Introduction
In September 1996, Staples and Office Depot, two of the largest office supply superstores in United States announced their plans to merge together their chains of stores. Office Depot had 571 North American outlets and staples had 553 superstores in 100 U.S markets so their combined operations would exceed 1100 stores and generate over $10 billion in revenue.
The merger announcement between the two intense competitors raised immediate suspicions of dominating market power. Many feared that the nearest competitor Office Max would be left in the dustbin. Supplier, competitors and customers filed comments of concern with the Federal Trade Commission. In response, On March 10, the FTC authorized its staff to seek a federal district court order to prevent Staples from acquiring Office Depot. The FTC argued in court that the Staples/Office Depot merger would violate federal antitrust laws by substantially reducing competition in the retail sale of office supply superstores in various markets throughout the country where each firm directly competes against each other. FTC research determined that there were higher retail prices for office supplies in markets where there was high industry concentration of office supply stores, and lower prices in markets where there was lower industry concentration. Furthermore, in entertained a much narrower definition of the market. Staples and Office Depot claimed that they were in competition not only with traditional office supply stores, but also with other large outlets such as Wal-Mart, and Circuit City who also sold office machines, paper, pens and pencils. In contrast FTC took a different view of the market, as "one-stop shop" office supply stores. In this narrower market definition, the merged company would have 70 percent market share.
In Staples the FTC applied new economic theory and introduced a new type of evidence. [1] The case has been described as the "biggest litigated antitrust case in a decade" [2] and "the most controversial and well-contested merger case in decades." [3] An antitrust expert declared "Had the FTC lost, it would have been in terrible trouble, because in the future many more merging parties would have felt it worth their while to take the FTC on. Instead, merging parties are more likely to be willing to negotiate consent decrees." [4] Crucial to the FTC's case was data showing that in markets where Staples was the only office supply superstore (OSS) firm, prices were higher than in markets where it competed against Office Depot and Office Max. Prices in Staples-only markets were 13% higher than in three-firm markets. [5] Similarly, prices in Office Depot-only markets were 5% higher than in three-firm markets. [6] The FTC's successful challenge of the Staples-Office Depot merger is one of the most important antitrust cases in recent years. One of the _______ describes the case of Staples versus Staples as below:
"The argument about anticompetitive effect in Staples turned on a single overwhelming fact: prices of office supplies could be shown on average to be substantially higher in cities where only one office supply super store chain was located than where two super store chains competed, and even higher than in cities where the three super store chains all faced each other in the market place."
Robert Pitofsky (1997)
In response to FTC arguments, the Staples defendants argued that higher prices in one-firm markets could result from a "host of factors other than superstore competition."11 They estimated a fixed-effects regression model which tried to hold these other factors constant. It indicated that prices would decline following the merger. In addition, Thomas Sternburg, founder and CEO of Staples, said that office products are sold in a fragmented $18 billion market in which the merged firm would have a meager 6 percent market share. Furthermore, with larger operations the new company would achieve efficiencies, eliminate redundancies, and develop greater buying power with their suppliers. In combination these would lead to price reductions and bigger saving for both business and individual customers. Staples and Office Depot placed ads in major newspapers around the United States.
This paper will proceed as follows. In second part we will introduce a brief summary on Office Depot and Staples 'company profiles and their business context as the largest chains in the office supply superstores market. In Section II we will summarize the narrative of the
Staples-Office Depot proposed merger and its legal outcome.2 In Section III we will describe the
UP-SP merger and its outcome.3 And Section IV will offer a brief conclusion
2. Case Background
Prior to 1986, small business and home office customers purchased office supplies primarily through small independent stationers, warehouse clubs and mail order firms. The "office super store" (OSS) retail concept was pioneered by Staples in 1986, as a large volume retail outlet for office supplies and other business-related products that focused on small land medium-sized businesses, home office customers, and individuals. Only the office supply superstores offered "one-stop shopping" providing an inventory, breadth of products and convenience not available elsewhere, including other retailers like BestBuy, Wal-Mart, Target, computer stores and independent stationers. The strategy of OSS involved a wide selection of items (5,000-6,000 in a store) and sharply discounted prices (typically 30%-70% below manufacturers' suggested list prices), based on direct-from-the-manufacturer purchases at substantial discounts. Prices at OSSs were often substantially below the prices for the same items that were being sold in local stationery stores and other outlets. Office Depot quickly followed Staples into the OSS category. This retail concept proved to be a great success. Both chains expanded rapidly. When th merger of the two firms was proposed, Office Depot and Staples were the first and second largest chains of OSSs in the U.S. As of 1997 (when the case went to trial) Office Depot operated over 500 stores in 38 states; it had worldwide revenues in 1996 of $7.3 billion and a 1996 year-end stock market value of $2.2 billion. Staples operated 550 stores in 28 states; it had worldwide revenues in 1996 of $4.5 billion, and a stock-market valuation of approximately $3 billion at year-end 1996.
In the decade following Staples' innovation, 23 other OSS chains attempted to replicate these two chains' success. By late 1996, only Staples, Office Depot and OfficeMax remained strong OSS competitors. Office1Superstore, the lone other OSS chain, was small and on the verge of exiting the market. Although the three remaining OSS rivals each had strong regional positions, they were beginning to expand into each other's territories as the chains were growing rapidly. Staples and Office Depot competed directly in more than forty metropolitan areas.
On September 4, 1996, Staples and Office Depot announced their plan to merge. At that time, each had approximately 500 stores and their combined annual sales exceeded $10 billion. In April 1997, the Federal Trade Commission (FTC) voted to oppose the transaction. In its complaint, the FTC alleged a product market consisting of "consumable office supplies sold through office supply superstores," and identified forty-two individual geographic markets in which it anticipated consumer harm resulting from the merger. Consumable office supplies, defined as those items repeatedly purchased by consumers (and thus quickly consumed), account for only 50% of OSS revenue. The remainder of OSS revenue arises from business services, office furniture and computer sales. While Staples and Office Depot argued that all products sold by their stores should be included in the product market, the data used by both the defendant and the government in the pricing studies described in this essay contained information about consumable office supplies only.
The FTC won a preliminary injunction against the merger in US District Court in June 1997. Subsequently, Staples and Office Depot abandoned the proposed transaction.
http://papers.ssrn.com/sol3/papers.cfm?abstract_id=139625 conclusion
http://www.ftc.gov/opa/1997/04/stapdep.shtm FTC REJECTS PROPOSEd SETTLEMENT IN STAPLES/OFFICE DEPOT MERGER
Staples, Inc. is one of the world's largest office superstore operators with about 780 stores as of mid-1998. The company sells a variety of office supplies, computers and other office machines, and office furniture. The typical Staples store stocks over 8,000 brand-name and private-label office products. It offers these products at a discount price and also guarantees to pay a customer 150 percent of the difference if a competitor sells the product at a lower price.
Staples competes with a variety of discount office supply stores, warehouse clubs, individual stationery stores, mail order chains, and consumer electronics chains. Half of its 30,000 employees work part-time. For its fiscal year that ended on January 31, 1998, Staples had sales of over $5 billion, an increase of 31 percent over the previous year.
The index http://books.google.com/books?id=okm5eTc2ILEC&pg=PA65&dq=staples+FTC+office+depot&cd=3#v=onepage&q=staples%20FTC%20office%20depot&f=false
http://books.google.com/books?id=-YoouhgqxFMC&pg=PA26&dq=staples+FTC+office+depot&cd=7#v=onepage&q=staples%20FTC%20office%20depot&f=false - FTC case
http://books.google.com/books?id=Rz2ty2s6fb8C&pg=PA155&dq=staples+FTC+office+depot&cd=8#v=onepage&q=staples%20FTC%20office%20depot&f=false BEGIN
Introduction
The first part of the paper is dedicated to the review of previous surveys related with liquidity and asset pricing. We will continue in the second part by raising hypothesis and using empirical results of the paper discussed in the previous part in order to validate the findings. Meanwhile, the last part refers to the methodology used to perform the empirical tests.
2. Review of prior research
There is a huge and expanding literature regarding liquidity. The paper will examine the literature that links liquidity with asset pricing. The review will be based in the studies of different researches utilizing their quantitative approaches and results on the relevant topic. Below we provide brief summaries of the main papers within the literature concentrating on both theoretical and empirical findings. We will be focused in different approaches toward the important influence of liquidity on asset pricing.
Amihud & Mendelson (1986, 2002, 2005)
In the paper of Amihud & Mendelson (1986) we find a development of a model that relates positively the returns to the relative spread (bid ask spread divided by price) at a decreasing rate (a concave relationship). The authors empirically test the implications of their theoretical model using a CAPM framework by supporting their result on the usage of the data of spreads and stock returns from the NYSE and AMEX over the period 1961-1980.
In separate regressions they find a linear relationship between the expected return and beta and find the relationship between excess returns and relative spread to be concave. The main result of the paper is that expected asset returns are positively related to the relative bid-ask spread.
In this model, risk neutral investors are assumed to have different holding periods and limited capital. These assumptions introduce a clientele effect into the solution where investors with long holding periods select stocks with high trading costs. The required return will then differ for different classes of investors, and the expected gross return becomes an increasing and concave function of the relative transaction cost.
Summing up, according to Amihud & Mendelson findings:
Asset with higher transaction costs are allocated to agents with longer investments horizons.
The expected return E (r) is an increasing and concave function of transaction costs.
The above imply that the liquid assets are allocated to agents with short investment horizons. These agents earn lower returns since they are the least capable of dealing with illiquidity.
Baks&Kramer (1999)
Another approach toward the significant impact of liquidity on asset prices has been presented in the working paper of Baks&Kramer (1999) regarding the global liquidity and asset prices. The authors assess the empirical effect of global liquidity on asset pricing through cross country spillover. By collecting the data on money and asset returns for different countries (G-7) for the period 1971 - 1998, they try to give an overview of the importance of liquidity for asset returns at an international level. The paper identifies two concepts of liquidity, market liquidity which refers to the ability of financial markets to absorb the variation in demand and supply without dislocation in prices and monetary liquidity which refers to short term interest rates. The model calculates three measures of liquidity: weighted growth rate series (liquidity is weighted by each country's GDP), simple sum US dollar aggregate and Divisia indices of global liquidity which uses narrow and broad money. Based on 3 aggregation methods excess money growth variables are constructed by subtracting the average growth rate of nominal GDP. Also construct ex-country (ROW) indices where one of the G3 countries is excluded. Narrow money is found to have a stronger relationship to asset prices than broad money as do Divisia and simple sum measures.
The main finding of the study is that there is a negative correlation between interest rates and liquidity meaning that an increase in liquidity is consistent with lower interest rates. Moreover, liquidity is positively correlated with stock return meaning that an increase in liquidity is consistent with higher stock returns. Other result was found regarding the spillovers from the volatility of liquidity to the volatility of asset returns across countries. In addition, liquidity may cause inflation if demand increases for a fixed supply of assets and a booming economy may lead to both increased liquidity and to a rise in asset prices.
Kempf & Uhrig-Homburg (2000)
As mention above, when focusing on the stock market, Amihud & Mendelson (1986) show that because investors are rational they prefer liquid stocks that can be traded immediately at lower costs. For this reason, the expected return is higher for illiquid stocks comparing with the expected return of liquid stocks. Kempf& Uhrig-Homburg (2000) present a continuous-time model that investigates the impact of liquidity on bond prices. The benefit of focusing only on bonds gives the possibility of investigating bond market as a risk free market, while the stock market reflects beside liquidity, also the risk. Moreover, the results on the impacts of liquidity will be separated from the ones on the impact of risk.
The model is tested using 1992-1994 data from bonds issued by the German government. The empirical results show that bond prices not only depend on interest rates, but also on the liquidity of bonds. Hence, bond liquidity should be used as an additional pricing factor. The contribution of this paper is in the pricing of illiquid bonds relative to liquid ones. The empirical results suggest about the importance of liquidity in determining the bond prices and that bond liquidity should be used as a pricing factor. We will use the empirical approach of Kempf& Uhrig-Homburg (2000) to perform a joint hypothesis testing about the prices of liquid and illiquid bonds.
Acharya&Pedersen ( 2004)
The paper of Acharya&Pedersen (2004) introduces a theoretical framework of a liquidity - adjusted capital asset pricing model. The model shows that the required return of an asset depends on its liquidity and also on the covariance of its own returns and liquidity with market return and market liquidity. More specifically, the expected return of the asset increases with the increase of illiquidity. The paper explores cross sectional predictions using AMEX and NYSE stocks over a period of 30 years. The main finding of the research is that the liquidity - adjusted capital asset pricing model stands better than the standard capital asset pricing model. The results are proven in terms of squared-R (correlation coefficient) for cross sectional returns and p-values in specification tests (showing higher significance). Furthermore, the paper presents not only how prices are affected by liquidity but also the commonality of liquidity.
Bekaert, Campbell & Lundblad (2007)
The paper examines the influence of liquidity on expected returns in emerging equity markets where liquidity appears to be a significant factor. Bekaert, Campbell & Lundblad (2007) proposes a measure of liquidity which is positively correlated with bid-ask spreads and negatively correlated with equity market turnover. The main finding of the paper is that unexpected liquidity shocks are positively correlated with returns and negatively correlated with dividend yields. The empirical results suggest that local market liquidity is a significant driver of expected returns in emerging markets, and that the liberalization process has not eliminated its impact. The authors also propose a pricing model that distinguishes the ways through which liquidity can affect expected returns. One refers to the transaction costs and the other to the liquidity as a systematic risk factor.
3. Hypothesis - Research questions
In this part we will draw some hypothesis and will try to prove their validity by using previous empirical tests. The first hypothesis will give us the answers about the relationship between liquidity and asset prices and its importance as an essential determinant in the asset pricing. The second hypothesis is not of less interest. As discussed above, bond liquidity is examined separately because it gives us the opportunity to study the liquidity in a free risk market - bond market. We consider as important to test the relation between liquidity and illiquidity price bonds.
The first hypothesis sheds light on the central question of the topic:
Does liquidity explain the asset prices? Does it make a significant factor that should be included when determining asset pricing?
If we think of the relationship between liquidity and asset prices, our hint tells us that they are related negatively with each other. In order to verify our hint we will use the empirical evidence of Amihud & Mendelson (1986). This survey develops an empirical model to test whether liquidity influences asset pricing. We can use these results to prove our hypothesis and draw the final conclusion.
In the first part of the model, the authors focus in the relationship between illiquidity and return which is determined to be increasing and concave. As a result, the return increases by less and less as the asset becomes more liquid. The literature presents different possible relationship between liquidity and asset prices. First, an increase in liquidity is followed by an increase in asset prices. Second, liquidity can increase the inflation in asset prices if excess liquidity causes an increase in the demand for a fixed supply of assets. Third, the decline in interest rates caused by an increase in liquidity leads to an increase in equity prices by reducing the discount factors that price future cash flows. The results show a link between expected return and bid - ask spread and also between expected return and turnover (reflecting holding period distances). Since expected returns are a decreasing function of liquidity, investors must be compensated for higher trading costs that they bear in less liquid markets. Amihud & Mendelson find a significantly positive relation between expected returns and the bid-ask spread for NYSE and AMEX stocks during 1961-80 periods. The average portfolio risk-adjusted returns increase when the bid-ask spread increases. In the same way, Amihud (2002) argues that expected stock returns partly represent an illiquidity premium and it also shows that stock returns are an increasing function of illiquidity.
We can sort many effects of liquidity in financial markets. Based on the empirical paper of Amihud & Mendelson (1986) liquidity explains the cross section of assets with different liquidity after taking in consideration the risk and the correlation between liquidity and asset returns. Liquidity explains why some assets that are difficult to be traded have lower prices and also it explains the price of bonds and stocks. Moreover, liquidity explains the risk free puzzle (the lower the risk the lower the required return of the asset), the equity premium puzzle (equity require higher return) and the small firm effect (small illiquid firms have higher returns).
Summing up, based on the above we can draw some final results. Asset returns depend on liquidity and their relationship appears to be increasing and concave which implies that the return is higher if the asset is less liquid. Liquidity is a significant factor that determines asset pricing and it is very important to include it as a pricing factor. Finally, we conclude that there is positive liquidity premium on stock returns.
Another test of great interest and importance would be the below one:
Are liquidity bond prices equal to illiquidity ones?
We will consider as null hypothesis "Ho: liquidity bond prices equal to illiquidity bond prices" and as alternative the hypothesis "Ha: liquidity bond prices are higher than illiquidity ones". Failure to reject the null hypothesis which defines liquid and illiquid bonds having the same price proves our null hypothesis. In order to validate this we will use the empirical results of Kempf& Uhrig-Homburg (2000). The paper tests the joint hypothesis by using a t statistical test. In addition, it uses only liquid bonds to estimate the term structure of interest rates that explain observed prices of liquid coupon bonds with a very low average absolute pricing error. Meanwhile, the discount factors are found using a quadratic linear programming approach to determine a discrete discount function. The results confirm a very significant t value which supports the alternative hypothesis, thus the null hypothesis is rejected. Concluding, the illiquid bonds have a higher price than liquid bonds.
4. Methodology
Amihud & Mendelson
The model Amihud & Mendelson (1986) performs a cross section analysis on AMEX and NYSE stock during 1960 - 1980. In order to test the relationship between the illiquidity and return they divide the stocks by their bid - ask spread, which is a measure of their illiquidity. Than, the authors perform a cross sectional estimation of the average return on each portfolio as a function of the bid-ask spread as well as of firm size and the unsystematic volatility. The result showed that the average portfolio return was significantly higher for stocks with higher spread. Furthermore, the function was increasing and concave, as predicted by the model.
The empirical estimation model Amihud (2002) employs a measure of illiquidity that calculates it from daily returns and volume of AMEX and NYSE stocks for the period 1964-1999. Stocks are sorted every year by their illiquidity and grouped in portfolios. The monthly illiquidity for each portfolio is calculated as the residual from the autoregressive of the portfolio illiquidity.
Baks&Kramer (1999)
The model presented uses principal components analysis and first principal component which has high explanatory power for all individual country growth rates. There are estimations of several relationships, from simple correlations to regressions and test of Granger causality. The results indicated negative correlation with interest rates and positive correlation with stock returns. The authors run different regressions to check for the direct effects in asset returns and inflation. For cross-country monetary spillovers they use ROW indices and GARCH model. The evidence appears consistent with a "push" channel rather than the "pull" channel. And in order to test the indirect effects they use Granger causality tests.
Kempf& Uhrig-Homburg (2000)
Kempf& Uhrig-Homburg (2000) introduces a theoretical continuous-time valuation framework that analyzes the influence of liquidity on bond prices. In the study are used 1992-1994 data from bonds issued by the German government. These bonds represent a market segment that is homogeneous in bankruptcy risk, taxes, age, and coupons, but the bonds differ with respect to their liquidity. The model prices fixed income securities within a two-factor Cox/Ingersoll/Ross-type model [7] which refer to the factors are the short-term interest rate and a second exogenous factor that accounts for the liquidity of assets. The survey includes two stochastic factors, the instantaneous risk free rate of a liquid investment and an unspecified factor that accounts for the price differences between liquid and illiquid bonds. Moreover, in order to support the proposed model by empirical validity, the paper uses a two-step approach: in - sample test and out-of sample test.
Acharya&Pedersen ( 2007)
Acharya&Pedersen (2007) studies a broad model of pricing liquidity risk and liquidity level. The authors put the expected return over risk rate as a function of illiquidity and estimate four systematic risk variables. The model identifies a positive market liquidity premium, when it adopts Amihud's (2002) market liquidity measure. Furthermore, in the paper are performed many tests of the liquidity-adjusted CAPM. The Wald test and p-value are estimated, and their values are of a great significance which determines that the liquidity-adjusted CAPM explains the data better than the standard CAPM.
Bekaert, Campbell & Lundblad (2007)
Bekaert, Campbell & Lundblad (2007) studies the time series liquidity-expected return relationship using data from 19 emerging markets and US that proxies the world index. The dynamics of returns and liquidity are analyzed by using various vector autoregressions (VAR ). The model estimates liquidity by using measures that rely on the incidence of observed zero daily returns in these markets. The advantage of this measure is that it requires only a time series of daily equity returns. In addition, they impose cross-country restrictions on the parameter space when examining the dynamics of expected returns and liquidity. The paper introduces three measures of liquidity, one that refers to the proportion of zero daily returns which is highly correlated with more traditional measures of transaction costs and the second includes information about the length of the non-trading interval.
