabinose at the first xylose position. In contrast, Arabidopsis xyloglucan does not contain arabinose, and is substituted by galactose at the second and third xylose position. Furthermore, the second galactose residue in Arabidopsis xyloglucan is usually fucosylated, a modification not found in solanaceous plants. Searching the database of expressed sequence tags (dbEST), we identified many likely glycosyltransferases in solanaceous plants, including tomato (Lycopersicon esculentum). AtMUR3 and AtGT18 search queries resulted in the identification of three putative glycosyltransferases in L. esculentum, which were tentatively designated LeGT1, Le1GT18, and Le2GT18. Based on phylogenetic considerations, Le2GT18 was thought to be a putative arabinosyltransferase. The gene was transformed into atmur3-3 and atgt18 mutant plants, and the resulting plants will be screened for homozygous plants with the inserted gene. The homozygous T2 plants can then be screened for changes in the composition of their cell walls. Because Le2GT18 is thought to be an arabinosyltransferase, the levels of arabinose may be increased in the xyloglucan fraction of the cell wall. If so, further testing can be performed to reveal the true function of Le2GT18.