Non-parametric estimation of sequential english auctions

I propose an empirical strategy to identify and to estimate non-parametrically the distribution and the density of latent valuations from the winning bids at sequential oral, ascending-price (hereafter English) auctions within the independent private-values paradigm. I evaluate the asymptotic and finite-sample properties of my approach, and the estimation strategy is applied to daily data from a fish auction held in Grenaa, Denmark, between January 2, 2000 and March 31, 2004.     

Semiparametric identification and estimation in multi-object,English auctions

Within the independent private-values paradigm, we derive the data-generating process of winning bids for two different objects sold sequentially at English auction, assuming the valuations across objects for a particular bidder are potentially dependent. We demonstrate that, within the Archimedean family of copulas, the model is identified using only observed winning bids, and then propose a semiparametric estimation strategy to recover the joint distribution of valuations. We implement our methods using data from fish auctions held in Denmark and estimate whether bundling is expected-revenue enhancing.     

Risk aversion and asymmetry in procurement auctions: Identification,estimation and application to construction procurements

This article studies a model of asymmetric risk averse bidding within the independent private value paradigm. The inherent asymmetry in cost and risk aversion imposes an original restriction on the observed bid data, an exact equality which leads to the model semiparametric identification and estimation. The unobserved arguments of this equality need to be simulated in order to estimate the bidders’ Constant Relative Risk Aversion or Constant Absolute Risk Aversion parameters and their heterogeneous cost distributions. In the Los Angeles City Hall construction contracts offered between 1994 and 2003, the model and methodology help reveal that financial asymmetries affect the firms’ cost distribution, while experience influences their degree of risk aversion.     

The econometrics of auctions with asymmetric anonymous bidders

We consider standard auction models when bidders’ identities are not-or are only partially-observed by the econometrician. We first adapt the definition of identifiability to a framework with anonymous bids and explore the extent to which anonymity reduces the possibility of identifying private value auction models. Second, in the asymmetric independent private value model which is nonparametrically identified, we generalize Guerre, Perrigne and Vuong’s estimation procedure [Optimal nonparametric estimation of first-price auctions, Econometrica 68 (2000) 525-574] and consider the asymptotic properties of our multi-step kernel-based estimator. Monte Carlo simulations illustrate the practical relevance of our estimation procedure in small data sets.     

Semiparametric estimation in models of first-price,sealed-bid auctions with affiliation

Within the affiliated private-values paradigm, we develop a tractable empirical model of equilibrium behaviour at first-price, sealed-bid auctions. The model is non-parametrically identified, but the rate of convergence in estimation is slow when the number of bidders is even moderately large, so we develop a semiparametric estimation strategy, focusing on the Archimedean family of copulae and implementing this framework using particular members—the Clayton, Frank, and Gumbel copulae. We apply our framework to data from low-price, sealed-bid auctions used by the Michigan Department of Transportation to procure road-resurfacing services, rejecting the hypothesis of independence and finding significant (and high) affiliation in cost signals.     

Identification of first-price auctions with non-separable unobserved heterogeneity

We propose a novel methodology for identification of first-price auctions, when bidders’ private valuations are independent conditional on one-dimensional unobserved heterogeneity. We extend the existing literature ( and ) by allowing the unobserved heterogeneity to be non-separable from bidders’ valuations. Our central identifying assumption is that the distribution of bidder values is increasing in the state. When the state-space is finite, such monotonicity implies the full-rank condition needed for identification. Further, we extend our approach to the conditionally independent private values model of Li et al. (2000), as well as to unobserved heterogeneity settings in which the implicit reserve price or the cost of bidding varies across auctions.     

Quantile-based nonparametric inference for first-price auctions

We propose a quantile-based nonparametric approach to inference on the probability density function (PDF) of the private values in first-price sealed-bid auctions with independent private values. Our method of inference is based on a fully nonparametric kernel-based estimator of the quantiles and PDF of observable bids. Our estimator attains the optimal rate of Guerre et al. (2000), and is also asymptotically normal with an appropriate choice of the bandwidth.     

Semi-nonparametric estimation of independently and identically repeated first-price auctions via an integrated simulated moments method

In this paper we propose to estimate the value distribution of independently and identically repeated first-price auctions directly via a semi-nonparametric integrated simulated moments sieve approach. Given a candidate value distribution function in a sieve space, we simulate bids according to the equilibrium bid function involved. We take the difference of the empirical characteristic functions of the actual and simulated bids as the moment function. The objective function is then the integral of the squared moment function over an interval. Minimizing this integral to the distribution functions in the sieve space involved and letting the sieve order increase to infinity with the sample size then yields a uniformly consistent semi-nonparametric estimator of the actual value distribution. Also, we propose an integrated moment test for the validity of the first-price auction model, and an data-driven method for the choice of the sieve order. Finally, we conduct a few numerical experiments to check the performance of our approach.     

Semiparametric efficient adaptive estimation of asymmetric GARCH models

In this paper we derive a semiparametric efficient adaptive estimator of an asymmetric GARCH model. Applying some general results from Drost et al. [1997. The Annals of Statistics 25, 786–818], we first estimate the unknown density function of the disturbances by kernel methods, then apply a one-step Newton–Raphson method to obtain a more efficient estimator than the quasi-maximum likelihood estimator. The proposed semiparametric estimator is adaptive for parameters appearing in the conditional standard deviation model with respect to the unknown distribution of the disturbances.     

Descending price multi-item auctions

We propose a simultaneous descending price auction mechanism to sell multiple heterogeneous items, each owned by a distinct seller, to a number of buyers. Each buyer has known private valuations on items, and wants at most one item. We show that if the sellers follow a descending price offer procedure and the buyers follow a greedy strategy for accepting the offers, the auction results in a nearly efficient allocation, and terminates close to a competitive equilibrium price vector. The descending price offer strategy of the sellers is close to a Nash equilibrium. However, we show that the buyers are better off waiting in our auction. There is a maximum limit (corresponding to the minimum competitive equilibrium price vector) till which they can wait without running into the risk of not winning any item. If the buyers wait within this limit, the prices can be brought arbitrarily close to a uniquely defined competitive equilibrium price vector.     

Efficient estimation in dynamic conditional quantile models

In this paper we consider the problem of semiparametric efficient estimation in conditional quantile models with time series data. We construct an M-estimator which achieves the semiparametric efficiency bound recently derived by Komunjer and Vuong (forthcoming). Our efficient M-estimator is obtained by minimizing an objective function which depends on a nonparametric estimator of the conditional distribution of the variable of interest rather than its density. The estimator is new and not yet seen in the literature. We illustrate its performance through a Monte Carlo experiment.     

Adaptive estimation of autoregressive models with time-varying variances

Stable autoregressive models are considered with martingale differences errors scaled by an unknown nonparametric time-varying function generating heterogeneity. An important special case involves structural change in the error variance, but in most practical cases the pattern of variance change over time is unknown and may involve shifts at unknown discrete points in time, continuous evolution or combinations of the two. This paper develops kernel-based estimators of the residual variances and associated adaptive least squares (ALS) estimators of the autoregressive coefficients. Simulations show that efficiency gains are achieved by the adaptive procedure.     

Identification and nonparametric estimation of a transformed additively separable model

Let r(x,z)$r(x,z)$ be a function that, along with its derivatives, can be consistently estimated nonparametrically. This paper discusses the identification and consistent estimation of the unknown functions H$H$, M$M$, G$G$ and F$F$, where r(x,z)=H[M(x,z)]$r(x,z)=H\left[M(x,z)\right]$, M(x,z)=G(x)+F(z)$M(x,z)=G\left(x\right)+F\left(z\right)$, and H$H$ is strictly monotonic. An estimation algorithm is proposed for each of the model’s unknown components when r(x,z)$r(x,z)$ represents a conditional mean function. The resulting estimators use marginal integration to separate the components G$G$ and F$F$. Our estimators are shown to have a limiting Normal distribution with a faster rate of convergence than unrestricted nonparametric alternatives. Their small sample performance is studied in a Monte Carlo experiment. We apply our results to estimate generalized homothetic production functions for four industries in the Chinese economy.     

A simple approach to the parametric estimation of potentially nonstationary diffusions

A simple and robust approach is proposed for the parametric estimation of scalar homogeneous stochastic differential equations. We specify a parametric class of diffusions and estimate the parameters of interest by minimizing criteria based on the integrated squared difference between kernel estimates of the drift and diffusion functions and their parametric counterparts. The procedure does not require simulations or approximations to the true transition density and has the simplicity of standard nonlinear least-squares methods in discrete time. A complete asymptotic theory for the parametric estimates is developed. The limit theory relies on infill and long span asymptotics and is robust to deviations from stationarity, requiring only recurrence.     

Corruption and auctions

We investigate how corruption affects the outcome of a first-price auction (bidding behavior, efficiency and the seller’s expected revenue). The auctioneer approaches the winner to offer the possibility of a reduction in his bid in exchange for a bribe. The bribe can be a percentage of the difference between the winning and the second-highest bid or a fixed amount. We show that there exists a symmetric bidding strategy equilibrium that is monotone, i.e., higher valuation buyers bid higher. Corruption does not affect efficiency but both the auctioneer’s expected bribe and the seller’s expected revenue depend on the format of the bribe payments. We also find the optimal bribe scheme.     

Partial parametric estimation for nonstationary nonlinear regressions

This paper proposes an estimation method for a partial parametric model with multiple integrated time series. Our estimation procedure is based on the decomposition of the nonparametric part of the regression function into homogeneous and integrable components. It consists of two steps: In the first step we parameterize and fit the homogeneous component of the nonparametric part by the nonlinear least squares with other parametric terms in the model, and use in the second step the standard kernel method to nonparametrically estimate the integrable component of the nonparametric part from the residuals in the first step. We establish consistency and obtain the asymptotic distribution of our estimator. A simulation shows that our estimator performs well in finite samples. For the empirical illustration, we estimate the money demand functions for the US and Japan using our model and methodology.     

Local polynomial estimation of nonparametric simultaneous equations models

We define a new procedure for consistent estimation of nonparametric simultaneous equations models under the conditional mean independence restriction of Newey et al. [1999. Nonparametric estimation of triangular simultaneous equation models. Econometrica 67, 565–603]. It is based upon local polynomial regression and marginal integration techniques. We establish the asymptotic distribution of our estimator under weak data dependence conditions. Simulation evidence suggests that our estimator may significantly outperform the estimators of Pinkse [2000. Nonparametric two-step regression estimation when regressors and errors are dependent. Canadian Journal of Statistics 28, 289–300] and Newey and Powell [2003. Instrumental variable estimation of nonparametric models. Econometrica 71, 1565–1578].     

Are there common values in first-price auctions? A tail-index nonparametric test

We develop a consistent nonparametric test of common values in first-price auctions and apply it to British Columbia Timber Sales data. The test is based on the behavior of the CDF of bids near the reserve price. We show that the curvature of the CDF is drastically different under private values (PV) and common values (CV). We then show that the problem of discriminating between PV and CV is equivalent to estimating the lower tail index of the bid distribution. Our approach admits unobserved auction heterogeneity of an arbitrary form. We develop a Hill (1975)-type tail index estimator and find the presence of common values in BC Timber Sales.     

Semi-nonparametric estimation and misspecification testing of diffusion models

Novel transition-based misspecification tests of semiparametric and fully parametric univariate diffusion models based on the estimators developed in [Kristensen, D., 2010. Pseudo-maximum likelihood estimation in two classes of semiparametric diffusion models. Journal of Econometrics 156, 239-259] are proposed. It is demonstrated that transition-based tests in general lack power in detecting certain departures from the null since they integrate out local features of the drift and volatility. As a solution to this, tests that directly compare drift and volatility estimators under the relevant null and alternative are also developed which exhibit better power against local alternatives.     

Implementing optimal procurement auctions with exogenous quality

In this paper, we consider the implementation of the optimal procurement auction outcome when there is a trade-off between price and quality for the buyer. The quality proposals of the competing firms are assumed to be exogenous and common knowledge. After defining the optimal auction, we characterize the discriminatory first score auction and the modified second score auction implementing it when the buyer is assumed to have the same information as the firms. Then we consider an auction which requires no specific information from the buyer and show that a second score auction followed by a negotiation stage can implement the optimal auction outcome under this more realistic assumption. Received: 30 March 2001 / Accepted: 14 December 2001 I am grateful to the anonymous referees for comments on an earlier version of this paper. The usual disclaimer applies.