Robert Bricker CASE WESTERN RESERVE UNIVERSITY
and
Kevin Brown CASE WESTERN RESERVE UNIVERSITY
THE USE OF HISTORICAL DATA IN ACCOUNTING RESEARCH: THE CASE OF THE AMERICAN SUGAR REFINING COMPANY
Abstract: In 1908, the American Sugar Refining Company (ASR) reversed its long-held policy of secrecy as to its financial condition and performance. Prior work, applying contemporary capital market methods to ASR security price data of that period, has suggested a value to ASR shareholders of this policy reversal. This paper examines the historical record of that time and presents additional evidence on this matter, particularly in terms of identifying potentially confounding events occurring during the period under study. The results of this analysis suggest a difficulty in attributing observed abnormal returns to ASR’s secrecy policy reversal on the basis of the results obtained from applying capital markets methods. This analysis is useful for scholars interested in applying modern capital market methods to historical data. It highlights the significance of the possible effects of contemporaneous historical events, focuses attention on the importance of a deep understanding of the historical period studied, and suggests a value in combining historical and empirical-markets methods to gain a richer understanding of the events and conditions in the time period under study.
The application of modern capital markets methods to explore accounting issues in historical time periods has attracted increased interest in recent years. One such study, Porter et al, (1995) (hereafter PSW) examine the effects of American Sugar Refining’s (ASR’s) 1908 reversal of a financial secrecy policy to determine the value of voluntary disclosure per se and voluntary disclosure policy. They find in part, using traditional capital markets methods and an event window around the time of the secrecy policy reversal, positive abnormal returns for ASR stock. These findings lead them to conclude that the secrecy policy reversal had measurable value to ASR stockholders. Narrowly, the purpose of this paper is to assess these findings and to consider whether other chronologically proximate events and conditions could plausibly have affected the valuation of American Sugar securities. Broadly, this paper seeks to address the use of data in capital markets studies of historical time periods, and the interpretation of results derived in such studies. Not addressed in this paper are underlying issues related to institutional arrangements, acceptable trading practices, and other matters pertinent to the issue of market efficiency during this historical period. We are not challenging the assumption that ASR security pricing was efficient with respect to publicly available information. Readers interested in studying issues of this period related to market efficiency are referred to Previts and Bricker (1994).
The remainder of this paper is organized as follows. First, we review the PSW study and summarize the historical events amidst which it is set, these being the illness and death of H.O. Havemeyer, the long time President of ASR, the succession of a new president, and the company’s annual meeting at which time ASR formally announced the reversal of its long-held financial secrecy policy. Several research issues pertinent to the PSW study, primarily related to the existence of other possibly significant historical events during the period covered by PSW, are identified. We then provide a contextual, historical analysis of contemporaneous events and conditions of the period and use this as a basis for evaluating the PSW study of ASR. Particular attention is focused on events reported in the financial press that may have affected ASR security pricing, dates on which press reports related to ASR’s financial secrecy policy are published, and the period of uncertainty surrounding the succession to the ASR Presidency that occurred following Havemeyer’s death. Next, using the PSW data and methods, tests extending the original PSW analysis are conducted based on the results of our historical analysis. Following a presentation and discussion of these results, we conclude the paper with a discussion of the opportunities and difficulties of applying capital markets methods to data from historical time periods.
BACKGROUND AND RESEARCH ISSUES
The PSW study follows a line of research that has applied contemporary capital markets methods and models to data from historical time periods. Prior work has included Chow (1983) and Sivakumar and Waymire (1993 and 1994). The focus of these studies on historical time periods makes them interesting and appealing. Chow’s work, however, has been specifically criticized by Merino et al (1987), and Previts and Bricker (1994) argue generally that such studies must fully consider historical contextual issues that are essential in both the design of such studies and the interpretation of results.
The PSW study finds that ASR shareholders earned abnormal returns as a result of the 1908 reversal of Henry 0. Havemeyer’s long-held financial secrecy policy. As summarized in PSW, Havemeyer, who opposed disclosure of any ASR information, died suddenly on December 4, 1907 after becoming ill on November 28, 1907. Havemeyer’s successor, W. B. Thomas, “announced the company’s intent to reverse its long-standing secrecy policy by making periodic reports” (PSW, 1995, p. 129) at the company’s annual meeting on January 8, 1908. On March 23, 1908, ASR made its first annual report available to shareholders. PSW examine cumulative abnormal returns to ASR shareholders for both the secrecy policy reversal on January 8, and the publication of the annual report on March 23 (and each preceding day). They find a positive wealth effect which is statistically significant for three sets of dates: January 8th, January 8th and March 23rd combined, and January 7th and 8th and March 21st and 23rd combined. On the basis of the abnormal returns of these dates, PSW conclude the existence of a value to voluntary reporting – for the first event as related to voluntary reporting policy, and for the second event as related to voluntary reporting per se, for ASR shareholders. We do not address the value of voluntary reporting per se in the remainder of this paper, but rather on the issue of a value to a policy of voluntary reporting, as studied by PSW in their assessment of the effects of ASR’s secrecy policy reversal.
The PSW study measures ASR returns from September 30th, 1907 to May 29th, 1908. PSW use a dummy variable is used to measure the difference in returns during the event window(s) in comparison to returns outside the event window. The event date for the secrecy policy reversal (disclosure precommitment date) is defined as January 8, the date on which the company’s policy reversal was officially announced during its annual meeting. The resulting model is significant (p = .023) and they find a significant cumulative positive abnormal return of 2.35 percent.
Several issues not fully addressed in PSW may bear on their results and the interpretation thereof, as listed below.
• The date used by PSW to isolate the financial secrecy policy reversal effect (January 8th) is the date of ASR’s annual meeting. It is conceivable that other potentially significant disclosures may have occurred on that date. Such disclosures would raise issues about distinguishing ASR’s secrecy policy reversal from its voluntary reporting, per se, (or other economic effects) as causes of ASR security price changes.
• January 8th is the sole date used by PSW in assessing the (separable) effect of ASR’s secrecy policy reversal. If, however, there are other dates on which ASR’s secrecy policy was addressed in financial press reports prior to January 8, then such dates should also be added to any empirical analysis addressing the effect of the secrecy policy reversal on ASR securities.
• The period of the illness and death of Havemeyer is included by PSW as a part of the period for estimating ASR normal returns. However, any systematic ASR pricing reaction to Havemeyer’s passing may influence the PSW results. Several research questions arise from these issues. First, is it reasonable to attribute any abnormal ASR price effect on January 8th to a change in the company’s financial secrecy policy, or do other disclosures confound such an interpretation? Second, were there other dates in addition to January 8th (the date used by PSW) which are plausible event dates for assessing the effect of the secrecy policy reversal? Third, was there an ASR security price reaction to the succession uncertainty that may have occurred with and following the illness and death of Havemeyer? If so, would this effect influence PSW’s findings?
ANALYSIS
Our analysis consists of two parts. First, we study the events and conditions roughly contemporaneous with January 8th to address the three questions posed above. Then, using the same data and methods as PSW, we conduct additional empirical testing based on the results of our first phase of analysis.
Research question one explores the attribution of the observed ASR abnormal return effect to the announcement of the secrecy policy reversal, and whether other information released on this date could have had information content and have affected ASR security pricing. An analysis of the financial press articles of the period in the Financial Chronicle, the. Wo// Street Journal, and the financial section of the New York Times raises issues potentially bearing on this matter. Significantly, January 8th was not just a date on which ASR announced the reversal of the secrecy policy, but (as noted in PSW) was the date of ASR’s annual meeting. For instance, it was on this day that the company made its initial report of its annual operations to shareholders, and also discussed other economic and political developments prospectively affecting ASR. It appears (and widely believed), for example, that sugar crop and tariff announcements were important factors in ASR security pricing. As reported on the following day by the Wall Street Journal, management discussed several topics during the annual meeting: that the past year had been the best that the company had experienced, that the policy of the company with new management would be less antagonistic and less secret, that various lawsuits would be adjudicated, that sugar crops were smaller and prices higher, that labor trouble in Cuba implied a smaller crop, and that government statements on tariffs were favorable to ASR. All of these reported disclosures, to the extent unanticipated, could be expected to increase ASR security prices. Even with the presence of some level of anticipation, the elimination of uncertainty accompanying the official company announcement would also tend to increase ASR security prices and returns. These several potentially significant information disclosures confound interpretations of stock price reactions for the day and certainly do not lend themselves to an unambiguous attribution of the day’s abnormal price effect to any single cause.
Research question two addresses the existence of additional dates which may have lead to a market anticipation of a reversal of ASR’s secrecy policy. We searched financial press reports of the period from the date of Havemeyer’s illness to the date of the official announcement of the policy reversal to identify all days on which the secrecy policy reversal of ASR was discussed, implied, or anticipated. We identified the following four dates:
• In The Wall Street Journal of Thursday, December 5th 1907 in an article following H. 0. Havemeyer’s death, it was asserted that “(Havemeyer’s) policy was not that of publicity, so that he belonged to a passing rather than to a new era of corporate finance.” This assertion implies that Havemeyer’s secrecy policy would not survive him.
• In the Wall Street Journal of Tuesday, December 10th, 1907, the following article appeared: “American Sugar Refining Co.: Havemeyer’s Death Likely to Result in Abandonment of Secrecy Policy”.
• In the Wall Street Journal of Tuesday, January 7th, 1908 in an article preceding the next day’s annual meeting, it was written that “. . . it is supposed that the policy of the company will change, the keynote of the new administration being less antagonistic and less secret”.
• In the Wall Street Journal of Wednesday, January 8th, in an article published prior to that day’s annual meeting, it was written that “Friends of (ASR) . . . admit that more publicity in regard to the company’s affairs would be better policy… President Thomas has given his word to this effect… the first step in this direction is
expected at the annual meeting tomorrow, when President Thomas may in his review of the year disclose more information than has been given out heretofore.” January 8th was also, of course, the date of the actual annual meeting at which the financial secrecy policy reversal was officially announced later that day (See January 9 –
New York Times).
Although PSW use event dates of January 7th and 8th as well as those of March 21st and March 23rd, only January 8th is tested individually for a secrecy policy reversal effect. PSW’s remaining tests address the joint hypothesis of a price response as a function of both voluntary reporting and the secrecy policy reversal. If PSW’s hypothesis about the value of a voluntary reporting and publicity policy, per se, is correct, then we would expect to see positive security price effects on each of the event days, with the effect on the 8th being limited to the elimination of whatever uncertainty remained regarding reversal of the secrecy policy.
As related to the third research question posed, any analysis of this period or interpretation of results requires consideration of any possible effect on ASR security prices of the illness and death of Henry Havemeyer. Particularly, as described subsequently, if Havemeyer’s death resulted in a period of uncertainty related to management succession, and if that period was included in an estimation window for calculating normal ASR returns, then the stationarity assumption of PSW’s beta estimate would be violated, with the result that in the estimation of ASR normal returns would be mathematically biased.
Specifically, the statistical analysis computing returns in the immediately subsequent period would be biased in favor of finding abnormal returns.
Havemeyer’s passing was clearly regarded as a significant event in the eyes of ASR, the financial markets, and its shareholders, according to financial press accounts of the time. These accounts reflect the belief, for example, of an association between Havemeyer’s illness and death and an ASR’s stock price decline that occurred. On December 4th, the Wall Street Journal’s daily column “Features of the Market” column noted “.American Sugar was naturally depressed on the serious illness of H.O. Havemeyer”. On December 6th, the same column noted that “(t)he first transaction in (ASR) was one point below the closing sale of the previous day on President Havemeyer’s death. It is admitted that the loss to the company is a very serious one . . .” On December 12, as ASR’s common stock price neared its nadir, “Features of the Market” stated that “(ASR) selling . . . comes through conservative commission houses . .. It is naturally based upon the very tangible loss the company . . . has sustained in H.O. Havemeyer’s death and represents investment holdings.” On December 16, the column reported that “Boston has been selling the stock since H. 0. Havemeyer’s death. This is a natural result, as investors there had great faith in the late president.” An article in December 17th’s Wall Street Journal stated “… One reason advanced forthe marked decline inthestockis the loss of H.O. Havemeyer, which, if be true, is indeed a tribute to his ability as the real genius of the American Sugar Refining Company.” On January 4th, well into ASR’s price rebound, and following a day on which the stock closed up over 6 1/2 points to 107, the “Features of the Market” column noted, “One reason for the exceptional strength in Sugar is the real scarcity of the stock … It was always a dangerous short sale, and was oversold when HO. Havemeyer died”.
As the matter of succession was not settled at ASR at the time of Havemeyer’s illness and death, it is plausible to believe that his passing initiated a significant process of increasing uncertainty as to management succession and ASR prospects. PSW, in contrast, address the possible impact of Havemeyer’s death from a different perspective:
“… our results may be confounded by Havemeyer’s sudden death. For example, investors may have expected Havemeyer’s death to lead to the adoption of a disclosure policy, and impounded any wealth effects at that time. Accordingly, we investigated ASR stock returns at the time of Havemeyer’s death. The evidence does not support the hypothesis that
favorable effects of the subsequent policy reversal were capitalized when Havemeyer died; the market adjusted return on the date of his death was -2.26 percent. No large price movements were observed in adjacent days.” (PSW, 1995, p. 136-137)
PSW, then, focus on the possibility that the death of Havemeyer itself may have signaled a reversal of the secrecy policy, and so test for a positive abnormal return effect. But they do not examine an alternative possibility that an uncertainty-increasing process began at the date of Havemeyer’s illness regarding management succession and ASR prospects and concluded at ASR’s announcement of his successor, resulting in a negative abnormal return effect for the period.
Chronologically, this succession-uncertainty process occurred several weeks prior to ASR’s annual meeting and the formally announced reversal of the secrecy policy. Havemeyer’s sickness began November 29th (he died December 4), and W.B. Thomas was announced as his successor on December 20th. A visual inspection of Figure 1, which is reproduced from PSW, shows that ASR returns turned negative about the time of Havemeyer’s sickness and death, and remained negative until the December 20th announcement regarding W.B. Thomas.
decline in ASR returns during this period, it does provide prima facia evidence of a decline in ASR security prices as well as an analyst and press belief that this was so. Such an effect, as described above, would bias an analysis of abnormal returns on January 8th in favcr of a rejection of the null hypothesis. Specific empirical tests on this are described later in the paper.
The empirical portion of our analysis began with a replication of a portion of PSW in order to establish comparability with its data and methods. Using the same parameters and process for selecting a data set, and duplicating the general model employed by PSW, we tested for a price effect on January 8th, the date of the financial secrecy policy reversal. Following PSW we employed a dummy variable for the date of January 8th as the secrecy policy reversal event date, and used the railroad index to proxy for the market. The results, shown in Table 1, Panel A, compare the PSW results (Item 1) with our replication (Item 2); The results are generally consistent, including an event-variable coefficient (probability) of .0235 (.023) and .0237 (.026) respectively. Therefore, we are comfortable that our data and method fairly closely approximates those of ¦ P SW.
Table 1 Estimation Results for Event Parameter Models
Panel A Eyen
(1) Porter Results .008 1.11.833 12.18 .0235 2.29.023.436 2.35
1/8/08)
(2) Replication Results .001 .91.798 11.31 .0237 2.24.026.387 2.37
(1/8/08)
(3) Four Event Dates Results .001 1.21 .796 11.16-.0061-1.14 .250 .386 -2.45
(12/5/07,12/10/07,
1/7/08,1/8/08)
(4) Uncertainty Results .001 1.72 .791 11.23 -.0070 -2.47 .014 .400 -10.47
(12/3/07 through 1/19/08)
(5)TwoEventDatesResults .001 1.01.795 11.12 .0031 .40.687.382 .61 (1/7/08,1/8/08)
Panel B Model Respecified to Include Two Event Parameters
a aft) P fift) 81 Sift) 51(p) 81CAR% 82 82ft) 82(p) 82CAR% r> .001 1.67.791 11.21 -.007 -2.458 .015 11.20 .0025 .387 .737 .50 .398
where 5] represents the “succession uncertainty” event of 12/3/07 to 1/19/08 and 52 represents the “secrecy policy reversal” event of 1/7 and 1/8/08.
Next, to empirically address research question 2, we tested for a policy reversal effect using the four dates identified above (days on which reversal of the secrecy policy was discussed in press articles). Although PSW use event dates of January 7th and 8th as well as those of March 21st and March 23rd, only January 8th is tested individually for a secrecy policy reversal effect. PSW’s remaining tests address the joint hypothesis of a price response as a function of both voluntary reporting and the secrecy policy reversal. Using the method described above, we tested for a secrecy policy reversal effect using December 5th, December 10th, January 7th, and January 8th. The results are given in Table 1, Panel A, Item 3 which shows no significant effect on the secrecy reversal variable for the dates with secrecy policy reversal references. Notably, the sign of the coefficient on the event variable is negative (-.0061).
Using the data set and general model described above, we tested the hypothesis derived from research question 3: that an uncertainty-increasing process began with the illness of Havemeyer and ended with the announced selection of Thomas. We employed a dummy variable for the dates of December 3rd, when Havemeyer’s illness was first reported in the New York Times, through December 19th, the day before the Wall Street Journal announced W.B. Thomas’s upcoming election as ASR president. We observe in Table 1, Panel A, Item 4 a highly significant effect (p. = .014) on the event days in the direction consistent with an uncertainty-increasing process (that is, negative). Using PSW’s method of calculating the size of the abnormal return as the product of the dummy coefficient (-.0070) and the number of days in the event period (15), we find a cumulative abnormal return of -10.50%. Overall, this evidence is consistent with a picture of investors who became increasingly distressed by the uncertainty of management succession until the experienced W. B. Thomas was identified as H. O. Havemeyer’s successor. As a further test of investor uncertainty during
The Case of the American Sugar Refining Company
this period, we compared an ASR market model using the period of September 30th 1907 through January 19th, 1908, with a model using a period of the same length immediately after. As shown in Table 2, ASR’s beta increases from .786 to .963 in the two models, and their difference is marginally significant at alpha = . 10.
Irrespective of the interpretation of the above results, they demonstrate that any model used for estimating ASR normal returns that includes the period from December 3rd through December 19th cannot assume beta stationarity; the inclusion of the period effectively incorporates a bias for rejecting a null hypothesis for any event date(s) proximately subsequent to that period. This occurs because the inclusion of the uncertainty-increasing period mathematically reduces the slope (coefficient) on the systematic (market) return. Given the evidence and findings described above, the PSW findings are reinforced by the occurrence of an event window that merely happened to fall following a period of significant negative abnormal returns included in the estimation period. Indeed, any event identified during the period immediately following the resolution of the succession issue would be biased towards the finding of positive abnormal returns for ASR investors. A second implication, as mentioned earlier, is that any policy reversal tests using event dates falling during this period could be confounded by the succession uncertainty effect. For this reason, our earlier results, which included two secrecy policy reversal event dates falling during this apparent succession uncertainty period, may be biased against finding a positive ASR abnormal return associated with a reversal in its secrecy policy.
1 It is hard to have much confidence in these results. As described in the next section, tests reveal a negative security market response to the succession uncertainty accompanying the illness and death of Havemeyer, which potentially confounds any analysis including event days in the period from November 29th to December 20th. The policy reversal test described above includes the event dates of December 5th and December 10th. Further complicating any interpretation of our results, news articles on ASR on the two remaining days, December 7th and 8th, refer to both the secrecy policy reversal and other matters of potential importance. We have previously described the topics covered in the press on January 8th. The day before ASR’s annual meeting (January 7th) The Wall Street Journal not only discussed the anticipated secrecy policy reversal, but based on favorable news about the sugar market and the favorable estimates made regarding ASR’s financial results, the article declares ASR’s “past fiscal year.. . was one of the
We, therefore, respecified our model to include a separate dummy variable for the event dates falling in the succession uncertainty period and used only the event dates of January 7th and January 8th (we removed December 5th and 10th). The results, shown in Table 1, Panel B again suggest no effect of the secrecy policy reversal. The coefficient on the policy reversal variable is positive (.0025) but not significant, with an implied CAR of .50%. The succession uncertainty variable is negative (-.007) and significant (p = .015), with an implied CAR of -11.2%. While this may appear odd, our investigation suggests that it may reflect (as reported by the financial press) C.A. Spreckels’ published attack on ASR which appeared on that day.
To test the sensitivity of the results to the separation of the succession uncertainty dates as a second event, we reran the analysis including the succession uncertainty dates as estimation dates (that is, along with all the other dates in the estimation period), continuing with January 7th and 8th as event dates. Table 1, Panel A, Item 5 shows a positive coefficient on the policy reversal variable that, while slightly larger than the result in Panel B, is still statistically insignificant.
CONCLUDING REMARKS
In conclusion, the results do not necessarily rule out that the positive wealth effect observed on January 8* for ASR shareholders resulted fromASR’s reversal of its secrecy policy. However, our analysis of the events and context surrounding January 8th suggests the following. First, there were several disclosures with potential import to investors. While PSW are correct in their finding of an empirically measurable abnormal return on January 8th, it is difficult to know how much of this effect can be attributed to the official policy reversal announcement versus ASR’s report of a good year, Cuban crop shortages, favorable government tariff policies, the methodological artifact stemming from the succession uncertainty following Havemeyer’s death (nonstationary beta), or the secrecy policy reversal. There are, thus, several possible factors which may have contributed to the ASR abnormal return effect found by PSW on January 8th.
Second, the financial press anticipated the reversal of the secrecy policy other dates. When included in an analysis, no positive abnormal best that the company has experienced.. .”
The Case of the American Sugar Refining Company
returns for those dates are found. However, these results may be suspect because the analysis includes event dates during the negative-return period that accompany the apparent succession uncertainty at the time of H.O. Havemeyer’s death. Nevertheless, when controlling for the succession uncertainty dates and reducing the event dates to January 7th and 8th, we found no secrecy policy reversal event effect, even when including the succession uncertainty dates in the “normal return” estimation period. Yet even this result cannot be relied upon, as financial press articles attribute ASR stock price weakness on the 7th to C.A. Spreckels published criticism of ASR on that date.
Finally, as mentioned just above, the positive abnormal return of January 8th is, in part, a statistical artifact resulting from the succession uncertainty following Havemeyer’s death. During this period of succession uncertainty ASR returns were below market returns. The nonstationary beta confounds the interpretation of any event dates occurring during the succession uncertainty period (including some of the secrecy reversal policy dates), and mathematically biases any statistical analysis using this period in the window for estimating normal returns in favor of finding a positive abnormal return for immediately subsequent dates.
We conclude that it is not possible with the present data and capital markets methods to empirically and statistically attribute ASR’s positive abnormal returns to its reversal in its secrecy policy. Our objective, however, is not to disprove that ASR’s secrecy policy reversal resulted in a stock price change. Rather, our point is to emphasize the importance in considering all the contextual factors salient to research focusing on historical time periods. This is particularly important when applying contemporary markets methods to data from such periods. Aside from issues of market efficiency during this period (Previts and Bricker, 1994), the nature of historical data may frequently place some limitations on the drawing of conclusions, particularly when the data are scarce and the historical context of the events renders data interpretation ambiguous. Correspondingly, it may be in such instances that historical analysis can be a useful and tool for more fully investigating, interpreting, and augmenting the results of empirical work. In this way, modern capital markets method and historical analysis may prove to be valuable complements in studying historical time periods.
It is helpful and desirable for accounting scholars to carefully construct research that applies contemporary methods to historical accounting data as a backdrop for considering present-day issues.
Empirical researchers are to be commended for conducting pioneering work in this area. Future research of this sort should continue to develop full and careful approaches to combining historical analysis with modern capital market methods, so that adequate consideration of the fundamental contextual factors can be used in the design of such research and in the interpretability of results.
REFERENCES
Chow, C. W., “The Impact of Accounting Regulation on Bondholder and
Shareholder Wealth: The Case of the Securities Acts,” The Accounting
Review, 1983, v58(3), 485-520. Merino, B. D., B. S. Koch and K. L. MacRitchie, “Historical Analysis – A
Diagnostic Tool for ‘Events’ Studies: The Impact of the Securities Act
of 1933,” The Accounting Review, 1987,v62(4), 748-762. New York Times. “H.O. Havemeyer Critically 111”, The New York Times,
Tuesday 12/3/07. New York Times. “No More Secrecy in Sugar Trust”, The New York Times,
Thursday 1/9/08. Porter, B., K. Sivakumar, and G. Waymire, “Disclosure Policies and
Shareholder Wealth in the Early Twentieth Century: The Case of the
American Sugar Refining Company,” Journal of Accounting, Auditing
& Finance, Winter 1995, 121-145. Previts, G. J., and R. J. Bricker, “Fact and Theory in Accounting History:
Presentmindedness and Capital Market Research,” Contemporary
Accounting Research, 1994, vlO(2), 625-641. Sivakumar, K. N., and G. Waymire, “The Information Content of Earnings
in a Discretionary Environment: Evidence from NYSE Industrials, 1905-
1910,” Journal of Accounting Research, 1993, v31(l), 62-91. Sivakumar, K. N., and G. Waymire, “Voluntary Interim Disclosure by Early
20th Century NYSE Industrials,” Contemporary Accounting Research,
1994, vlO(2), 673-698. The New York Times. “New Sugar Trust Head: May Be Chosen Next Week
and May Be H.O. Havemeyer’s Son.” The New York Times, Friday
12/06/07. Wall Street Journal. “Features of the Market,” Wall Street Journal, 12/4/07,
12/6/07, 12/12/07, 12/16/07, 1/4/08. Wall Street Journal. “H.O. Havemeyer,” Wall Street Journal, Thursday
12/5/07. Wall Street Journal. “American Sugar Refining Co: Havemeyer’s Death
Likely to Result in Abandonment of Secrecy Policy,” Wall Street
Journal, Tuesday 12/10/07. Wall Street Journal. “American Sugar Refining: Trade and Company
Conditions Good” Wall Street Journal, Monday 12/16/07.
The Case of the American Sugar Refining Company
Wall Street Journal. “American Sugar Refining Co. Again Center of Interest”
Wall Street Journal, Tuesday 12/17/07.
Wall Street Journal. “American Sugar” Wall Street Journal, Friday 12/20/07. Wall Street Journal. “American Sugar Refining Co. Completes a Successful
Year” Wall Street Journal, Tuesday 1/07/08.
Appendix A
American Sugar Refining Prices November 1,1907 through January 31,1908
Dow Jones
American Railroad
Sugar index Notes
Havemeyer is ill 11/28
Havemeyer illness reported in NY Havemeyer
Havemeyer death reported. Secrecy reversal—date 1
Secrecy policy reversal—date 2
W.B. Thomas announced as Havermeyer successor
Date
11/1 100.625 84.360
11/2 100.625 84.140
11/4 104.000 85.050
11/6 105.000 85.720
11/7 103.625 84.270
11/8 103.625 85.090
11/9 103.625 85.070
11/11 105.875 85.910
11/12 104.500 85.410
11/13 104.250 84.800
11/14 101.875 84.150
11/15 100.500 82.500
11/16 101.250 82.970
11/18 103.000 84.600
11/19 101.125 82.930
11/20 101.750 82.380
11/21 101.000 81.410
11/22 101.500 81.490
11/23 103.000 83.400
11/25 103.000 81.720
11/26 104.000 83.510
11/27 103.000 84.090
11/29 103.750 85.800
11/30 106.000 87.130
12/2 107.750 88.280
12/3 106.000 87.320
12/4 106.500 89.110
12/5 107.250 90.300
12/6 107.750 90.560
12/7 107.000 90.120
12/9 105.000 89.190
12/10 102.000 87.880
12/11 98.375 87.850
12/12 98.375 87.850
12/13 94.125 86.940
12/14 94.125 87.760
12/16 95.625 86.730
12/17 95.750 86.610
12/18 96.250 87.230
12/19 98.000 87.390
12/20 100.000 88.780
Appendix A (continued)
Dow Jones
American Railroad
Date Sugar index Notes
12/21 100.125 89.350
12/23 99.125 88.110
12/24 98.750 87.610
12/26 96.875 87.010
12/27 100.250 88.410
12/28 99.250 88.350
12/30 100.875 89.500
12/31 99.500 88.770
1/2/08 101.375 89.810
1/3 107.000 90.380
1/4 106.250 90.370
1/6 106.500 91.120
1/7 104.750 91.150 Secrecy policy reversal—date 3
1/8 107.000 90.820 Annual Meeting. Secrecy policy
1/9 108.250 92.860 —date 4
1/10 108.250 92.030
1/11 110.250 93.750
1/13 112.250 94.270
1/14 113.375 95.060
1/15 113.250 95.100
1/16 112.125 94.680
1/17 113.625 95.270
1/18 113.625 95.750
1/20 113.250 94.670
1/21 111.000 93.090
1/22 112.375 92.760
1/23 111.875 92.460
1/24 111.500 91.660
1/25 113.500 92.220
1/27 113.500 92.730
1/28 114.500 93.400
1/29 112.500 92.400
1/30 113.500 92.440
1/31 114.250 92.190
Appendix B Pertinent Articles from the Wall Street Journal
1. “Features of the Market” extract – dates as noted:
Wed. 12/04/07 “American Sugar was naturally depressed on the serious illness of H. 0. Havemeyer. Washington has been selling a little of the stock lately, which has often shown some tendency to sell off about the time of the meeting of Congress. American Sugar, however, has been very much out of politics for a long time past.”
Fri. 12/06/07 “The first transaction in American Sugar was one point below the closing sale of the previous day on President Havemeyer’s death. It is admitted that the loss to the company is a very serious one, as he raised it from the most sensational gambling counter in the unlisted department to one of the best held and distributed industrial stocks in the market.”
Wed. 12/11/07
“. . . Sugar also was subjected to considerable pressure. There was a poor market in it.. .”
Thurs. 12/12/07
“Inquiry from people in close touch with the Havemeyer family and the American Sugar situation disposes of any rumors of a change in the holding of the stock arising out of the death of the president. The selling, however comes through conservative commission houses, and in the opinion of specialists does not indicate a short position of any consequence. It is naturally based upon the very tangible loss the company has sustained in H. O. Havemeyer’s death, and represents investment holdings.”
Fri. 12/13/07 “American Sugar opened active and higher. H. Content
who is thought to have sold something like 25,000 shares on Wednesday, was reported active in making a market, helped by J. Carlisle, the specialist in the stock. It used to be one of the best trading industrials on the board, and the present opportunity is considered favorable for reviving interest.”
“Halsted & Hodges, who were among the heavy sellers of stocks in the early trading of Wednesday, sold American Sugar and Amalgamated Copper in considerable quantity on the opening strength. Traders called this good selling, and took it to be long stock.”
Sat. 12/14/07 “American Sugar showed the severest break in the
industrial list, losing over 4 points. . .
Mon. 12/16/07 “Houses which used to do a good deal of business for lower Wall Street when American Sugar was the most active industrial on the list, have been selling that stock. Some of this is probably for shorts.
“There was renewed pressure in American Sugar at the outset, with selling by J. Carlisle, the specialist in the stock, CD. Barney & Co., and other influential commission houses. Boston has been selling the stock since H.O. Havemeyer’s death. This is a natural result, as investors there had great faith in the late president. One disturbing factor, however, in the market has been some selling from Washington in the past few days.
Wed. 12/18/07 “Trade interests seem to be the principal bears in American Sugar. One view is that the short crop will hurt earnings. The fact possibly that trade enemies of the company see a chance of retaliation for the Havemeyer attitude towards them in past years.”
Fri. 12/20/07 “American Sugar, after opening quiet, advanced a point and a half in the first ten minutes. Traders began to suspect from the beginning of the week that short selling was being encouraged, and a sharp demand for the stock in the loan crowd on Wednesday confirmed this view. The Washington houses have been borrowers of the stock.”
Mon. 12/23/07 “There was some disposition to test the strength of American Sugar, but J. Carlisle, the specialist in the stock, had buying orders, and seemed able to support it without much difficulty. It is well and widely held, and soon becomes oversold.”
Tues. 12/24/07 “Traders think that a little of the American Sugar bought to support in the recent decline has been coming out quietly. Lower Wall Street advices are bearish, but such speculators do not seem to have made much by their preliminary campaign after H.O. Havemeyer’s death.
Fri. 12/27/07 “With very little pressure Sugar disclosed a rather weak technical position. There is no doubt that the rally to par forced in most of the shorts. The bear talk still comes from lower Wall Street, and it looked as if a little long stock were coming to the market, to judge by its more frequent appearance in loans.”
“The selling of Sugar in the forenoon was taken as indicating that the New Street bear party was trying to make an active market. S.L. Blood & Co. and E. F. Hutton & Co. were conspicuous among the sellers, while the specialists supported.”
Sat. 12/28/07 “Bear points on American Sugar were again current. They have reference to the election of the new president on Jan. 8, when lower Wall Street believes some permanent difference among leading holders as to policy may make itself felt.”
Sat. 1/04/08 “One reason for the exceptional strength in Sugar is the real scarcity of the stock. It loaned at 1.64 premium in the first hour, and has been in demand flat in the loan crowd for some time past. It was always a dangerous short sale, and was oversold when H.O. Havemeyer died.”
Wed. 1/08/08 “American Sugar was 2 points lower at the opening on the violent attack by Claus A. Spreckles published in a morning paper. While the Street does not regard him as altogether impartial witness, traders admit that the attack is the most formidable made upon the stock in many years.”
Thurs. 1/09/08 “… Speculators who attacked American Sugar on the protests of Mr. Spreckles found a market which was too much for them, and probably took losses during the course of the day. ..”
2. Additional Articles – dates and titles as noted:
Tues. 12/03/07 “Answers to Inquires”
An investing advice column states: “… [American] Sugar stock is in danger of being adversely affected by an industrial slow-up, and also by agitation of the tariff question.”
Thurs. 12/05/07 “H. O. Havemeyer”
Announcement of the death of H. 0. Havemeyer: “… Under his control the American Sugar Refinery Co. has . made a remarkable record of profitable operation. Mr. Havemeyer’s methods imparted steadiness and efficiency to his corporation, but his policy was not that of publicity, so that he belonged to a passing rather than to a new era of corporate finance.”
“H. O. Havemeyer Dead”
Article downplays the impact of H. O. Havemeyer’s death:”While the removal of death of Henry O. Havemeyer, founder and president of the American Sugar Refining Co., for between sixteen and seventeen years, is a loss which is bound to be severely felt, the splendid organization which had been perfected under his leadership, in the opinion of those most familiar with its affairs, has gathered sufficient momentum to carry forward the business without interruption or diminution of prosperity. . .”
Tues. 12/10/07 “American Sugar Refining Co.: Havemeyer’s Death Likely to Result in Abandonment of Secrecy Policy”
Article predicts more open communication with stockholders.
Wed. 12/11/07 “Dividends and Meetings”
“The directors of the American Sugar Refining Co. have elected Horace Havemeyer, a director of the company to succeed his father, H. O. Havemeyer, deceased. Vice-President Thomas was appointed acting President until the next annual meeting, Jan 8.
Appropriate resolutions were passed by the board regarding the death of Mr. Havemeyer.”
“American Sugar Refining: Trade and Company Conditions Good”
“Secretary Heike, of the American Sugar Refining Co., says: ‘The condition of the sugar trade and of the company is very good. We are doing a normal business for this season of the year, and the general business recession has not as yet affected the sugar market. The sugar crop is a little short this year.’
In the sugar trade it is stated that there is no apparent cause for the selling of sugar stock. The American Sugar refineries are all running. Its sugar accumulations are sufficient for thirty days, and of these, 50,000 tons were bought by Mr. Havemeyer shortly before his death, at the lowest prices current for sugar for the year and as l/4c. below present values, which transaction netted a profit of $200,000.
The company’s purchases of sugar in October and November show their cost was below present value. The total stocks of all refiners is only 100,000 tons.
Mr. Thomas, the acting president, has been in the management for many years and has wide knowledge of the sugar trade.
It is said the financial results for the present fiscal year are the best that the company has yet experienced. The profits from the Beet Sugar industry along are said to have increased folly 25%.”
“American Sugar Refining Co. Again Center of Interest”
“Not for years has the American Sugar Refining stock been so active as during the past week when over 107,000 shares changed hands at prices ranging from 106 1/2 to 92 3/4, the latter being the lowest point recorded in years. Considerable interest has been aroused as to the causes of this activity, and all the more so, since the company’s officials maintain the trade is normal for this time of year, and that conditions, generally, are favorable. Injustice to the 17,000 stockholders it should be pointed out that in the sugar trade there is utmost confidence as to the future. The crop is a little short of last year’s crop, but on the other hand much better prices prevail.
For the fiscal year, about to close, it is believed, and there is sufficient ground for so doing, that the earnings will be larger that in any year heretofore. . . .
. . . One reason advanced for the marked decline in the stock is the loss of H. O. Havemeyer, which, if be true, is indeed a tribute to his ability as the real genius of the American Sugar Refining Co. The company was apparently too well organized for such to have a radical effect, though it undoubtably did occasion considerable selling.”
Wed. 12/18/07 “American Sugar”
“Many thousand shares of American Sugar common have been picked up by individual investors on the decline to 92 3/4, and in the cause of several Boston houses, Sugar buying last week was the investment feature. ..”
Fri. 12/20/07 “American Sugar”
“… The notice states that proxies will be voted for the re-election of Messrs. Parsons and Frazier and the election of Horace Havemeyer to succeed his father. It is the intention of the board, as Mr. Heike [Secretary], at the annual election of the officers after the stockholders meeting, to elect W. B. Thomas as president of the company.”
Tues. 1/07/08 “American Sugar Refining Co. Completes a Successful Year”
“The American Sugar Refining Co.’s past fiscal year, which will be reviewed at the annual meeting on Wednesday, Jan 8., was one of the best that the company has experienced. . .
With the inauguration of the new management it is supposed that the policy of the company will change, the keynote of the new administration being less antagonistic and less secret. It is presumed that the various lawsuits inherited from the Havemeyer regime will be adjudicated.
. . . The strength of the stock in the past week, during which it showed a net gain of seven points, is attributable to various causes. Primarily, the crop is smaller than last year’s. This in itself, is sufficient cause for the higher prices prevailing. . . Another bull point as far as the stock is concerned is the spread of the labor troubles in Cuba. Still another factor pointed to with considerable satisfaction by the sugar traders is the attitude of Secretary Taft taken in regard to the duty on sugar and the regulation of the supply from the Philippines. . .”
Wed. 1/08/08 “American Sugar”
“The officers of the American Sugar Refining Co. brand the Spreckel’s charges against the Sugar Co. as utterly absurd. Acting President Thomas says, ‘The story is, of course, untrue,’ and Secretary Heike is equally emphatic in denying the truth of the account. Friends of the American Sugar Refining Co. in the sugar trade admit that more publicity in regard to the company’s affairs would be better policy. The new management has already committed itself to more open and less antagonistic methods. President Thomas has given his word to this effect. It will however, take some time to bring about this result completely but the first step in this direction is expected at the annual meeting tomorrow, when President Thomas may in his review of the year disclose more information that has been given out heretofore.”
