Maximizing sellers’ welfare in online auction by simulating bidders’ proxy bidding agents

This paper presents an agent-based simulation model to estimate the effects of auction parameters on the auction outcomes, and to find an optimal or, at least, a near-optimal infrastructure. In particular, this study intends to study how bidders’ personalities and bidding strategies with other auction parameters affect the closing prices in two auction mechanisms: English and Yankee auction. Experimental results show that the aggregated outcome of multiple English auctions is more favorable to auctioneers and sellers than that of a corresponding single Yankee auction. It is also shown that raising minimum bid increment or bid starting price positively affects the closing prices in both auction types. However, two auction systems respond differently to the changes in parameter values in terms of magnitude and robustness. Further, it is noted that the closing price of an auction is negatively related to the proportion of risk-takers who adopt sniping strategy to submit their bids. As the proportion of risk-takers increases, the closing price significantly decreases in hard-ending auctions (eBay-like system), but only marginally in soft-ending auctions (Amazon-like system). In particular, soft-ending auctions with less snipers and a longer time window of automatic extension return the most favorable closing price to sellers.     

Visualizing combinatorial auctions

We propose a novel scheme to visualize combinatorial auctions; auctions that involve the simultaneous sale of multiple items. Buyers bid on complementary sets of items, or bundles, where the utility of securing all the items in the bundle is more than the sum of the utility of the individual items. Our visualizations use concentric rings divided into arcs to visualize the bundles in an auction. The arcs’ positions and overlaps allow viewers to identify and follow bidding strategies. Properties of color, texture, and motion are used to represent different attributes of the auction, including active bundles, prices bid for each bundle, winning bids, and bidders’ interests. Keyframe animations are used to show changes in an auction over time. We demonstrate our visualization technique on a standard testbed dataset generated by researchers to evaluate combinatorial auction bid strategies, and on recent Federal Communications Commission (FCC) auctions designed to allocate wireless spectrum licenses to cell phone service providers.     

A secure multi-item e-auction mechanism with bid privacy

The recent focus within the auction field has been multi-item auctions where bidders are not restricted to buying only one item of the merchandise. It has been of practical importance in Internet auction sites and has been widely executed by them. In this paper, we concentrate on the use of the multi-item auction for task assignment scenarios and propose a novel PUPA auction protocol to solve the problem of bid privacy in multi-item auctions. A verifiable technique of shared key chain is proposed to find the winners without revealing the losing bid and bidder’s privacy. It can be shown that our new scheme is robust against cheating bidders.     

Optimal reserve prices in weighted GSP auctions

Most search engines use the weighted Generalized Second Price (wGSP) auction to sell keyword-based text ads, generating billions of dollars of advertising revenue every year. Designing and implementing near-optimal reserve prices for these wGSP auctions are naturally important problems for both academia and industry.In this paper, we show how to calculate and implement the near-optimal reserve price of the wGSP mechanism in realistic settings. Unlike reserve prices in standard single-item auctions, optimal reserve prices in wGSP auctions are discriminatory, different even for advertisers bidding on the same keyword. The optimal reserve price results can be extended to support CPA/CPC/CPM¹ hybrid auctions.Our simulations indicate that setting a proper reserve price will transfer some bidder utility (payoff) to auctioneer utility, resulting in higher revenue for the search engine. We describe a practical methodology to implement optimal reserve prices in production systems.     

A secure reverse Vickrey auction scheme with bid privacy

In recent years auctions have become more and more important in the field of multi-agent systems as useful mechanisms for resource allocations, task assignments and electronic commerce. In this paper, we concentrate on the use of the reverse Vickrey auction for task assignment scenarios and propose a novel RVP auction protocol as a method to solve problems to bid privacy in reverse Vickrey auctions. A verifiable technique of encryption key chain is used to find the second lowest bid without revealing the losing bid and unnecessary information. Through analysis, it is verified that our new scheme is robust against cheating bidders.     

Real time prediction of closing price and duration of B2B reverse auctions

Nowadays, online auctions have become the most successful business model in the electronic marketplace. To the best of the authors’ knowledge, no other work has been devoted to the prediction of closing price and duration of Business-to-Business (B2B) English reverse online auctions in which goods or service providers compete with each other to win contracts by lowering offering prices with each bid, which is conducted on a virtual platform hosted on the Internet. This research designs and proposes a new methodology to predict closing prices and duration within the first few bids of the corresponding auctions based on real time bidding information rather than static auction information. In this article, we employ real time information and prediction rules to forecast the behavior of live auctions. This is in contrast to the static prediction approach that takes into consideration only information available at the beginning of an auction such as products, item features, or the seller’s reputation. This simulation is based on discretized auction data derived from a B2B online auction marketplace over a two-year period. Three measurements including accuracy, coverage, and benefit are used to evaluate the methodology. Results show that after observing the first 4 bids, this methodology can predict closing prices and duration with 84.6 and 71.9% accuracy, respectively.     

Improving the Performance of Online Auctions Through Server-side Activity-based Caching

Online auction sites have very specific workloads and user behavior characteristics. Previous studies on workload characterization conducted by the authors showed that (1) bidding activity on auctions increases considerably after 90% of an auction’s life time has elapsed, (2) a very large percentage of auctions have a relatively low number of bids and bidders and a very small percentage of auctions have a high number of bids and bidders, (3) prices rise very fast after an auction has lasted more than 90% of its life time. Thus, if bidders are not able to successfully bid at the very last moments of an auction because of site overload, the final price may not be as high as it could be and sellers, and consequently the auction site, may lose revenue. In this paper, we propose server-side caching strategies in which cache placement and replacement policies are based on auction-related parameters such as number of bids placed or percent remaining time till closing time. A main-memory auction cache at the application server can be used to reduce accesses to the back-end database server. Trace-based simulations were used to evaluate these caching strategies in terms of cache hit ratio and cache efficiency. The performance characteristics of the best policies were then evaluated through experiments conducted on a benchmark online auction system.     

Implications of a Reserve Price in an Agent-Based Common-Value Auction

Auction sellers can use a reserve price to require a minimum bid before items are sold. Theoretical and experimental research has tested the influence of a reserve price in an independent private values auction, but little focus has been given to the influence of a reserve price in a first-price common-value auction. We establish an agent-based first-price common-value auction to determine the impact of the reserve price with two buyers and with three buyers. An agent-based approach to this problem is both a unique contribution to the literature and appropriate since finding analytical solutions with common-value auctions is difficult. The agent-based model approach also allows us to consider buyers that have non-symmetric bid functions. Furthermore, we introduce a combination of numerical integration techniques with a new particle swarm learning algorithm. The buyers in the model choose their expected-net-revenue-maximizing bid price, and sellers choose their expected-revenue-maximizing reserve price. In the two-buyer and three-buyer auction, a reserve price increases the equilibrium winning bid price, decreases the probability that the item is sold, and increases the seller’s expected revenue. A reserve price in a two-buyer auction increases the winning bid price more than including an additional buyer in the auction with no reserve price. However, due to only receiving a salvage value when the item does not sell in the auction, the seller’s expected revenue is higher in the three-buyer first-price common-value auction without a reserve price than in the two-buyer auction with a reserve price.     

Designing Electronic Auctions: An Internet-Based Hybrid Procedure Combining Aspects of Negotiations and Auctions

We discuss design issues pertaining to multiple issue auctions in the WWW environment. Based on a critical evaluation of existing auctions, we propose NegotiAuction^tm, an algorithmic Internet-based auction procedure, which combines certain elements of negotiations and auctions. It can be used either in reverse or forward auctions. When defining a multiple issue (multiple attribute) auction, the auction owner has more control over the bidding process than is possible in traditional auctions, signaling bid requirements to the bidders individually. This will result in a preferred set of auction winners. It is believed to reduce the total transaction time and eliminate the necessity of holding subsequent negotiations with the set of winners.     

MoRVAM: A reverse Vickrey auction system for mobile commerce

Internet auctions bring buyers and sellers together for the purpose of trading goods and services online. In order to get the goods, a buyer must search for items through several auction sites. When the auction starts, the buyer needs to connect to these auction sites frequently so that he/she can monitor the bid states and re-bid. In this paper, we propose an automated negotiation model between two participants, for mobile commerce, using collaborative mobile agents called MoRVAM, which mediates between the buyer and the sellers, and executes bidding asynchronously and autonomously. A new RVT protocol is also implemented to achieve unconditional bid privacy. Advantages of the RVT protocol are addressed as well. All the bidding process can be implemented without revealing losing bid and unnecessary information.     

Distributed ascending proxy auction — A cryptographic approach

In recent years, auctions have become a very popular price discovery mechanism in the Internet. The common auction formats are typically centralized in nature. The peer-to-peer paradigm demands gearing up auctions for decentralized infrastructures. In this context, this paper proposes a distributed mechanism for ascending second-price auctions that relies on standard cryptographic algorithms. In essence, the auction protocol has the capability of preserving the privacy of the winning bidder’s true valuation. The auction protocol makes use of a high number of auctioneers divided into several groups. A bidder creates an encrypted chain of monotonously increasing bidding steps, where each bidding step can be decrypted by a different auctioneer group. This considerably reduces the attack and manipulation possibilities of malicious auctioneers. In addition, this secure approach does not require bidders to be online unless they are submitting their bid chain to the auctioneers.     

Simultaneous Independent Online Auctions with Discrete Bid Increments

Decentralized multi-item auctions offer great opportunities for integrating fragmented online auction markets into larger markets with more efficient outcomes. This paper extends the theory of multi-item ascending auctions of substitutes by considering any finite positive bid increment and allowing the bidders to bid asyn-chronously instead of bidding in a round-robin fashion. We consider a setup where the bidders utilities over multiple items are additive and bound the maximum inefficiency in the allocation when the bidders follow a simple greedy strategy. We also obtain the limits within which the prices of individual items can vary from one outcome to another. For the special case of single unit bidder demand, we also bound the maximum surplus which a bidder can extract by unilaterally switching to some other strategy. The paper suggests an upper bound for the minimum required bid increment which would be necessary for competitive price discovery and truthful bidding in a practical online implementation.The work was done while the author was with IBM Research.     

Calibration based on entropy minimization for a class of asset pricing models

We consider the problem of calibrating pricing models based on the binomial tree method to market data in a network of auctions where agents are supposed to maximize a given utility function. The calibration is carried out using the minimum entropy principle to find a probability distribution that minimizes a weighted misfit between predicted and observed data. Numerical results from calibrating the mid prices from the bid–ask pairs of the buyer and seller to Taobao data demonstrated the feasibility of this approach in the case of pricing goods in a sequential auction. Further numerical test cases have been presented and have shown promising results. This work can equip those engaged in electronic trading with computational tools to improve their decision-making process in an uncertain environment.     

Robust combinatorial auction protocol against false-name bids

This paper presents a new combinatorial auction protocol that is robust against false-name bids. Internet auctions have become an integral part of Electronic Commerce (EC) and a promising field for applying agent and Artificial Intelligence technologies. Although the Internet provides an excellent infrastructure for combinatorial auctions, we must consider the possibility of a new type of cheating, i.e., an agent tries to profit from submitting several bids under fictitious names (false-name bids). If there exists no false-name bid, the Generalized Vickrey Auction protocol (GVA) satisfies individual rationality, Pareto efficiency, and incentive compatibility. On the other hand, when false-name bids are possible, it is theoretically impossible for a combinatorial auction protocol to simultaneously satisfy these three properties.

Our newly developed Leveled Division Set (LDS) protocol, which is a modification of the GVA, utilizes reservation prices of auctioned goods for making decisions on whether to sell goods in a bundle or separately. The LDS protocol satisfies individual rationality and incentive compatibility even if agents can submit false-name bids, although it is not guaranteed to achieve a Pareto efficient social surplus. Simulation results show that the LDS protocol can achieve a better social surplus than that for a protocol that always sells goods in one bundle.     

A graphical formalism for mixed multi-unit combinatorial auctions

Mixed multi-unit combinatorial auctions are auctions that allow participants to bid for bundles of goods to buy, for bundles of goods to sell, and for transformations of goods. The intuitive meaning of a bid for a transformation is that the bidder is offering to produce a set of output goods after having received a set of input goods. To solve such an auction the auctioneer has to choose a set of bids to accept and decide on a sequence in which to implement the associated transformations. Mixed auctions can potentially be employed for the automated assembly of supply chains of agents. However, mixed auctions can be effectively applied only if we can also ensure their computational feasibility without jeopardising optimality. To this end, we propose a graphical formalism, based on Petri nets, that facilitates the compact represention of both the search space and the solutions associated with the winner determination problem for mixed auctions. This approach allows us to dramatically reduce the number of decision variables required for solving a broad class of mixed auction winner determination problems. An additional major benefit of our graphical formalism is that it provides new ways to formally analyse the structural and behavioural properties of mixed auctions.     

Optimal Lot Sizing Policies For Sequential Online Auctions

This study proposes methods for determining the optimal lot sizes for sequential auctions that are conducted to sell sizable quantities of an item. These auctions are fairly common in business to consumer (B2C) auctions. In these auctions, the tradeoff for the auctioneer is between the alacrity with which funds are received, and the amount of funds collected by the faster clearing of inventory using larger lot sizes. Observed bids in these auctions impact the auctioneer's decision on lot sizes in future auctions. We first present a goal programming approach for estimating the bid distribution for the bidder population from the observed bids, readily available in these auctions. We then develop models to compute optimal lot sizes for both stationary and non-stationary bid distributions. For stationary bid distribution, we present closed form solutions and structural results. Our findings show that the optimal lot size increases with inventory holding costs and number of bidders. Our model for non-stationary bid distribution captures the inter-auction dynamics such as the number of bidders, their bids, past winning bids, and lot size. We use simulated data to test the robustness of our model.     

Probabilistic Automated Bidding in Multiple Auctions

This paper presents an approach to develop bidding agents that participate in multiple auctions with the goal of obtaining an item with a given probability. The approach consists of a prediction method and a planning algorithm. The prediction method exploits the history of past auctions to compute probability functions capturing the belief that a bid of a given price may win a given auction. The planning algorithm computes a price and a set of compatible auctions, such that by sequentially bidding this price in each of the auctions, the agent can obtain the item with the desired probability. Experiments show that the approach increases the payoff of their users and the welfare of the market.     

Competitive analysis of incentive compatible on-line auctions

This paper studies auctions in a setting where the different bidders arrive at different times and the auction mechanism is required to make decisions about each bid as it is received. Such settings occur in computerized auctions of computational resources as well as in other settings. We call such auctions, on-line auctions.

We first characterize exactly on-line auctions that are incentive compatible, i.e. where rational bidders are always motivated to bid their true valuation. We then embark on a competitive worst-case analysis of incentive compatible on-line auctions. We obtain several results, the cleanest of which is an incentive compatible on-line auction for a large number of identical items. This auction has an optimal competitive ratio, both in terms of seller's revenue and in terms of the total social efficiency obtained.     

Online auctions enabling the secondary computer market

The Internet is enabling new forms of commerce and novel markets. One example is the secondary computer market, facilitating exchange between quality sensitive sellers, oftentimes businesses, and price sensitive buyers. As this market does not have a viable physical counterpart with reference prices, it is developing via online auctions. One question of interest in the evolution of this market is the determinants of prices. Using a dataset of 2,000 laptop auctions in a seven-month period, this research provides support for accepted auction theory while raising questions that deserve further explanation. The negative relationship between supply and auction price supports standard supply and demand theory, while higher prices for better features is consistent with vertical differentiation. Even within accepted theory this research broadens the understanding of auction behavior. There is clear support for the “price decline anomaly” where prices in sequential auctions decline, violating the “law of one price.” One result that deserves further attention is that midweek auctions realize higher prices. A second is that price changes over time are not monotonic. Future research should replicate and explain these results, as well as extend them to other auction settings. As the secondary computer market evolves it will impact the primary computer market.