Predicting bid prices by using machine learning methods

It is well-known that empirical analysis suffers from multicollinearity and high dimensionality. In particular, this is much more severe in an empirical study of itemized bids in highway procurement auctions. To overcome this obstacle, this article employs the regularized linear regression for the estimation of a more precise interval for project winning bids. The approach is put to the test using empirical data of highway procurement auctions in Vermont. In our empirical analysis, we first choose a set of crucial tasks that determine a bidder’s bid amounts by using the random forest variable selection method. Given the selected tasks, project bid forecasting is conducted. We compare our proposed methodology with the least square linear model based on the bias and the standard root mean square error of the bid estimates. There is evidence supporting that the suggested approach provides superior forecasts for an interval of winning bids over the competing model. As far as we know, this article is the first attempt to provide reference bids of highway construction contracts.     

The effect of information on the bidding and survival of entrants in procurement auctions

In government procurement auctions of construction contracts, entrants are typically less informed and bid more aggressively than incumbent firms. This bidding behavior makes them more susceptible to losses affecting their prospect of survival. In April of 2000, the Oklahoma Department of Transportation started releasing the internal cost estimates to complete highway construction projects. Using newly developed quantile regression approaches, this paper examines the impact of the policy change on aggressive entrants. First, we find that the information release eliminates the bidding differential between entrants and incumbents attributed to informational asymmetries. Second, we argue that the policy change affects the prospects of survival of entrants in the market. We find that those who used to exit the market relatively soon are now staying 37% longer, while at the median level bidding duration increased by roughly 68%. The policy has the potential to encourage entry in government procurement auctions and thus increase competition.     

JOINT BIDDING,GOVERNANCE AND PUBLIC PROCUREMENT COSTS:A CASE OF ROAD PROJECTS*

ABSTRACT *** : To utilize public resources efficiently, it is important to take advantage of competition in public procurement auctions to the maximum extent. Joint bidding is a common practice that potentially facilitates competition. By pooling financial and experiential resources, more firms are expected to enter the market, but it will also directly reduce competition if more than one bidder who is solely qualified makes a coalition. In theory joint bidding may or may not be beneficial to auctioneers, depending on the model. The paper empirically examines the impacts of joint bidding on firms' entry as well as bidding behaviour, using data on public road projects in developing countries. It shows that coalitional bids, in particular by local firms, would be competitive, but foreign joint ventures would undermine competition. It is also found that good governance can encourage firms' entry into the tendering and facilitate joint bidding practices.     

PROCUREMENT EXPERIMENTS WITH UNKNOWN COSTS OF QUALITY

Abstract.  We experimentally examine the efficiency and profitability of two different procurement auctions allowing for quality differences across products. We compare the vector auction with more competition on the sellers' side with a half-auction, reflecting actual procurement practice – an auction for the cheaper variant and bargaining with the contractor about the additional cost of the better quality variant. Our main hypothesis, that buyers are better off when using the vector auction instead of the half-auction, is confirmed when quality differences of variants are large and the uncertainty of the cost difference is also large.     

Are bid preferences benign? The effect of small business subsidies in highway procurement auctions

Bid preferences in procurement auctions allow firms from an identifiable group an advantage in bidding against unfavored firms. While economic efficiency is expected to fall as a result of bid preferences, government procurement costs may either increase or decrease depending on the competitive response of favored and unfavored firms. This paper uses data from California auctions for road construction contracts, where small businesses receive a 5-percent bid preference in auctions for projects using only state funds and no preferential treatment on projects using federal aid. I show that while firms' bidding behavior matches theoretical predictions, procurement costs are 3.8 percent higher on auctions using preferences. The higher procurement cost in preference auctions is attributed to reduced participation by lower cost large firms.Structural estimates of latent firm costs are then used to evaluate how efficiency and the division of surplus between firms and the government are impacted by bid preferences. Firm profits are 3.1 percent lower under bid preferences, however this is overwhelmed by the efficiency loss due to reduced large firm participation. The efficiency loss conditional on firm participation is estimated to represent around 0.1 percent of overall procurement costs. Including the adverse effect of preferences on the participation of large firms increases the estimated efficiency loss to 3.6 percent, which represents 27 cents for each additional dollar awarded to small businesses through the program. Counterfactual simulations indicate that if participation were instead inelastic to bid preferences, the 5-percent bid preference would be close to the optimal level.     

The (in) effectiveness of procurement auctions in the public sector

This article studies procurement auctions in the public sector using game theoretical tools. The article shows that when participants in an auction are agents with low abilities (low type), as is common in the public sector, they place the same contract request. As a result, the auction mechanism will rarely produce real competition, thus making this mechanism, which limits bidders to those already within the system, ineffective for use in the public sector job market.     

COMPARING PUBLIC PROCUREMENT AUCTIONS

This article contrasts two auction formats often used in public procurement: first price auctions with ex post screening of bid responsiveness and average bid auctions (ABAs), in which the bidder closest to the average bid wins. The equilibrium analysis reveals that their ranking is ambiguous in terms of revenues, but the ABA is typically less efficient. Using a data set of Italian public procurement auctions run alternately under the two formats, a structural model of bidding is estimated for the subsample of first price auctions and used to quantify the efficiency loss under counterfactual ABAs.     

Comparing competition and collusion: a numerical approach

Summary. Collusion is a serious problem in many procurement auctions. In this research, I study a model of first price sealed bid procurement auctions with asymmetric bidders. I demonstrate that the equilibrium to the model is unique and describe three algorithms that can be used to compute the inverse equilibrium bid functions. I then use the computational algorithms to compare competitive and collusive bidding. The algorithms are useful for structural estimation of auction models and for assessing the damages from bid-rigging. Received: January 14, 2000; revised version: February 28, 2001     

First-price equilibrium and revenue equivalence in a sequential procurement auction model

We analyze first-price equilibrium bidding behavior of capacity-constrained firms in a sequence of two procurement auctions. In the model, firms with a cost advantage in completing the project auctioned off at the end of the sequence may enter the unfavored first auction hoping to lose it. Equilibrium bidding in both auctions deviates from the standard Symmetric Independent Private Value auction model due to opportunity costs of bidding created by possibly employed capacity. For this sequential auction model with non-identical objects, we show that revenue equivalence applies.     

BIDS AND COSTS IN COMBINATORIAL AND NONCOMBINATORIAL PROCUREMENT AUCTIONS—EVIDENCE FROM PROCUREMENT OF PUBLIC CLEANING CONTRACTS

Combinatorial procurement auctions enable suppliers to pass their potential cost synergies on to the procuring entity and may therefore lead to lower costs and enhance efficiency. However, bidders might find it profitable to inflate their stand‐alone bids in order to favor their package bids. Using data from standard and combinatorial procurement auctions, we find that bids on individual contracts in simultaneous standard auctions without the option to submit package bids are significantly lower than the corresponding stand‐alone bids in combinatorial auctions. Further, no significant difference in procurer's cost as explained by auction format is found. (JEL D44, H57, L15)     

Sequential versus Bundle Auctions for Recurring Procurement

We compare sequential and bundle procurement auctions in a framework of successive procurement situations, where current success positively or negatively affects future market opportunities. We find that in bundle auctions procurement cost is lower and less risky than in sequential standard auctions, but still higher than in the optimal sequential auction. Only a sequential second price auction leads to the efficient outcome.     

Renegotiation on incomplete procurement contracts

This article examines the impact of incomplete contracts on procurement costs in road construction auctions. Ex ante contracts in these auctions often fail to specify all of the potential construction contingencies, and consequently, changes in scope are necessary after construction begins. Using Vermont road construction contract data, this study finds evidence that there is a statistically significant difference in costs of firms between auctions with and without extra work adjustments. Substantial adaptation costs are responsible for the higher procurement outlays in incomplete contract. I also find that bidders inflate their bids to incorporate risk premiums in incomplete contracts; however, my estimates suggest that this bidding behaviour does not affect their profit margins.     

Entry decisions and bidding behavior in sequential first-price procurement auctions: An experimental study

Though many real life auctions are run independently of each other, from the bidders' point of view they often form sequences of auctions. We investigate how behavior responds to the additional incentives that are present in such auction sequences. Comparing subjects' decisions in single first-price procurement auctions with their decisions in a game consisting of two subsequent first-price procurement auctions, we find that, in line with the theoretical prediction, entry and bidding behavior is crucially affected by the opportunity cost of early bid submission. Though, entry decisions and average bids in the auction sequence systematically deviate from the perfect Bayesian equilibrium prediction. While the nature of the opponent (human being or computer) has no significant effect on these findings, giving subjects additional feedback on winners and prices seems to reduce the deviations from the equilibrium prediction.     

Valuation structure in first-price and least-revenue auctions: an experimental investigation

In many auctions the valuation structure involves both private and common value elements. Existing experimental evidence (e.g. Goeree and Offerman in Am. Econ. Rev. 92(3):625–643, 2002) demonstrates that first-price auctions with this valuation structure tend to be inefficient, and inexperienced subjects tend to bid above the break-even bidding threshold. In this paper, we compare first-price auctions with an alternative auction mechanism: the least-revenue auction. This auction mechanism shifts the risk regarding the common value of the good to the auctioneer. Such a shift is desirable when ex post negative payoffs for the winning bidder results in unfulfilled contracts, as is often the case in infrastructure concessions contracts. We directly compare these two auction formats within two valuation structures: (1) pure common value and (2) common value with a private cost. We find that, relative to first-price auctions, bidding above the break-even bidding threshold is significantly less prevalent in least-revenue auctions regardless of valuation structure. As a result, revenue in first-price auctions is higher than in least-revenue auctions, contrary to theory. Further, when there are private and common value components, least-revenue auctions are significantly more efficient than first-price auctions.     

Auction fever: Rising revenue in second-price auction formats

The prevalent term “auction fever” visualizes that ascending auctions – inconsistent with theory – are likely to provoke higher bids than one-shot auctions. To explore and isolate causes of auction fever experimentally, we design four different strategy-proof auction formats and order these according to expected rising bids based on pseudo-endowment effect arguments (psychological ownership and disparity between willingness to pay and willingness to accept). Observed revenues in the experiment in the four formats rank as expected if bidders have private uncertain values (the private information of a bidder is the distribution of her value). A control treatment supports our view that the traditional private certain values approach prevents auction fever in the laboratory. Another control treatment with a procurement auction relates the auction fever bids to bids in a one-shot auction with real endowments. We conclude that, when bidders are uncertain about their valuations, auctions that foster pseudo-endowment may raise bids and revenues.     

Oligopoly equilibria in electricity contract markets

This paper examines the implications of forward contracting on oligopoly environments by extending the model of Allaz and Vila to an environment with multiple firms and increasing marginal cost. Estimates of key parameters of this model are taken from existing electricity markets to predict the market impact of one round of public contracting, such as those seen in auctions for retail provision and resource procurement. The results imply that, when forward contracts are present, the importance of supplier concentration is greatly magnified relative to other determinants unilateral market power such as demand elasticity.      

Entry and Competition Effects in First-Price Auctions: Theory and Evidence from Procurement Auctions

Motivated by several interesting features of the highway mowing auction data from the Texas Department of Transportation (TDoT), we study three competing procurement auction models with endogenous entry. Our entry and bidding models provide several interesting implications. For the first time, we show that even within an independent private value paradigm, as the number of potential bidders increases, bidders' equilibrium bidding behaviour can become less aggressive, and the expected procurement cost may rise because the "entry effect" is always positive and may dominate the negative "competition effect". We then develop structural models of entry and bidding corresponding to the three models under consideration, controlling for unobserved auction heterogeneity, and use the recently developed semi-parametric Bayesian estimation method to analyse the data. We select the model that best fits the data, and use the corresponding structural estimates to quantify the "entry effect" and the "competition effect" with regard to the individual bids and the procurement cost.     

The law of small numbers: investigating the benefits of restricted auctions for public procurement

A commonly accepted view in the academic literature is that dispensing with competition may only be beneficial when tendering complex contracts. However, restricted auctions are frequently used among EU member states to procure small contracts. In this article, we investigate this paradox. Using an original data set of 180 contracts used by a local public buyer of social housing between 2006 and 2009, we show that limiting competition may enable economies to be made on transaction costs while the most efficient bidders still come forward, and that abuses such as corruption or favouritism do not result. To our knowledge, this article is the first to shed light on the advantages of using restricted auctions when tendering small simple contracts.     

Industry concentration in common value auctions: theory and evidence

We examine theoretically and experimentally two countervailing effects of industry concentration in common value auctions. Greater concentration of information among fewer bidders reduces competition but increases the precision of private estimates. We demonstrate that this generally leads to more aggressive bidding. However, the reduction in competition dominates the informational effects, resulting in lower prices. We examine these hypothesized effects experimentally by conducting a series of auctions with constant informational content but distributed among a varying number of bidders. The experimental results are consistent with our theoretical predictions. The authors would like to thank Octavian Carare, Eric Friedman, Luke Froeb, Ron Harstad, Toshi Iizuka, Mike Rothkopf, Charles Thomas, and two anonymous referees, for many useful comments and suggestions.     

Bidding in private-value auctions with uncertain values

Auctions are often used to sell idiosyncratic goods difficult for potential bidders to value ex ante. Laboratory auctions with uncertainty over final values in this experiment resulted in 18% and 27% of bids above the expected value of the item in private-value first-price and English auctions, respectively. Risk-seeking preferences as measured on an individual decision task cannot explain overbidding and the first-price auction results suggest that risk aversion may not be a good explanation for bidding behavior observed with certain values. Several candidate explanations fail to explain overbidding, rather it appears to stem from some bidders who are prone to overbidding. Relative to first-price auctions, the size and frequency of overbids are significantly larger in English auctions, while more English auctions are won by overbidders. Differences between the formats appear to be driven by the dynamic nature of English auctions which is consistent with popular notions of “auction fever.”