• The Carnegie Steel Co., its dominant property, plus other holdings previously held on the books of the Steel Co., including:
• The Frick Coke Co. The Steel Co. previously held a 29.55% interest in the Coke Co., but the Carnegie Co. bought out the other shareholders shortly after it was organized, so its interest increased to 100%.
• A five-sixths interest in the Oliver Mining Co., which held large ore leases in the Mesabi ore range.
• A variety of railroads, steamship lines, and Great Lakes docks, many of them newly developed and used primarily by the Carnegie subsidiaries.
The Steel Company’s Earnings
Steel Co. earnings grew very strongly throughout the 1890s. Carnegie Steel was so much more productive than the competition that it was able to fight a rail price war in 1897, take significant market share, and still rack up record earnings, even as most of its competitors were booking losses. The earnings below include the Steel Co.’s earnings from its 29.55% share of the Coke Co. and from its other subsidiaries. For reference, the Coke Co.’s 1899 earnings were a record $4.2 million, of which $1.25 million accrued to the Steel Co.’s account. Earnings from the twelve other subsidiaries taken together were lower than from Coke.
TABLE 1
Annual Earnings—Carnegie Steel Co.
1893
$3,000,000
1894
4,000,000
1895
5,000,000
1896
6,000,000
1897
7,000,000
1898
11,500,000*
1899
21,000,000
Source: ACLC
The U. S. steel market grew at a 20–25% annual clip in 1898 and 1899, and steel prices rose sharply. In the first quarter of 1900, Carnegie was confidently forecasting $40–50 million steel profits for the year. But growth turned down in the second quarter, and the market virtually collapsed in the third quarter.
Annual financials for 1900 for either Carnegie Co. or the Steel Co. do not seem to have been preserved, but Steel Co. monthly earnings for the first eleven months of the year can be documented from board-level reports, while the missing December number can be inferred from Carnegie’s personal financials.
TABLE 2
Monthly Earnings, 1900—Carnegie Steel Co.
Jan
$3,638,642
Feb
3,541,679
Mar
4,700,032
Apr
3,219,879
May
2,381,127
Jun
1,850,047
Jul
978,102
Aug
641,662
Sep
470,441
Oct
940,446
Nov
361,857
Dec
1,046,086
Total
23,770,000
Average
1,980,833
Source: ACLC; author’s calculations
The Coke Co. was transferred to the Carnegie Co.’s (the holding company’s) books in April 1900. The Steel Co.’s earnings in Table 2 therefore reflect its 29.55% share of the Coke Co.’s earnings, or $511,600, only through March.
As the table shows, there was a dramatic profit collapse in the second half of the year. Total Steel Co. second half earnings were only $4.4 million, down from $19.3 million in the first half. The Homestead Works, which by itself booked almost $4 million earnings just in the first quarter, ran a loss of $160,000 in the four second half months for which there are plant-level figures. It might be noted that the Homestead losses were policy-driven, in the sense that a decision was made in July that, except for rails, they would “take all the business going at low prices.”
The 1900 Steel Market Break
The earnings outcomes are consistent with the very sharp break in the steel market during 1900. Table 3 shows Pittsburgh steel prices for a selection of standard products. Prices roughly doubled during 1899, stayed more or less steady at the outset of 1900, then gave up most of the 1899 increases by the third and fourth quarter. (All data from Iron Age.) The price collapse was apparently steeper than even these prices suggest. Iron Age noted in June 1900 that despite the “nominal Pittsburgh prices . . . the market is now an open one and . . . Pig Iron and Steel are both being offered at much lower prices.” In its year-end review, Iron Age summed up 1900 as a year with “long stretches of dullness . . . serious diminution of consumption . . . labor troubles of a most vexatious kind . . . and a sharp decline in prices.” A measure of the severity of the second half market break is that, after growing 24.8% in 1898 and 19.1% in 1899, and despite blistering growth for the first several months of 1900, total steel output declined by 4.2% over the full year.
TABLE 3
Selected Pittsburgh Steel Price Quotes
Jan-99
Jan-00
Feb-00
Mar-00
Apr-00
May-00
Jun-00
Jul-00
Aug-00
Sep-00
Oct-00
Nov-00
Dec-00
Steel
$16.25
$35.00
$33.00
$33.00
$33.00
$30.00
$28.00
$25.00
$19.00
$18.00
$16.50
$18.00
$19.75
billets
Wire rods
22.25
50.00
50.00
50.00
50.00
48.00
48.00
35.00
35.00
33.50
33.00
33.00
33.00
Steel bars
1.00
2.20
2.20
2.25
2.25
1.95
1.80
1.40
1.00
1.10
1.05
1.10
1.25
Beams
1.30
2.25
2.25
2.25
2.25
2.25
2.25
1.90
1.90
1.50
1.50
1.50
1.50
Source: Iron Age
Commentary at Carnegie Steel Co. board meetings underlines the seriousness of the crash. (Peacock and Bope are sales executives.)
May: (Operating Committee) “[Peacock] stated that Jones & Laughlin are practically shut down. The Illinois Steel Co., aside from the rail mill, are shut down. The National Steel Company are running about one-half of their different plants.”
June: (Schwab) “As near as I can find out there is no new business coming up in the Pig Iron or Steel business. All manufacturers that have talked with [me] tell me they have nothing to do and that things are too rapidly approaching a standstill.”
July: (Peacock) “Pig Iron Association says they have sold practically no Pig Iron for the past two months.”
September: (Peacock) “Bars are selling as low as 37½ cents per hundred pounds and plates are as low as we have ever known or heard them to be.”
October: (Bope) “As to general market, conditions are unchanged. It is a waiting market and not a great deal of business is being placed.”
Rail orders were so low that the board discussed shutting down the Edgar Thomson Works in November. Schwab merely cautioned to be sure that they had exhausted all their rail orders before they did so. Total Rail Pool allocations had shrunk two-thirds. I could not find a date for a shutdown, but there was a note in a December minute that the rail mills had “reopened” on December 5, when markets were showing signs of recovery. On July 28, Carnegie asked whether they should consider delaying interest on Carnegie Co. bonds to finance his investment
program. Schwab thought that was not necessary.
Carnegie’s Personal Balance Sheet
The eleven-month numbers retrievable from board reports track with Carnegie’s personal account statement for 1900, which shows an item “Earnings of Carnegie Steel” at $12.87 million. His custom in previous years was to book his equity share of Carnegie Steel’s earnings as personal earnings. At the start of the year, he held 58.5% of the company. Subtracting his transfers of shares to Schwab and several of the other partners, he would have been left with 54.14%, which suggests total earnings of $23.77 million for the year (and $1.05 million in earnings in December, which is consistent with the recovery getting under way at that time. I use that number in Table 2). Properly speaking, he shouldn’t have made that booking, since he had exchanged his shares for holding company shares, which did not flow through equity-method earnings. But this is a personal account, not a company account, and Rockefeller followed a similar practice, even after the 1911 Standard breakup.
The Carnegie Company’s Other Properties
I found a report listing the first half earnings of the holding company’s properties. It is not an earnings booking, but just an informational report, and includes 100% of the Coke Co. earnings, and those of other subsidiaries, for the first quarter (i.e., prior to their acquisition). They are set out in Table 4.
Carnegie Co. was the holding company, and did not flow through the earnings of its subsidiaries, instead reporting just dividends and interest accrued and received. (Corporate lawyers advised maintaining the fiction that the holding company did not exercise management control.) Since this document is obviously just a background document, it includes all earnings for the listed companies, including for periods prior to their ownership by the Carnegie Co., and there is, additionally, a certain amount of double-counting in the presentation. Since we also know the Steel Co. H2 earnings, I applied the H1 steel/non-steel earnings ratio to those to produce a first-order estimate of the full-year earnings for the entire company. (The Steel Co. was the primary, or sole, customer for all the other subsidiaries.) Some notes on the details follow:
1. Frick Coke included four subsidiaries—the Youghiogeney Railroad and the three water companies. The total of the Q1 Frick Coke entries and the four subsidiaries multiplied by the 29.55% Carnegie Steel ownership percentage produces the $511,601, already included in the “Frick” amount shown in the Carnegie Steel entries. All of the Q1 Frick and subsidiary earnings in the table, therefore, should be excluded from Carnegie 1900 earnings, since the owned portion has already been counted.
TABLE 4
Six-Month Earnings of Carnegie Co. Holdings, 1900*
Jan.
Feb.
Mar.
Apr.
May
June
Total
Carnegie Steel
$3,638,549
$3,591,365
$4,700,244
$3,219,872
$2,387,455
$1,823,241
$19,360,726
HC Frick Coke
687,061
593,066
428,946
803,373
700,250
647,305
3,860,001
Youghiogeney RR
2,388
2,475
3,006
3,766
2,987
2,626
17,248
Mt Pleasant Water
2,359
2,297
2,308
2,264
2,178
2,038
13,444
Youghiogeney Water
927
822
907
772
796
669
4,893
Trotter Water
1,720
1,510
1,513
1,566
1,440
1,322
9,071
Union Supply
29,824
46,421
29,728
30,648
30,692
28,125
195,438
Carnegie Natural Gas
80,162
71,094
82,607
103,061
103,082
91,897
531,903
Union RR
–41,687
–79,405
–65,573
–5,176
39,528
73,202
–79,111
Pitts, Bess, and
L Erie RR
–56,594
–63,866
–33,893
–46,113
74,518
85,774
–40,174
Pitts Steamship
0
0
–431
–12,813
119,885
75,068
181,709
Pitts and Conneaut Dock
–5,874
–4,215
–892
–9,610
71,407
39,652
90,468
Pitts Limestone Co.
4,229
3,200
2,595
2,531
2,337
2,039
16,931
Oliver Iron Mining
500,000
500,000
500,000
500,000
500,000
500,000
3,000,000
Totals
4,843,064
4,664,764
5,651,065
4,594,141
4,036,555
3,372,958
27,162,547
Source: ACLC
2. The remaining entries for the first quarter, except for the Oliver earnings, are all also included in the Carnegie Steel entries. They are, in any case, slightly negative (by $2,570). Taking the Oliver numbers at face value (but see my reservations below), they should be reduced by the 16.7% held by outside shareholders, or by $250,500. Adding the net Oliver to the Carnegie Steel entries produces Q1 earnings of $13.18 million.
3. Q2 I simply take as presented, less the $250,500 for external Oliver ownership, which produces a total of $11.75 million, for a total net first half of $24.93 million.
4. Including net Oliver and all Frick (and eliminating the $511,601 Frick double-count in the Steel Q1 earnings), the ratio of non-steel to steel earnings in the first half is 38.7%.
5. Second half steel earnings were $4.44 million. Assuming non-steel earnings in H2 were in the same ratio as in H1 produces an additional $1.72 million for a total of $6.16 million—or a grand total of $31.09 million for the year.
Having said that, I remain extremely skeptical of the claimed Oliver earnings, the more so since they look like a “plug” entry. The Steel Co. did not include Oliver earnings in its monthly profit bookings. I did find a year-end Oliver entry, for 1898. It was an estimated booking for the Steel Co.’s share of Oliver earnings for the year. Oliver had estimated its 1898 earnings to be $800,000, which the Steel Co. reduced to $600,000 “to be safe” (making an entry of $500,000 to reflect its 83.3 percent share). Even at the $800,000 figure, however, the ratio of ore to steel earnings was only half that suggested by Table 4. In addition, in 1899 the contract between Oliver and the Steel Co. was revised so non-Bessemer ore was transfer-priced at cost. That should have accounted for at least half of the Steel Co.’s output, which would have commensurately reduced Oliver’s earnings.
Furthermore, there are several other spring 1900 accounting entries that may shed light on the ore figures:
1. There are two offsetting entries of approximately $20 million increasing both payables and inventory for “ore at mines.” These are apparently related to a discussion at the board about Oliver’s problems with customers who were deferring shipments of previously ordered ore. The Steel Co., as a pre
cedent for other Oliver customers, agreed to be invoiced for ore at an earlier stage, instead of waiting until the ore had been delivered to a Lake dock. Oliver presumably booked the invoices as a sale, which would have inflated its earnings. (It would have had the effect of advancing future earnings into the current period.) Modern Profit-and-Loss accounting would have reduced Carnegie Steel’s earnings by a commensurate amount, since the invoice would be treated as an expense. But nineteenth-century companies did not keep P&Ls, but reported only balance sheets and balance-sheet changes. Since monthly Steel Co. earnings seem to track sales during this period, I strongly suspect that the big receivable increase at Oliver was not offset by an expense at Steel, which would have artifically inflated the total earnings of the holding company.
2. A further entry in May 1900 increases March steel earnings from the $4.7 million shown in this table and in previous board reports by about $350,000. That is the same way the ore earnings were booked in 1898, as a post-close item when Oliver got around to reporting them. If this number represents the Steel Co. share of Q1 Oliver earnings (I can’t think of what else it could be), it would be much closer to, although somewhat lower than, the 1898 ratio. (But that would be consistent with the transfer pricing of non-Bessemer ore.)
Since there seem to be good grounds for skepticism of the ore earnings, I make an alternative estimate simply by applying the 1898 ratio of ore to steel earnings, net of the amount held by external investors, ignoring the apparent agreement to transfer-price non-Bessemer ore. That would reduce projected full year ore earnings by $2 million.
The Tycoons: How Andrew Carnegie, John D. Rockefeller, Jay Gould, and J. P. Morgan Invented the American Supercompany Page 40