In October of 1999, Warren Buffett announced a deal to purchase approximately 76 percent of MidAmerican Energy for a per share price of $35.05 in cash, which was a 29 percent premium over the preannouncement closing price. Berkshire Hathaway acquired common shares as well as preferred convertible shares and fixed income securities in this deal, which was structured to give Berkshire Hathaway an economic interest of 76 percent in the company but a voting interest of only just under 10 percent of the company. This structure was adopted in part to avoid breaching any of the complex regulations related to the Public Utility Holding Company Act of 1935 (PUHCA). This act significantly restricted the business activities of holding companies of public utilities, and presumably if Berkshire had a larger voting interest than 10 percent, it would have been subject to these restrictions. In total, Berkshire Hathaway paid approximately $2 billion in this transaction, and there were also two notable coinvestors: Walter Scott, who had been on the board of Berkshire Hathaway since 1988 and had introduced the deal to Buffett, and David Sokol, the entrepreneurial CEO of MidAmerican at the time.1
Diving right into the investment case, if I were a potential investor at the time, I would have asked two main questions about the company: Is this a quality business? And is the company for sale at a good price? The 1998 year-end report of MidAmerican is a good place to start because that is the latest annual report that would have been available at the time when the deal was announced in late 1999.
As the annual report shows, MidAmerican was a diversified energy company that was involved foremost with electricity production but also with electricity distribution and “up-stream” gas field exploration. According to the segmental reporting in the notes to financial statements, MidAmerican broke its business into three core business segments and its corporate overheads (see Table 18.1).
Overview of business segments
| Segment |
Revenue |
Operating income |
Comments |
| Domestic generation |
$583 million |
$314 million |
|
| Foreign generation |
$224 million |
$143 million |
Primarily in Philippines |
| Foreign utility |
$1843 million |
$173 million |
Primarily in United Kingdom |
| Corporate |
$33 million |
−$10 million |
|
| Total |
$2683 million |
$619 million |
|
The annual report also includes a list of the company’s generation plants of the company at year-end 1998.
Table 18.2.
Summary of generation plants (1998)
aThe company operates all such projects other than Teesside Power Limited, Quad Cities Power Station, Ottumwa Generation Station, and Desert Peak.
bTable 18.2 excludes three projects in Indonesia, two of which are currently in arbitration. One unit became operational in March 1998.
cActual MW may vary depending on operating and reservoir conditions and plant design. Facility net capacity (in MW) represents facility gross capacity (in MW) less parasitic load. Parasitic load is electrical output used by the facility and not made available for sale to utilities or other outside purchases. Net MW owned indicates current legal ownership, but, in some cases, does not reflect the current allocation of partnership distributions.
dPNOC-Energy Development Corporation (PNOC-EDC); Government of the Philippines (GOP) and Philippine National Irrigation Administration (NIA) (NIA also purchases water from this facility); Northern Electric plc (Northern). The Government of the Philippines undertaking supports PNOC-EDC’s and NIA’s respective obligations. Southern California Edison Company (Edison); San Diego Gas & Electric Company (SDG&E); Utah Power & Light Company (UP&L); Bonneville Power Administration (BPA); New York State Electric & Gas Corporation (NYSEG); Texas Utilities Electric Company (TUEC); Niagara Mohawk Power Corporation (NIMO); and MidAmerican Energy Company (MEC).
As we can see, the business segment responsible for the largest portion of income was domestic generation, with most of the U.S. generation plants in Iowa and Illinois. These plants encompass coal, gas, geothermal, hydro, and nuclear power plants. The foreign power plants, which include three power plants in the Philippines and two in England, constitute the assets of the foreign generation segment.
The domestic and foreign generation businesses were similar in their operations. In both, the core activity was building and operating power plants—purchasing fuel, such as coal, and selling electricity. There are nuances to what differentiates more-profitable from less-profitable power plants—their substitute fuels and costs, their technologies, and their efficiency—but the largest difference overall between the two businesses was in government regulation. Power generation is one of the most heavily regulated industries across the world, and as a result, a generation company’s profitability is tied substantially to the regulatory environment of the country in which it operates.
In the United States at the time, there were several regulatory agencies and policies that were relevant for MidAmerican. At the national level, since the passage of the Public Utilities Regulatory Policies Act (“PURPA”) just prior to 1980, independent energy producers were encouraged, and utilities had to buy their electricity. The regulation also extended to price levels. At the state level, there were also numerous regulations, often related to national policies. For example, in Iowa, direct profit regulation stipulated that if MidAmerican’s annual jurisdictional return on common equity exceeded 12 percent, some of those profits had to be shared with customers. Likewise, unless MidAmerican’s annual return on common equity fell below nine percent, it would not be able to increase its prices.
In England, the regulation of the power generation industry included a full market for electricity trading called the “Pool.” Since the phase-in of the Electricity Act of 1989, virtually all electricity generated in England and Wales had to be purchased and sold through this Pool, which set prices. This meant that even if a company produced electricity and sold it as a utility, it had to sell its electricity to the Pool at a specified price and repurchase it from the Pool in order to resell to its utility customers. This and several other measures made price regulation prevalent in England.
Overall, the effect of governmental regulation—both domestic and abroad—was to confine profitability to a band of healthy but not stellar numbers. Within the timeframe of a stipulated contract, an energy company could increase profits if they could increase efficiency, but only for so long. Once regulatory agencies noticed these high profits, new regulation would set in.
The company’s final two business segments were foreign utility and corporate. Foreign utility was dominated by the Northern Electric Distribution Limited, which was a distributor of electricity in England. In 1998, this business had approximately 17,000 km of overhead and 26,000 km of underground electricity cables and had rights to supply electricity to approximately 1.5 million customers. In addition to being a distributor of electricity, Northern also supplied and distributed natural gas. Besides Northern, MidAmerican had other businesses, including CE Gas UK Limited, that were involved in gas exploration and production. Hence, abroad, especially in England, MidAmerican had a traditional utility business, including the full value chain of energy supply from exploration to generation to distributing and selling to customers. The final segment of the business, corporate, included corporate functions like legal and finance, but also several niche businesses like Homeservices, which was a realty business belonging to the overall MidAmerican entity.
Turning to the financial statements, in 1998 MidAmerican reported net income of $127 million on a revenue basis of $2.55 billion for the holding company. On a diluted per share basis, this represented $2.01 per share for the 74.1 million shares on a diluted basis outstanding at year-end 1998. Calculating for an operating profit, the EBIT is $491 million. This excludes $220 million in net interest expense, a $93 million provision for income taxes, a $41 million minority interest charge, and $11 million in extraordinary charges. Given the extent of net interest expenses (close to double the net income), the EBIT is probably more reflective of the inherent earnings ability of the business than the net income. In margin terms, EBIT margin was 19.2 percent, and net income margin was five percent.
As a potential investor in the company, I would also consider the capital intensity and return economics of the business and would have calculated for the ROTCE in order to do this. Here, taking an EBIT of $491 and subtracting the 35 percent corporate tax prevalent in 1998, the resulting NOPAT would have been $319 million. The tangible equity base would have looked as follows:
Table 18.3.
| Category |
Amount in $ |
As a % of revenues |
| PPE |
$4236m |
166% |
| Inventories |
— |
0% |
| Accounts receivable |
$528m |
21% |
| Accounts payable |
−$306m |
−12% |
| Total capital employed (TCE) |
$4458m |
174% |
As one can see, the capital employed in the power generation and utility business from MidAmerican is quite significant. In total it is 174 percent of revenues, which is considerably higher than in most production businesses. The resulting ROTCE, based on this TCE and the NOPAT of $319 million, is 7.2 percent. This is fair, but not exceptional. Considering that some of this PPE at year-end was on the balance sheet but did not contribute to the earnings in that year, a less-conservative TCE of year-end 1997, $3731 million, would also be reasonable. Based on this figure, the resulting ROTCE would have been 8.5 percent—better, but still not very good. As an investor, I would have concluded that the ROTCE economics of the MidAmerican business were good but certainly not great.
Of course, there are additional factors to consider. Given that MidAmerican is a utility business with a high level of consistency, it likely had access to very cheap capital and thus was able to generate better returns than suggested by its ROTCE economics. Calculating the ROE is a good way to test this hypothesis. Taking the net income of $127 million on a common equity base of $827 million, the ROE of MidAmerican is 15 percent. This indeed seems to support the hypothesis that MidAmerican was able to achieve fairly good returns from the business at least in part due to its use of financial leverage of reasonably priced capital. Overall, I would conclude that while MidAmerican was a company that was better operated than its peers and also had higher growth, its core business was not inherently a superb compounder.
Finally, valuation: Given the $35.05 per share price that Berkshire Hathaway purchased MidAmerican at, the valuation would have been as follows:
Table 18.4.
EV/EBIT multiples
| EV/EBIT |
1998 |
1997 |
| EV* |
$7867m |
$7867m |
| EBIT |
$491m |
$343m |
| EV/EBIT |
16.0× |
22.9× |
*Calculation of EV is based on share price of $35.05 × 72.64m diluted shares outstanding at June 30, 1999 plus net debt at June 30, 1999 of $5321m. Net debt is calculated based on figures of $247m cash + $130m marketable securities + $385m restricted cash + $190m equity investments − $2017m parent co. debt − $ 4256m project debt. Note that this sum based on the latest available 10Q report is somewhat lower than the approximate $9 billion EV quoted in the press release document for the acquisition. Two factors explain the divergence: (1) changes in net debt between June 30 and press release date (2) calculation of the value of preferred shares.
Table 18.5.
PER multiples
| PER |
1998 |
1997 |
| Share price |
$35.50 |
$35.50 |
| EPS as reported (diluted) |
$2.01 |
Negative |
| PER |
17.4× |
Not applicable |
These multiples seem extremely high, especially given the previous analysis that MidAmerican was not a great compounder. To be sure, there were one or two businesses within MidAmerican (such as the realty business) that had a clear value but were not considered in the valuation, but unless the earnings were at cyclically depressed levels, MidAmerican would not have seemed undervalued to me as a potential investor at the time. Given the seven percent ROTCE and 15 percent ROE achieved in 1998, I would have considered cyclically depressed earnings unlikely; as power generation clearly seemed to be capital intensive, I would have guessed cyclical high ROTCE to be no more than about 12 or 13 percent. Assuming that in a very good year earnings were 50 percent higher on the same asset basis, the comparable EV/EBIT multiple would have been approximately 11 times, and the PER would have been approximately 12 times—hardly dirt cheap.
The only aspect that would have seemed impressive to be me would have been the growth. Between 1994 and 1998, as the selected financial data show, revenues grew from $154 million to $2.5 billion. Similarly, net income grew from $37 million to $127 million during this period. The commentary in the annual report makes it clear that growth was both organic and acquisitions based. All in all, I would have seen MidAmerican as a stable business with good growth, but only average returns. I would not have seen it as an especially attractive investment, at least with respect to the common equity.
So what might Buffett have seen differently in this case? The October 25, 1999 press release from Berkshire Hathaway about this acquisition offers clues. First, the transaction was not structured as a simple acquisition through the purchase of all common stock. To keep the percentage of voting rights low for regulatory purposes, Berkshire Hathaway was to invest approximately $1.25 billion in common stock and nondividend-paying convertible stock and a further $800 million in a nontransferable-trust preferred stock. As is revealed in the Berkshire annual letter to shareholders, the $800 million trust preferred stock was equivalent to a fixed income security with an 11 percent yield and was considered as such by Buffett. Given this, one can view Buffett’s investment in MidAmerican as one in which he only commits $1.25 billion as equity, with the benefit of having an 11 percent fixed-income product with relatively little risk. While the setup is somewhat complicated, this transaction seems akin to a private equity transaction. Buffett and his associates do not purchase the 72 or so million diluted shares outstanding2 at $35.05 per share, which would have cost approximately $2.5 billion. Buffett only pays $1.25 billion, gets to invest in a fixed-income product yielding 11 percent, and gets a 76 percent ownership of the earnings of the company. The remaining cash is made up with debt in the form of bonds and a smaller $300 million investment by coinvestors Walter Scott and David Sokol. In this setup, it seems that Buffett benefits twice. First, he enjoys a rather attractive fixed-income investment; second, he is able to benefit from leveraging a fairly stable growth business, resulting in higher returns on his equity in comparison to the moderate ROTCE inherent in the business. While difficult to quantify, it seems that this deal is significantly better than it would be for an ordinary investor investing in the common equity of MidAmerican. It also seems clear that avoiding regulatory requirements was not the only reason the deal was structured in this complex manner.
Another point that seems essential for Buffett in this investment case is the focus on the management team and the board. In describing the acquisition of MidAmerican, Buffett states, “If I only had two draft picks out of American business, Walter Scott and David Sokol are the ones I would choose for this industry.”3 This seems to be a case where Buffett was also investing in two partners he believed to be superb managers. MidAmerican’s annual report does not give a strong sense of the management team, but from the company’s history of growth it is clear that management had been competent. Buffett clearly benefited from his personal knowledge of the capacity of the management team.
Reading the commentaries of Buffett about the business in later years,4 I was struck by Buffett’s keen eye for the numbers. Specifically, he comments that MidAmerican carries a significant amount of amortization of goodwill (also called purchase price allocation), which was going away. The 1998 annual report indeed includes a figure of goodwill amortization, but the figure is hidden in the income statement under the line “depreciation and amortization.” Although only $42 million in 1998, the goodwill amortization should not be counted as a real cost since it is an accounting cost only rather than a cost that a business requires in continuing its operations. Considering this factor, the true EBIT minus PPA would have been $533 million rather than $491 million—about 10 percent higher. The resulting EV/EBIT multiple would change from 16.0 times to an EV/EBIT minus PPA multiple of 14.6 times—still high but lower than before.
Taking these factors together, Buffett seems to have invested foremost in a team of managers whom he trusted and believed would continue to grow the business. He also invested in a specific setup that was likely significantly better than a private investor would have gotten if that investor had purchased on the common equity at the same price. However, in terms of price, which is a bit lower than what I would have focused on, Buffett seems to be paying a full price in comparison to the previous earnings of the MidAmerican business. Perhaps it is true that, during 1999, which was the tail-end of a large run of stock prices, Buffett had to pay more than previously and was prepared to pay for a business that would be able to employ a large amount of capital at a decent, albeit not great, return.
Five-year financial summary (1994–1998)
($ in thousands)
Source: MidAmerican Energy Holdings Co., 10K Report 1998, 61.
aReflects the acquisition of KDG.
bReflects the acquisitions of Northern, Falcon Seaboard, and the Partnership Interest owned for a portion of the year.
cReflects the acquisition of Magma Power Company owned for a portion of the year.
*Includes the $87,000, $1.29 per basic share, $1.27 per diluted share, nonrecurring Indonesian asset impairment charge.
Balance sheet (1997–1998)
Years ended Dec. 31
($ in thousands)
| Assets |
1998 |
1997 |
| Cash and cash equivalents |
1,604,470 |
1,445,338 |
| Joint venture cash and investments |
1,678 |
6,072 |
| Restricted cash |
515,231 |
223,636 |
| Restricted investments |
122,340 |
— |
| Accounts receivable |
528,116 |
376,745 |
| Properties, plants, contracts, and equipment, net |
4,236,039 |
3,528,910 |
| Excess of cost over fair value of net assets acquired, net |
1,538,176 |
1,312,788 |
| Equity investments |
125,036 |
238,025 |
| Deferred charges and other assets |
432,438 |
356,112 |
| TOTAL ASSETS |
9,103,524 |
7,487,626 |
| Liabilities and stockholders’ equity |
| Liabilities |
| Accounts payable |
305,757 |
173,610 |
| Other accrued liabilities |
1,009,091 |
1,106,641 |
| Parent company debt |
2,645,991 |
1,303,845 |
| Subsidiary and project debt |
3,093,810 |
2,189,007 |
| Deferred income taxes |
543,391 |
509,059 |
| Total liabilities |
7,598,040 |
5,282,162 |
| Deferred income |
58,468 |
40,837 |
| Company—obligated mandatorily redeemable convertible preferred securities of subsidiary trusts |
553,930 |
553,930 |
| Preferred securities of subsidiary |
66,033 |
56,181 |
| Minority interest |
— |
134,454 |
| Common stock and options subject to redemption |
— |
654,736 |
| Stockholders’ equity |
| Common stock, par value $0.0675 per share* |
5,602 |
5,602 |
| Additional paid-in capital |
1,233,088 |
1,261,081 |
| Retained earnings |
340,496 |
213,493 |
| Accumulated other comprehensive income |
45 |
−3,589 |
| Common stock and options subject to redemption |
— |
−654,736 |
| Treasury stock—23,375 and 1,658 common shares at cost |
−752,178 |
−56,525 |
| Total stockholders’ equity |
827,053 |
765,326 |
| TOTAL LIABILITIES AND STOCKHOLDERS’ EQUITY |
9,103,514 |
7,487,626 |
Source: MidAmerican Energy Holdings Co., 10K Report 1998, 73.
*Authorized 180,000 shares, issued 82,980 shares, outstanding 59,605 and 81,322 shares, respectively.
Income statement (1996–1998)
Year ended Dec. 31
($ in thousands except per share data)
| |
1998 |
1997 |
1996 |
| Revenue: |
| Operating revenue |
2,555,206 |
2,166,338 |
518,934 |
| Interest and other income |
127,505 |
104,573 |
57,261 |
| Total revenues |
2,682,711 |
2,270,911 |
576,195 |
| Costs and expenses: |
| Cost of sales |
1,258,539 |
1,055,195 |
31,840 |
| Operating expense |
425,004 |
345,833 |
132,655 |
| General and administration |
46,401 |
52,705 |
21,451 |
| Depreciation and amortization |
333,422 |
276,041 |
118,586 |
| Loss on equity investments in Casecnan |
— |
5,972 |
5,221 |
| Interest expense |
406,084 |
296,364 |
165,900 |
| Less interest capitalized |
−58,792 |
−45,059 |
−39,862 |
| Nonrecurring charge—asset valuation impairment |
— |
87,000 |
— |
| Total costs and expenses |
2,410,658 |
2,074,051 |
435,791 |
| Income before provision for income taxes |
272,053 |
196,860 |
140,404 |
| Provision for income taxes |
93,265 |
99,044 |
41,821 |
| Income before minority interest |
178,788 |
97,816 |
98,583 |
| Minority interest |
41,276 |
45,993 |
6,122 |
| Income before extraordinary item and cumulative effect of change in accounting principle |
137,512 |
51,823 |
92,461 |
| Extraordinary item, net of tax |
−7,146 |
−135,850 |
— |
| Cumulative effect of change in accounting principle, net of tax |
−3,363 |
— |
— |
| Net income/loss available to common stockholders |
127,003 |
−84,027 |
92,461 |
| Per share |
| Income before extraordinary item and cumulative effect of change in accounting principle |
2.29 |
0.77 |
1.69 |
| Extraordinary item |
−0.12 |
−2.02 |
— |
| Cumulative effect of change in accounting principle |
−0.06 |
— |
— |
| Net income/loss |
2.11 |
−1.25 |
1.69 |
| Per share—diluted |
| Income before extraordinary item and cumulative effect of change in accounting principle |
2.15 |
0.75 |
1.54 |
| Extraordinary item |
−0.10 |
−1.97 |
— |
| Cumulative effect of change in accounting principle |
−0.04 |
— |
— |
| Net income/loss |
2.01 |
−1.22 |
1.54 |
Source: MidAmerican Energy Holdings Co., 10K Report 1998, 74.
Cash-flow statement (1996–1998)
Years ended Dec. 31
($ in thousands)
| |
1998 |
1997 |
1996 |
| Cash flows from operating activities: |
| Net income/loss |
127,003 |
−84,027 |
92,461 |
| Adjustments to reconcile net cash flow from operating activities: |
|
|
|
| Nonrecurring charge-asset valuation impairment |
— |
87,000 |
— |
| Extraordinary item, net of tax |
7,146 |
— |
— |
| Cumulative effect of change in accounting principle |
3,363 |
— |
— |
| Depreciation and amortization |
290,794 |
239,234 |
109,447 |
| Amortization of excess of cost over fair value of net assets acquired |
42,628 |
36,807 |
9,139 |
| Amortization of original issue discount |
42 |
2,160 |
50,194 |
| Amortization of deferred financing and other costs |
21,681 |
31,632 |
11,212 |
| Provision for deferred income taxes |
34,332 |
55,584 |
12,252 |
| Income on equity investments |
−10,837 |
−16,068 |
−910 |
| Income/loss applicable to minority interest |
5,313 |
−35,387 |
1,431 |
| Changes in other items: |
| Accounts receivable |
−135,124 |
−34,146 |
−13,936 |
| Accounts payable, accrued liabilities, and deferred income |
−41,803 |
29,799 |
2,093 |
| Net cash flows from operating activities |
344,538 |
312,588 |
273,383 |
| Cash flows from investing activities: |
| Purchase of KDG, Northern, Falcon Seaboard, Partnership Interest, and Magma, net of cash acquired |
−500,916 |
−632,014 |
−474,443 |
| Distributions from equity investments |
17,008 |
23,960 |
8,222 |
| Capital expenditures relating to operating projects |
−227,071 |
−194,224 |
−24,821 |
| Philippine construction |
−112,263 |
−27,334 |
−167,160 |
| Indonesian construction |
−83,869 |
−146,297 |
−76,546 |
| Acquisition of UK gas assets |
−35,677 |
— |
— |
| Domestic construction and other development costs |
−36,047 |
−12,794 |
−73,179 |
| Decrease in short-term investments |
1,282 |
2,880 |
33,998 |
| Decrease/increase in restricted cash and investments |
20,568 |
−116,668 |
63,175 |
| Other |
−33,787 |
60,390 |
−2,910 |
| Net cash flows from investing activities |
−990,772 |
−1,042,101 |
−713,664 |
| Cash flows from financing activities: |
| Proceeds from sales of common and treasury stock and exercise of stock options |
3,412 |
703,624 |
54,935 |
| Proceeds from convertible preferred securities of subsidiary trusts |
— |
450,000 |
103,930 |
| Proceeds from issuance of parent company debt |
1,502,243 |
350,000 |
324,136 |
| Repayment of parent company debt |
−167,285 |
−100,000 |
— |
| Net proceeds from revolver |
— |
−95,000 |
95,000 |
| Proceeds from subsidiary and project debt |
464,974 |
795,658 |
428,134 |
| Repayments of subsidiary and project debt |
−255,711 |
−271,618 |
210,892 |
| Deferred charges relating to debt financing |
−47,205 |
−48,395 |
−36,010 |
| Purchase of treasury stock |
−724,791 |
−55,505 |
−12,008 |
| Other |
21,701 |
13,142 |
10,756 |
| Net cash flows from financing activities |
797,338 |
1,741,906 |
757,981 |
| Effect of exchange rate changes |
3,634 |
−33,247 |
4,860 |
| Net increase in cash and cash equivalents |
154,738 |
979,146 |
322,560 |
| Cash and cash equivalents at beginning of year |
1,451,410 |
472,264 |
149,704 |
| Cash and cash equivalents at end of year |
1,606,148 |
1,451,410 |
472,264 |
| Supplemental disclosures |
| Interest paid (net of amounts capitalized) |
341,645 |
316,060 |
92,829 |
| Income taxes paid |
53,609 |
44,483 |
23,211 |
Source: MidAmerican Energy Holdings Co., 10K Report 1998, 76.
