Sources of Value in the Formation of U



Merger for Monopoly:

The Formation of the United States Steel Corporation

Preliminary and not for quotation.

Charles Reback

Clemson University

Clemson, SC 29634

chuckreback@

July 10, 2005

The author would like to thank Mike Maloney, Skip Sauer, Matt Lindsey and the participants in Clemson’s Doctoral Dissertation Seminar, the participants of the Southern Economics Association annual meeting in November 2001 in Tampa, and the participants of the Southwestern Economics Association Annual Meeting in March 2001 in Fort Worth. All of the errors are those of the author.

I. Introduction

In 1901 the United States Steel Corporation (US Steel) became the first billion-dollar corporation in the United States, controlling 60% of the nation’s primary steel capacity. It represented the culmination of the first great merger wave and the height of J. P. Morgan’s influence over the American economy. US Steel was capitalized at $1.4 billion, at a time when the total capitalization of American manufacturing was $9 billion. This $1.4 billion was larger than the national debt and over three times the size of the federal budget.

It has been well established by Parsons and Ray (1973), McCraw and Reinhardt (1989), Mullin, Mullin and Mullin (1995) and others that in the early part of the twentieth century US Steel exerted market power over the steel industry. However, the question remains whether it was formed with the expectation that it would become a monopolist. This is the first paper since Stigler (1968) to specifically address this issue and I find evidence to support the contention that US Steel was formed with the expectation that it would exert pricing power in the steel industry.

Thee are three potential reasons for the formation of US Steel:

• Swindle the public

• Monopoly

• Efficiency

The swindle hypothesis posited that US Steel was formed in order to sell overpriced securities to an unsuspecting public. Contemporary accounts emphasize US Steel’s $1.4 billion capitalization compared with the $700 million capitalization of its component firms. Stigler (1968) reports that A. S. Dewing (1941) made the most influential statement on this hypothesis in the 4th edition of his Financial Policy of Corporations. Dewing compares the merger movement to a “virus,” continuing

…It was the harvest-time of promoters… During 1900 and 1901, the movement continued, but the new promotions were fewer in number, owing to the fact that most of the opportunities for the formation of ‘trusts’ had already been fully exploited by bankers and promoters. Accordingly, the ground was combed over again. The trusts themselves were consolidated. A pyramid was built of pyramids. The United States Steel Corporation, capitalized at over 1300 millions of dollars, was built up out of half a dozen smaller ‘trusts,’ themselves, in several cases, the combination of smaller combinations. By 1902 signs were apparent that many of the trusts had not justified the predictions of their promoters.[1]

This write up from $700 million to $1.4 billion in capitalization was used as a “classic example of watered stock” in finance texts.[2],[3] Under this hypothesis, there was no business or economic reason for the steel firms to consolidate. The sole purpose of the consolidation was to sell overpriced stock to an unsuspecting public.

Stigler (1968) advocated the second hypothesis, monopoly, postulating that US Steel was “…formed for the monopoly power that it achieved.”[4] Under this hypothesis, US Steel became the dominant firm in the American steel industry and was able to increase its profits by reducing output and raising prices. Parsons and Ray (1973) and Mullin, Mullin and Mullin (1995) confirm Stigler’s conclusions that US Steel was a monopolist but shed little light on the reasons behind its formation.

Burton (1985) advocated the efficiency hypothesis in an unpublished dissertation. However, he did not reject the possibility that US Steel was more efficient and a monopoly.

In theory, each of these hypotheses has its own testable implications and can be disproved through an event study. This paper attempts to resolve the issue of why US Steel was formed by examining the stock price reactions of several constituencies to the US Steel consolidation. I conclude that US Steel was formed in order to dominate the primary steel market.

The results invalidate the swindle hypothesis. Most of the wealth increase resulting from the consolidation accrued to the stockholders of the component and competitor steel companies, not to the organizers and promoters of US Steel. My empirical results are consistent with the monopoly hypothesis, although do not allow for a complete rejection of the efficiency hypothesis. The stock prices of the component firms reacted positively to the formation, as did the stock prices of US Steel’s future competitors. These are the results predicted by Stigler’s Dominant Firm model. The securities of the customer firms exhibited no statistically significant reactions.

Also, my findings that there was no insider trading prior to the announcement of the formation of US Steel are consistent with Banerjee and Eckard’s (2001) conclusion that insider trading was no more prevalent during the mergers of the First Great Merger Wave (1897-1903) than it is in the modern era. However, my conclusion that US Steel was formed for market power reasons contradicts their earlier (1998) conclusion that mergers during the First Great Merger Wave were mainly a result of firms’ desires for increases in efficiency. While one this one case study cannot refute their study of a large number of mergers, it does suggest that their results may merit a closer look.

II. Theory

The formation of US Steel seems to be an ideal candidate for an event study, an examination of the reactions of securities to new information. Given the assumption that markets react rationally and that prices incorporate new information rapidly, then any change in value that cannot be explained by changes in overall market conditions reflects new information. It should be possible to answer the question of why US Steel was formed by examining the stock price reactions of the various US Steel constituencies. I place the firms to be analyzed into four portfolios: components, competitors, non-railroad customers, and railroad customers. The components, as the name implies, are those companies who comprised US Steel. The competitors are companies engaged in primary steel production, not finished goods, which did not become part of US Steel.[5] The non-railroad customers are steel companies that produced finished goods who, likewise, did not become part of US Steel. The railroad customers are a sampling of some of the leading railroads of the day. At the time, railroads were the largest consumers of steel, and steel was a significant portion of their costs.

Under the first hypothesis, the merger will have no impact on any firms that compete with or are customers of the US Steel component firms. With no economic basis for the merger, the component firms’ expected future cash flows do not change. Since there is no change in the economic fundamentals of the component firms, then the economic fundamentals of competitors and customers will not change. With no change in the underlying economic characteristics of any of the firms, there is no new information conveyed and there should be no stock price reaction to the announcement of the consolidation. This hypothesis is almost laughable but should be formally discussed. To think that J.P. Morgan or any other party could offer a security at a price exceeding its value assumes investor irrationality, questions the entire underpinnings of modern financial theory, and invalidates the entire literature of event studies.

If the second hypothesis, monopoly, is correct, then the consolidation will affect other firms. The component firms become part of a dominant firm that can set prices and quantity in order to generate higher profits. As fringe firms, the competitors can free ride on US Steel’s pricing power and earn abnormal profits. In fact, these firms should be more profitable than US Steel, since they do not have to enforce the restrictions on industry output. Therefore, under the second hypothesis both the component firms and the competitor firms will exhibit positive reactions to the announcement of the consolidation, with the competitors having larger positive abnormal returns than the US Steel component firms since the consolidation would benefit the competitive fringe more than the dominant US Steel. Customers of a monopolistic industry will be negatively impacted since their input costs will rise. Therefore, the customers’ stock price will decline.

If the third hypothesis, efficiency, is correct then the announcement of the consolidation should also affect firms other than those directly involved in the merger. The component firms will be part of a more efficient entity, with reduced costs and increased profits. However, the competitor firms will be hurt by having to face a more efficient US Steel. So, under the efficiency hypothesis, the component firms should react positively and the competitor firms should react negatively. The efficiency hypothesis also prescribes a positive reaction by the customer firms. A more efficient US Steel will lower steel prices and its customers will benefit by having lower input prices for their goods. These effects of these three hypotheses are summarized in Table 1.

I should emphasize that the efficiency and monopoly hypotheses are not mutually exclusive. There is no reason that US Steel could not have been formed in order to become a more efficient monopolist.

III. Previous Analyses

Stigler (1968) rejected Dewing’s hypothesis by examining ex post returns on the common stock of US Steel and several other steel firms. He found that US Steel outperformed the other steel firms by a wide margin from 1901-1924. From this he concludes that the organizers of US Steel did not exploit the stockholders, and that US Steel was formed to exploit monopoly powers. Stigler did not explicitly consider the possibility that US Steel was a more efficient firm but was not a monopoly. It would have been more appropriate had Stigler examined the ex ante reactions of the firms in order to judge their expectations of future profits given this change in the market structure of the steel industry. By examining the stock returns over a long period of time, Stigler implicitly addressed the issue of what US Steel became rather than why was it formed.

Parsons and Ray (1975) found that the formation of US Steel created a non-competitive market and allowed the industry to capture monopoly profits. However, unlike Stigler, they explicitly reject the efficiency argument, basing their conclusions on an examination of the structure of the steel market in the years following the consolidation. They find evidence of increases in American steel prices relative to world prices and price discrimination in exports of American steel, and conclude that US Steel had monopolistic power. Again, whether or not these facts were true after the fact, were these conditions expected at the time of the consolidation?

Burton (1985) also examined post-merger performance of the companies and concurred with Stigler that the merger did not exploit the stockholders. However, Burton rejects Stigler’s dominant firm model in the case of US Steel. He concludes that US Steel formed a more efficient firm, but was not necessarily a monopoly. He based this conclusion on his discovery that while stock prices of US Steel components rose, the stock prices of competitors fell. If US Steel had been a monopoly, the competitive fringe firms would be more profitable than US Steel. Burton concluded that the impetus for the consolidation was efficiency, and that US Steel was a more efficient firm. Examining post-merger trends and activities of US Steel and of the steel industry, he found that US Steel was more profitable than its competitors, and in the contractions of 1904, 1908, and 1911, US Steel’s market share declined by less than its competitors. These observations are consistent with efficiency but not with Stigler’s dominant firm model. However, Burton could not completely discount some degree of monopolization. [Why?] Unlike this study Burton found that the stock prices of the competitor firms reacted negatively to the announcement of the consolidation. However, Burton did not perform a traditional event study, but examined the change in stock prices over a multi-year period. Information regarding the effect of US Steel on the structure of the steel market could have been lost amid other economic changes during that period.

Mullin, Mullin and Mullin (1995) examined stock price reactions to the US Steel dissolution suit, which was active from 1911-1920. In some ways, their paper is the mirror image of this one. In United States v. U.S. Steel, the federal government sued US Steel for being a monopoly in violation of antitrust laws. They examine stock price reactions of various constituencies, and based on those constituencies’ reactions to the progress and setbacks of the suit, conclude that US Steel was a monopoly, and that if US Steel had been dissolved, steel prices would have declined and output increased. Had US Steel simply been a more efficient competitor then a dissolution of this more-efficient firm would have led to higher prices and lower output. Uniquely, their analysis utilizes the stock price reactions of railroads, which were major consumers of steel, and for whom steel costs were a significant cost component. At the time of the suit, the railroads operated under a regulated rate regime and could not pass increased costs onto customers. The railroads reacted positively to progress in the suit and negatively to setbacks in the suit.

IV. Background and the Merger

At the turn of the century, the steel industry could be categorized into two broad groups. Some firms (“primary goods producers”) produced primary products such as steel bars and pig iron. The most important of these firms were Carnegie Steel, Federal Steel, National Steel, and Republic Steel and Iron. Other firms (“finished goods producers”) purchased these primary goods in order to produce finished products such as plate, wire nails, tubes, rails, hoops, etc. Many of these finished goods producers were the result of horizontal mergers beginning in the later 1890’s resulting in monopolies in their narrow niches.[6] While most firms specialized in one of these areas, some firms spanned both the primary and finished goods markets. In the summer of 1900, a slump in profits caused the finished goods firms to consider vertically integrating backwards; and the primary goods firms, most notably Carnegie, began to consider vertically integrating forward [Connaught Mill]. J.P. Morgan, who controlled some of the finished goods firms, along with Federal Steel, the second largest primary steel manufacturer, was not happy about this alteration in the balance of power. Morgan did not believe in unbridled competition, preferring cartel or monopoly behavior with him at the center.

In 1898, J. P. Morgan and Elbert Gary formed Federal Steel, which by 1900 was the second largest steel company in the United States, behind only Carnegie Steel. Carnegie Steel began to integrate vertically and Andrew Carnegie’s plans included competing with National Tube, another Morgan company. Despising competition, Morgan berated Carnegie “as someone who would ‘demoralize’ the industry with price cuts rather than do the smart, gentlemanly thing: join a cartel.”[7]

On December 12, 1900 Morgan attended a dinner at Manhattan’s University Club honoring Charles Schwab, Andrew Carnegie’s chief lieutenant, who was also the principal speaker.[8] Schwab presented a vision of a steel industry dominated by a vertically integrated firm that would effectively eliminate competition. Supposedly, Morgan was completely enthralled, so much so that he forgot to light his cigar.[9] Schwab may have been speaking on behalf of Carnegie, but to this day it is uncertain whether he was operating with Carnegie or behind his back.

Morgan and Schwab immediately began working on the consolidation that would become US Steel. The combination included Carnegie Steel and the Morgan controlled firms of Federal Steel and National Tube, along with National Steel, American Steel and Wire, American Sheet Steel, American Hoop Steel, American Tinplate, and American Bridge Company. See table 2 for listing and description of these firms. On February 1, 1901 rumors of a gigantic combination in the steel industry began appearing in newspapers, and by April 4, 1901 the combination was essentially complete. Most of the firms were acquired in stock swaps with US Steel, although Carnegie, who would have no management role in the new firm, exchanged his holdings in Carnegie Steel for $300 million in US Steel bonds. [Expand this discussion for publication]

V. Data and Analytical Procedures

The age of the event and the lack of regulation in the economy and securities markets present some interesting challenges to the modern researcher. Disclosure and insider trading rules were nonexistent. Disclosure was neither full nor immediate. This creates a potential problem in conducting an event study, since there may not be a clearly defined announcement date. Furthermore, announcements were confounded by rumors and possibly by deliberate misinformation. In the case of US Steel, although rumors of a “gigantic” combination in the steel industry first appeared in newspapers on February 1, 1901, it was not until the end of that month that the general public knew the form of the merger, the companies involved, the exchange ratios, and whether any premium over market would in fact be paid.

Although the dinner of December 12, 1900 was a widely cited milestone in US Steel’s creation, it was not mentioned contemporaneously in either The Wall Street Journal or The Iron Age, an industry publication. Since insider trading was legal at the time, I was concerned that the December 12 may have led to some pre-announcement trading. I examined the stock returns for US Steel component and competitor companies for the weeks ending December 13, 1900 through January 29, 1901. I found no evidence of abnormal returns and conclude that there was no insider trading or leakage of the meeting’s information. This is consistent with the findings of Banerjee and Eckard (2001), who found no more insider trading during the First Great Merger Wave (1897-1903) than is found today. So, I begin the event period in this analysis at the week ending February 1, 1901, the date The Wall Street Journal printed its first story of a rumor of a combination in the steel industry.

In order to establish the details and chronology of the merger, I examined The Wall Street Journal and The Iron Age for the details of the events. Table 3 highlights the merger chronology. I used The Wall Street Journal because it was published daily rather than the weekly, as was the Commercial and Financial Chronicle. The Iron Age was the steel industry trade publication of the day. It too was published weekly.

I obtained stock prices from the Commercial and Financial Chronicle, the leading financial publication of the day. The Chronicle published some daily and some weekly stock prices. I used each Friday’s prices. The Chronicle reported the day’s high and low prices, and I averaged these two to obtain my quote. When there was no trading, the Chronicle reported the bid and asked prices. I averaged these. If only the bid or the asked price was reported I used that. Finally if there was nothing available for that day, I used the most recent price for that week. If there were no prices at all that week, I dropped the observation. The Chronicle also reported dividends and I include these in my stock returns.[10] Usually, the ex-dividend date was not reported, so I assumed that this was the day after the firm’s books closed, consistent with the discussion in Meeker (1922).

I estimate the abnormal returns using two models. The first is the standard market model. The second is a mean-adjusted model. The market model may not be appropriate because three of the twelve firms in the Dow Jones Industrial Average (DJIA) were steel firms and two of these three firms, Federal Steel and American Steel and Wire, became components of US Steel. I used the mean-adjusted model to account for this. The results of both models were qualitatively the same, although the market-model has greater explanatory power, evidenced by higher r-squareds and larger F-statistics.

Rather than estimate a separate regression for each security, I created four portfolios: components, competitors, and non-railroad customers, and railroad customers. The component portfolio is comprised of the publicly traded firms that had existed at the beginning of 1900. Unfortunately, the largest component of US Steel, Carnegie Steel, was privately held and thus could not be included in the analysis. This particular portfolio was value weighted based on the market values as of December 1900 [verify exact date]. The competitor and non-railroad customer portfolios were equally weighted and are comprised of firms for whom I had stock price data and for whom I could identify a main product. The railroad customer portfolio is a randomly chosen sample of ten railroads whose common stock was relatively liquid. Table 4 lists the companies used to form these portfolios.

The market model is

[pic] (1)

where,

Ri,t is the return on the ith security in week t

Rm,t is the return on the market in week t

D1,t =1 during the week ending 2/1/1901, 0 otherwise

D2,t =1 during the week ending 2/8/1901, 0 otherwise

:

D10,t =1 during the week ending 4/4/1901, 0 otherwise

[pic] is an error term with mean zero and constant variance.

For the market index I used Schwert (1990), who computed daily returns for a combination of the DJIA and the Dow Jones Railroad Average, plus an estimated dividend yield from Cowles (1939). The relative weight of each index was based on the number of stocks it contained: 12 for the Industrials and 20 for the Railroads. Thus, for the time period relevant to this analysis, the market index used is 37.5% DJIA and 62.5% DJRR, plus dividends.

The mean-adjusted model is

[pic] (2)

where,

Ri,t is the return on the ith security in week t

D1,t =1 during the week ending 2/1/1901, 0 otherwise

D2,t =1 during the week ending 2/8/1901, 0 otherwise

:

D10,t =1 during the week ending 4/4/1901, 0 otherwise

εi,t is an error term with mean zero and constant variance.

The regressions were estimated over the 62-week period from the week ending January 5, 1900 to the week ending April 4, 1901, with the event period, dummy variables D1 through D10, extending from the week ending February 1, 1901 to the week ending April 4, 1901. The results are summarized in Table 5.

For the ten-week event period, both the components and competitors exhibit statistically significant positive abnormal returns. Under the market model, the components had a cumulative aggregate abnormal return (CAAR) of 14.9%, with a t-statistic of 2.31, significant at the 5% level, while the competitors had a CAAR of 23.9%, with a t-statistic of 3.13, significant at the 1% level. Using the mean-adjusted model, the components’ abnormal return is qualitatively similar, a CAAR of 24.4%, with a t-statistic of 2.31. The competitors likewise had a positive CAAR of 36.4%, with a t-statistic of 2.73. Furthermore, as Table 6 shows, the competitors’ CAAR did not result from outliers. Under the Market Model, of the nine competitor securities, eight had positive ten-week CARs, and for five of the eight, the CAR was statistically significant. Only Tennessee Coal, Iron & Railroad had a negative CAR, and it was statistically insignificant (t=0.21). The results from the Mean-Adjusted Model are qualitatively similar, with all nine securities having positive CARs, and five of the nine are statistically significant.

The customers’ CAARs convey little information in that they were not statistically significant over the ten-week event period. Using the market model, the customers’ CAARs were 8.0%, but had a t-statistic of only 0.52. The results from the mean-adjusted model were similar, with a CAAR of 15.1% and a t-statistic of 1.36. I also used a portfolio of ten railroads to proxy the customer group. As with the other customer portfolio, the railroad portfolio had statistically insignificant CAARs and conveys no information regarding the reasons for the merger.

In examining the abnormal returns for individual weeks, the results are somewhat more ambiguous,. There were no weeks in which the components and competitors both had statistically significant abnormal returns. However, in seven of the ten weeks, the two constituencies had the abnormal returns that were the same sign: six weeks where they both were positive and one week where they both were negative. This directional equivalence is consistent with the monopoly theory of the formation of US Steel: the monopolist (or dominant firm) raises steel prices, this increasing profits for all primary steel producers. However, this is not surprising given the manner in which the information of the combination was revealed to the public. Friday, February 1, saw the first announcement of rumors of a merger in the steel industry. Beyond stating rumors that Andrew Carnegie was selling Carnegie Steel, no specific companies were mentioned. This date coincides with the end of event week 1, in which the components had abnormal returns (AR) of 1.2% (t=0.63) and the competitors had ARs of -1.9% (t=-0.85) under the market model, neither of which was statistically significant. Week 2 saw further rumors which began to mention specific companies: Federal Steel, American Steel and Wire, Carnegie Steel, National Tube, American Tin Plate, American Bridge, and Tennessee Coal, Iron and Railroad. In this week, the market model results in only the components having statistically significant positive ARs of 6.8% (t=3.60). Week 2 saw the competitors have an AR of 3.6% (t=1.60), although the AR was not significant.

The next significant announcement in the event chronology occurred in Week 5, the week ending March 1, 1901, when rumors appeared that the federal government may object to the merger. Additionally, that week saw the release of definitive information on the merger, specifically which firms were involved and what the exchange ratios would be. But, it appears that the market discounted the latter more than the former. That week witnessed an AR of –3.8% (t=-2.05) for the components and an insignificant AR of 2.5% (t=1.13) for the competitors. Weeks 8, 9 and 10 saw the tendering of shares, the addition of a few smaller firms, and the completion of the merger. Throughout this three-week period, the ARs of the components and of the competitors were positive with varying degrees of statistical significance.

VI. Results

The results clearly support the merger for monopoly theory. The period from the announcement to the completion of the formation of US Steel saw positive CAARs for the components and for the competitors, the other primary steel producers. While either monopoly or efficiency could explain the components’ positive CAAR, only monopoly is consistent with the competitors earning positive ARs. If US Steel had only been a more efficient competitor, then the competitors’ stock prices would have reacted negatively as the more-efficient US Steel took some of their business, which would have reduced their profits. The competitors’ profits increase only when a monopolistic US Steel is able to increase the price of steel by controlling the market. I cannot discount the possibility that US Steel was a more efficient firm than the sum of its components, but it clearly was able to exert some degree of market power. Had the customer’s reacted negatively, as they would have to the prospect of higher steel prices, then this would lend further support to this theory of merger for monopoly. However, the customers’ reactions were inconclusive, neither supporting nor contradicting the conjecture. The customers should have had some reaction.

Furthermore, I can reject Dewing’s swindle hypothesis, since the component firms did have a positive reaction to the announcement. This rejection is further supported by the following analysis of the sources and uses of funds in the consolidation.

VII. Gains from Consolidation

Another area examined was the gains from consolidation. Dewing and contemporary accounts discuss the huge increase in capitalization, on the order of $700 million, resulting from the formation of the company. There are several reasons to be leery of blindly accepting this figure. In those days, size was measured by total par value rather than by market capitalization. Market values, particularly of common stock, diverged greatly from the par values. In the case of US Steel, the $1.4 billion value assumes that the common and preferred each traded at $100 per share. While the preferred did trade at $95-100, the common only traded at $45-50.

As Table 7 details, on April 4, 1901 US Steel had an enterprise value of $1.034 billion based on the total market values of its common stock and preferred stock, and on the par value of its corporate debt; excluding the debt of component companies that it assumed. Of this amount, $970.5 million represents the market values of subsidiary companies and cash, leaving only $63.4 million as an increase in market value flowing to the organizers.

Again, the analysis of the market values supports Stigler’s assertion that US Steel was formed for business reasons rather than for the promoters’ profit. Most of the gains flowed to the shareholders of the subsidiary companies, not to the promoters. While it is true that Federal Steel and National Tube were J. P. Morgan companies, the individual who benefited the most from the formation of US Steel was Andrew Carnegie, no friend of Morgan’s. John D. Rockefeller, who owned much of Lake Superior Iron Ore, may have been the second largest beneficiary of the merger, and was one of US Steel’s largest shareholders[11].

As is detailed in Table 8, on January 25, 1901, the market value of the component firms for which sufficient data was available to include in the regression analysis was $262.3 million. Increasing this value by the 14.94% CAAR from the market model, which had more explanatory power than the mean-adjusted model, implies that the shareholders of these firms reaped a $39.2 million wealth increase directly attributable to the formation of US Steel. Two other firms that became components of USS, American Bridge, American Sheet Steel, were publicly traded on these dates although were not included in the regression analysis due to lack of data. On January 25, 1901 the market values of these two firms totaled $64.6 million. By April 4, 1901 their market value had increase by $18.8 million to $83.4 million. Since this 29% increase is approximately 150 basis points less than the 30% increase in market value of the firms included in the regression, I assume that the abnormal component of this 29% increase is likewise 150 basis points lower than the regression firms’ 14.94%, or 13.39%. Thus I am able to infer that the formation of US Steel increased the wealth of the shareholders of these two firms by $8.6 million. So I estimate that the shareholders of the publicly traded components of US Steel profited by $47.8 million.

US Steel’s largest component, Carnegie Steel, was not publicly traded and therefore its market value prior to the merger cannot be estimated with any degree of certainty. We know that Carnegie Steel was acquired for $441.4 million. Assuming that Carnegie Steel had experienced the same 14.94% increase in value as the other components, then the merger enriched Carnegie shareholders by $57.4 million. However, this estimated wealth increase should be seen as a lower bound, not an unbiased estimate. Carnegie Steel was the largest and most efficient steel producer of the era. It was the linchpin of the deal; without Carnegie Steel, US Steel would not have been born. It is reasonable to assume that Carnegie Steel would have commanded a higher price than the other firms. We know, at least anecdotally, that Morgan was willing to “pay-up” for Carnegie. It has been often quoted [Chernow (1990) p. 84] that in a subsequent meeting, Andrew Carnegie told JP Morgan that he should have asked for an additions $100 million, to which Morgan replied that he would have paid it.

So, between the publicly traded steel firms and Carnegie Steel, we have a wealth increase of accruing to the US Steel shareholders of at least $105.2 million.[12] If the organizers’ profits were $63.4 million, then the shareholders in the component firms made at least 66% more than them. If the House of Morgan was attempting to dupe an unsuspecting public, it did not do a very good job. [In refuting the swindle hypothesis, should I consider Gains to competitor firms? Hypothetical losses to customers from monopolization?]

VIII. Conclusion and Areas for Further Research

It is clear that US Steel was not formed to swindle the public. The major beneficiaries of the formation of US Steel were the shareholders of the component firms, particularly Andrew Carnegie, and the shareholders of the competitor firms. The organizers/promoters did not benefit much beyond their commissions for managing the offering. There was no information leakage to the public prior to the early February announcements, and no evidence of insider trading.

The evidence supports the conclusion that US Steel was formed in order to monopolize the primary steel market. This is the only explanation for the increases in prices of the component firms and of the competitor firms. While a negative reaction by the customer firms would assist in validating this conclusion, the lack of any substantive reaction on their part does not contradict the conclusion that US Steel was formed in order to monopolize the market. Furthermore, none of this is to say that US Steel was not a more efficient firm than the sum of its parts. Monopoly and efficiency are not mutually exclusive. But from the competitors’ perspective, the increase in profits from a monopolistic market outweighed the any loss in profits due to facing a more-efficient competitor.

REFERENCES

Banerjee, Ajeyo and Eckard, E. Woodrow. “Are Mega-Mergers Anticompetitive? Evidence from the First Great Merger Wave.” RAND Journal of Economics. Vol. 29 (1998), pp. 803-827.

Banerjee, Ajeyo and Eckard, E. Woodrow. “Why Regulate Insider Trading? Evidence from the First Great Merger Wave (1897-1903).” The American Economic Review, (December 2001), pp. 1329-1349.

Berglund, Abraham. “The United States Steel Corporation” in Studies in History Economics and Public Law. Edited by the Faculty of Political Science of Columbia University. New York: The Columbia University Press. 1907.

Bureau of Corporations. Report of the Commissioner of Corporations on the Steel Industry. Washington, DC: Government Printing Office. 1911.

Burton, Michael E. The 1901 Establishment of the U.S. Steel Corporation: For Monopoly and/or Efficiency? Unpublished PhD dissertation. UCLA. 1985.

Casson, Herbert N. The Romance of Steel. New York: A. S. Barnes & Company. 1907.

Chernow, Ron. The House of Morgan. New York: Touchstone. 1990.

Chernow, Ron. “The Deal of the Century.” American Heritage. July/August 1998, pp. 12-13.

Chernow, Ron. Titan. New York: Random House. 1998.

Dewing, Arthur Stone. “A Statistical Test of the Success of Consolidations.” Quarterly Journal of Economics. Vol. XXXVI (1921), pp. 84-101.

Dewing, Arthur Stone. The Financial Policy of Corporations. Volume II. Fourth Edition. New York: The Ronald Press Company. 1941.

Dittmer, Timothy. “United States Steel Corporation.” December 1997.

Eckbo, B. Espen and Wier, Peggy, “Antimerger Policy Under the Hart-Scott-Rodino Act: A Reexamination of the Market Power Hypothesis.” Journal of Law & Economics. Vol. XXVIII (1985), pp. 119-149.

Fridolfsson, Sven-Olof and Stennek, Johan. “Why Event Studies Do Not Detect Anti-Competitive Mergers,” The Research Institute of Industrial Economics. Working Paper #542. December 14, 2000.

Gallett, Craig A. “Mergers and Market Power in the US Steel Industry.” Applied Economic Letters. Vol. 3 (1996), pp. 221-223.

Haney, Lewis H. Business Organization and Combination. New York: The MacMillan Company. 1937.

The Iron Age. New York: David Williams Company. January 4, 1900-June 27, 1901 (various issues).

McAfee, R.P. and Williams, M.A. “Can Event Studies Detect Anticompetitive Mergers?” Economics Letters. Vol. 28 (1988), pp. 199-203.

McCraw, Thomas K. “Thinking About Competition.” Business and Economic History, Second Series. Vol. 17 (1988), pp. 9-29.

McCraw, Thomas K. and Reinhardt, Forest. “Losing to Win: U.S. Steel’s Pricing, Investment Decisions, and Market Share, 1901-1938.” The Journal of Economic History, Vol. XLIX (1989), pp. 593-619.

Meeker, Edward J. The Work of the Stock Exchange. New York: The Ronald Press Company. 1922.

Misa, Thomas J. A Nation of Steel. Baltimore: The Johns Hopkins University Press. 1995.

Mullin, George L., Mullin, Joseph C. and Mullin, Wallace P. “The Competitive Effects of Mergers: Stock Market Evidence from the U. S. Steel Dissolution Suit.” RAND Journal of Economics. Vol. 26 (1995), pp. 314-330.

Mullin, Joseph C. and Mullin, Wallace P. “United States Steel’s Acquisition of the Great Northern Ore Properties: Vertical Foreclosure or Efficient Contractual Governance?” The Journal of Law, Economics, & Organization. Vol. 13 (1997), pp. 74-100.

Mullin, Wallace P. “Railroad Revisionists Revisited: Stock Market Evidence from the Progressive Era.” Journal of Regulatory Economics. Vol. 17 (2000), pp. 25-47.

Parsons, Donald O. and Ray, John Edward. “The United States Steel Consolidation: The Creation of Market Control.” The Journal of Law & Economics. Vol. XVIII (1975), pp. 181-220.

Prager, Robin A. “The Effects of Horizontal Mergers on Competition: the Case of the Northern Securities Company,” RAND Journal of Economics. Vol. 23 (1992), pp. 123-133.

Robinson, Maurice H. “The Gary Dinner System: An Experiment in Cooperative Price Stabilization:” The Southwestern Political and Social Science Quarterly. Vol. 7 (September 1926) pp.137-61.

Schwert, G. William. “Using Financial Data to Measure the Effects of Regulation.” The Journal of Law and Economics. Vol. XXIV (1981), pp. 121-158.

Seager, Henry R. and Gulick, Charles A. Trust and Corporation Problems. New York: Harper & Brothers Publishers. 1929.

Stigler, George. “The Dominant Firm and the Inverted Umbrella.” In The Organization of Industry. Edited by George J. Stigler. Richard D. Irwin, Inc. Homewood, IL. 1968.

Strouse, Jean. Morgan: American Financier. New York: Random House. 1999.

Swetnam, George. Andrew Carnegie. Boston: Twayne Publishers. 1980.

Tiffany, Paul. “Opportunity Denied: The Abortive Attempt to Internationalize the American Steel Industry, 1903-1929.” Business and Economic History, Second Series. Vol. 16 (1987), pp. 229-247.

Tiffany, Paul A. The Decline of American Steel. New York: Oxford University Press. 1988.

Walker, James Blaine. The Epic of American Industry. New York: Harper & Brothers. 1949.

The Wall Street Journal. New York: Dow, Jones & Co. December 1900-April 1901 (various issues).

Warren, Kenneth. Big Steel: The First Century of the United States Steel Corporation 1901-2001. Pittsburgh: University of Pittsburgh Press. 2001.

Table 1

Stock Price Reactions to Announcement of US Steel Formation

| | | |Customer |

|Hypothesis |Component Reaction |Competitor Reaction |Reaction |

| | | | |

|Swindle Public |0 |0 |0 |

|Monopoly |+ (< Competitor Firms) |+ (> Component Firms) |– |

|Efficiency |+ |– |+ |

Table 2

Components of US Steel

| | |Tangible Assets April | | | |

| | |1, 1901 | |Market Share | |

|Company |Date Organized |($Millions) |Primary Business | |Controlled by |

| | | | | | |

|Carnegie Steel |March 1900 |198 |Semifinished steel |18% |Carnegie |

|Federal Steel |September 1898 |81 |Semifinished steel |15% |Morgan |

|National Steel |February 1899 |34 |Semifinished steel |12% |Moore |

| | | | | | |

|American Steel and Wire |January 1899 |53 |Wire | |Gates |

|National Tube |1899 |67 |Wrought-tube |90% |Morgan |

|American Tin Plate |December 1898 |25 |Tinplate |90% |Moore |

|American Steel Hoop |April 1899 |16 |Bars, hoops, cotton ties | |Moore |

|American Sheet Steel |February 1900 |18 |Sheet making |70% |Moore |

|American Bridge |May 1900 |35 |Bridge building |90% |Morgan |

|Shelby Steel Tube |1900 |3 |Seamless tube |90% | |

|Lake Superior Cons. Mines |1903 |31 |Iron mines | |Rockefeller |

|Bessemer Steamship | | | | |Rockefeller |

|Pittsburgh Steamship | | | | | |

|Oliver Mining | | | | |Carnegie |

Table 3

Chronology

|Week Ending |Event Week | |

| | | |

|December 12,1900 | |Dinner at University Club |

| | | |

|February 1, 1901 |1 |Rumors of “gigantic steel combination |

|February 8, 1901 |2 |“Semi-official” announcement that Carnegie will sell |

| | |out to Morgan |

|March 1, 1901 |5 |Federal government may object to merger |

| | |Public announcement of firms and exchange ratios |

| | |All companies accept terms |

|March 22, 1901 |8 |>60% of stocks turned in |

| | |US Steel syndicate has $200 million on call to support |

| | |stock prices |

|March 29, 1901 |9 |US Steel to pay $40 million for Lake Superior |

| | |Consolidated |

| | |US Steel will control >75% of iron ores |

| | |Rockefeller said to be largest US Steel shareholder |

|April 4, 1901 |10 |97% of stock tendered |

Table 4

Composition of Portfolios Used in Regressions

Components

• Federal Steel Preferred (15.08%)

• Federal Steel Common (9.01%)

• American Steel and Wire Preferred (12.79%)

• American Steel and Wire Common (8.02%)

• National Tube Preferred (15.10%)

• National Tube Common (9.00%)

• National Steel Preferred (9.23%)

• National Steel Common (4.51%)

• American Tin Plate Preferred (6.05%)

• American Steel Hoop Preferred (4.02%)

• American Steel Hoop Common (2.11%)

Competitors

• Cambria Steel Common

• Colorado Fuel and Iron Preferred

• Colorado Fuel and Iron Common

• Diamond Steel Common

• Republic Iron and Steel Preferred

• Republic Iron and Steel Common

• Sloss-Sheffield Steel Preferred

• Sloss-Sheffield Steel Common

• Tennessee Coal, Iron and Railroad Common

Customers (Non-railroad)

• American Car and Foundry Preferred

• American Car and Foundry Common

• US Cast Iron and Pipe Preferred

• Pressed Steel Car Preferred

Customers (Railroad)

• Atchison, Topeka & Santa Fe Common

• Baltimore & Ohio Voting Trust Certificate

• Brooklyn Rapid Transit Common

• Chesapeake & Ohio Common

• Chicago, Burlington & Quincy Common

• Chicago, Milwaukee & St. Paul Common

• Chicago, Rock Island & Pacific Common

• Delaware & Hudson Common

• Erie Common

• Louisville & Nashville Common

Table 5

Regression Results

Market Model

| |Components | |Competitors | |Customers (Non-RR) |

|Week Ending |Abnormal | | |Abnormal | | |Abnormal | |

| |Return |t-stat | |Return |t-stat | |Return |t-stat |

| | | | | | | | | |

|2/1 |.0120 |0.63 | |-.0189 |-0.85 | |.0021 |0.08 |

|2/8 |.0678 |3.60** | |.0355 |1.60 | |-.0034 |-0.13 |

|2/15 |.0035 |0.19 | |.0115 |0.52 | |-.0253 |-1.02 |

|2/21 |-.0047 |-0.25 | |-.0023 |-0.10 | |.0056 |0.22 |

|3/1 |-.0382 |-2.05* | |.0249 |1.13 | |.0015 |0.06 |

|3/8 |.0017 |0.09 | |.0500 |2.27* | |.0285 |1.14 |

|3/15 |-.0061 |-0.33 | |.0154 |0.70 | |-.0060 |-0.24 |

|3/22 |.0836 |4.44** | |.0112 |0.51 | |.0851 |3.38** |

|3/29 |.0249 |1.33 | |.0478 |2.17* | |.0001 |0.00 |

|4/4 |.0050 |0.26 | |.0638 |2.85** | |-.0082 |-0.32 |

| | | | | | | | | |

|CAAR |.1494 |2.31* | |.2389 |3.13** | |.0799 |0.92 |

| |F(11,51)=15.06 | |F(11,48)=14.26 | |F(11,48)=4.87 |

| |Adj R2 = 0.714 | |Adj R2 = 0.712 | |Adj R2 = 0.436 |

Mean-Adjusted Model

| |Components | |Competitors | |Customers (Non-RR) |

|Week Ending |Abnormal | | |Abnormal | | |Abnormal | |

| |Return |t-stat | |Return |t-stat | |Return |t-stat |

| | | | | | | | | |

|2/1 |.0396 |1.28 | |.0167 |0.43 | |.0241 |0.75 |

|2/8 |.0882 |2.86** | |.0619 |1.60 | |.0128 |0.40 |

|2/15 |.0051 |0.17 | |.0140 |0.36 | |-.0246 |-0.76 |

|2/21 |-.0340 |-1.10 | |-.0390 |-1.00 | |-.0189 |-0.58 |

|3/1 |-.0423 |-1.37 | |.0202 |0.52 | |-.0024 |-0.08 |

|3/8 |.0091 |0.29 | |.0598 |1.54 | |.0339 |1.05 |

|3/15 |-.0010 |-0.03 | |.0222 |0.57 | |-.0025 |-0.08 |

|3/22 |.1062 |3.44** | |.0404 |1.04 | |.1030 |3.19** |

|3/29 |.0352 |1.14 | |.0613 |1.58 | |.0079 |0.24 |

|4/4 |.0379 |1.23 | |.1059 |2.73** | |.0181 |0.56 |

| | | | | | | | | |

|CAAR |.2442 |2.31* | |.3636 |2.73** | |.1514 |1.36 |

| |F(10,52)=2.71 | |F(10,49)=1.68 | |F(10,45)=1.32 |

| |Adj R2 = 0.217 | |Adj R2 = 0.103 | |Adj R2 = 0.055 |

** Significant at the 1% level

* Significant at the 5% level

Table 5

Regression Results

Market Model

| |Customers (RR) | | | | |

|Week Ending |Abnormal | | | | | | | |

| |Return |t-stat | | | | | | |

| | | | | | | | | |

|2/1 |.0040 |0.29 | | | | | | |

|2/8 |-.0025 |-0.18 | | | | | | |

|2/15 |-.0046 |-0.33 | | | | | | |

|2/21 |-.0009 |-0.06 | | | | | | |

|3/1 |.0025 |-0.18 | | | | | | |

|3/8 |.0188 |1.36 | | | | | | |

|3/15 |.0117 |0.84 | | | | | | |

|3/22 |.0120 |0.86 | | | | | | |

|3/29 |-.0065 |-0.47 | | | | | | |

|4/4 |.0190 |1.35 | | | | | | |

| | | | | | | | | |

|CAAR |.0536 |1.13 | | | | | | |

| |F(11,55)=14.49 | | | | |

| |Adj R2 = 0.692 | | | | |

Mean-Adjusted Model

| |Customers (RR) | | | | |

|Week Ending |Abnormal | | | | | | | |

| |Return |t-stat | | | | | | |

| | | | | | | | | |

|2/1 |.0277 |1.10 | | | | | | |

|2/8 |.0151 |0.60 | | | | | | |

|2/15 |-.0030 |-0.12 | | | | | | |

|2/21 |-.0255 |-1.01 | | | | | | |

|3/1 |-.0007 |-0.03 | | | | | | |

|3/8 |.0253 |1.01 | | | | | | |

|3/15 |.0162 |0.64 | | | | | | |

|3/22 |.0314 |1.25 | | | | | | |

|3/29 |.0024 |0.10 | | | | | | |

|4/4 |.0471 |1.87 | | | | | | |

| | | | | | | | | |

|CAAR |.1360 |1.59 | | | | | | |

| |F(10,56)=0.88 | | | | |

| |Adj R2 = -0.018 | | | | |

** Significant at the 1% level

* Significant at the 5% level

|Table 6 |

|Competitor Firms |

|Cumulative Abnormal Returns |

|Week Ending Feb 1, 1901 - Week Ending April 4, 1901 |

| | | | | | |

| |Market Model | |Mean Adjusted Model |

| |CAR |t-stat | |CAR |t-stat |

|Portfolio Average |0.2389 |3.13** | |0.3636 |2.73** |

|  |  |  | |  |  |

|Cambria Steel Common |0.0231 |0.19 | |0.1810 |1.03 |

|Colorado Fuel and Iron Preferred |0.2472 |2.01* | |0.4792 |2.25* |

|Colorado Fuel and Iron Common |0.0698 |0.98 | |0.0894 |1.25 |

|Diamond Steel Common |0.5967 |2.54* | |0.7259 |2.9** |

|Republic Iron and Steel Preferred |0.2384 |1.20 | |0.5157 |1.74 |

|Republic Iron and Steel Common |0.2179 |2.57* | |0.3032 |2.81** |

|Sloss-Sheffield Steel Preferred |0.5195 |2.77** | |0.6112 |3.06** |

|Sloss-Sheffield Steel Common |0.2242 |3.60** | |0.2471 |3.85** |

|Tennessee Coal, Iron and Railroad Common |-0.0306 |-0.21 | |0.1532 |0.75 |

|** Significant at the 1% level | | | | | |

|* Significant at the 5% level | | | | | |

| | | | | | |

| |Table 7 | | | | |

| |Uses of Funds from US Steel Capitalization | | | |

| |April 4, 1901 | | | |

| | | | | | |

|Sources of Funds: | | | | |

| | | | | | |

| |US Steel Common Stock | | 243,740,324 | | |

| |US Steel Preferred Stock | | 486,240,024 | | |

| |US Steel Bonds | | 304,000,000 | | |

| | | | | | |

|Total Sources | | $1,033,980,348 | | |

| | | | | | |

|Uses of Funds: | | | | |

| | | | | | |

| |Acquisition of Major Components: | | | | |

| | | | | | |

| |Federal Steel Co. | 80,956,643 | | | |

| |American Steel and Wire | 69,302,331 | | | |

| |National Tube | 74,600,000 | | | |

| |National Steel | 51,147,390 | | | |

| |American Tin Plate | 44,110,000 | | | |

| |American Steel Hoop | 22,430,000 | | | |

| |American Sheet Steel | 34,666,934 | | | |

| |American Bridge | 48,641,142 | | | |

| |Shelby Steel Tube Co | 2,677,370 | | | |

| |Lake Superior Consolidated Mines | 57,081,359 | | | |

| |Carnegie Steel | 441,424,284 | | | |

| | | 927,037,453 | | | |

| | | | | | |

| |Cash Acquisitions: | | | | |

| | | | | | |

| |Bessemer Steamship |8,500,000 | | | |

| |Pittsburgh Steamship & Oliver Mining (est.) | 10,000,000 | | | |

| | | 18,500,000 | | | |

| | | | | | |

| | | | | | |

| |Cash |25,003,000 | | | |

| | | | | | |

| |Total Business Uses | | $ 970,540,453 | | |

| | | | | | |

| |Gain to Organizers | | $ 63,439,895 | | |

|Table 8 |

|Market Value Increase |

| | | | | | | | | | | | | | | | | | | | | |Increase | | | | | | | | |Raw |Market | |Attributable | | |25-Jan-01 | |4-Apr-01 | |Increase | |Return |Return |CAAR |to USS | | |($000) | |($000) | |($000) | | | | |($000) | |Component Firms Included | | | | | | | | | | | in Regression Model | 262,298 | | 342,546 | | 80,248 | |30.59% |12.98% |14.94% | 39,187 | | | | | | | | | | | | | |American Sheet Steel | 22,907 | | 34,667 | | 11,760 | |51.34% | | | | |American Bridge | 41,651 |  | 48,641 |  | 6,990 | |16.78% | | | | | | | | | | | | | | | | |Subtotal | 64,558 |  | 83,308 |  | 18,750 | |29.04% |12.98% |13.39% | 8,644 | | | | | | | | | | | | | |Total Publicly Traded | | | | | | | | | | | | Component Firms | 326,856 |  | 425,854 |  | 98,998 |  |30.29% |  |  | 47,831 | |

-----------------------

[1] Arthur Stone Dewing, The Financial Policy of Corporations, Volume II, Fourth Edition, pp. 924-925. (1941).

[2] George J. Stigler, “The Dominant Firm and the Inverted Umbrella” in The Organization of Industry, p. 108. (1968). “Watered stock” seemed to be applied to any security whose market value exceeded the value of its tangible assets. The discounted future value of the firm’s earnings stream does not appear to have been considered. It may be interesting to formally investigate the use of this term. It may be that watered stocks were high earnings and/or high growth stocks and did not necessarily represent fraudulent activities by the promoters.

The term “watered stock” has an interesting etymology. Legend has it that Wall Street speculator Daniel Drew was the first to apply it to stocks. When Drew was a boy he lived on a farm in New York. One of his chores was to drive the cattle to market. Immediately before and during the drive he would dehydrate the cattle. Just before arriving at the stockyard, he came to a stream and allowed the cattle to drink. The dehydrated cattle gorged themselves on the water and became temporarily bloated and overweight. Drew then took the cattle to the stockyard where they were sold by weight. Presumably the cattle’s excess weight did not last very long. Upon embarking on his Wall Street career, Drew saw the parallels between bloated cattle and “overpriced” stocks.

[3] Others who cited US Steel for excessive capitalization include Seager and Gullick (1929) p. 224, Casson (1907) pp. 229-231, and Haney (1937) p. 263.

[4] George J. Stigler, “The Dominant Firm and the Inverted Umbrella” in The Organization of Industry, p. 109. (1968).

[5] As will be discussed later, many of the finished (steel) goods markets had become monopolized prior to the formation of US Steel.

[6] American Tin Plate, formed in 1898, controlled 90% of tinplate production. National Tube, formed in 1899, controlled 90% of the steel tube production. Shelby Steel Tube, formed in early 1900, controlled 70% of the seamless steel tube production. American Sheet Steel controlled 70% of sheet making. American Steel and Wire, formed in 1899, controlled the steel wire production

[7] Ron Chernow, “The Deal of the Century,” American Heritage, July/August 1998, p. 12.

[8] In what was probably pure coincidence, the week before Morgan attended a dinner in his honor hosted by then-Vice President and later “Trust-Buster” Theodore Roosevelt.

[9] Ron Chernow, “The Deal of the Century,” American Heritage, July/August 1998, p. 12.

[10] The customer railroad portfolio does not follow this convention. The returns in this portfolio exclude dividends. The customer railroad portfolio is compared to a market index that excludes dividends. Otherwise, the market index is identical to that described below.

[11] Ron Chernow, Titan. Random House: New York, 1998, p. 393.

[12] I exclude Consolidated Lake Superior Coal from this analysis due to a lack of data. Also excluded are several smaller acquisitions.

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download