The Washington Post newspaper was founded in 1877 by Stilson Hutchins. The paper was bought and sold to several private owners in the ensuing decades, including John McLean, who also owned the Cincinnati Enquirer. After being mismanaged by McLean’s son, the Washington Post found itself in insolvency by the 1930s.
In 1933, the company was purchased out of bankruptcy by Eugene Meyer, whose son-in-law, Philip Graham, and later his daughter, Katharine Graham, played major roles in the paper’s history. The Washington Post Company became a publically listed company in 1971. When the Chairman, Fritz Beebe, died unexpectedly in 1973, Katharine Graham became the first female chairman of a Fortune 500 company.
It was during this timeframe (1972–1973) that Warren Buffett began accumulating shares in the company. By late 1973, Buffett had accumulated an ownership of about 10 percent.1
The annual reports of the Washington Post Company for the years ending 1971 and 1972 represent the information a potential investor would have had about the company, immediately before Buffett purchased into it. I have included the four pages of consolidated financial statements at the end of the chapter.
Figure 8.1.
First, we must examine the businesses. The Washington Post Company had three major business divisions.2 I have summarized in table 8.1 the relative importance of each division to the overall company based on their financials for 1972.
Table 8.1.
Overview of business segments (1972)
| Business segment |
Revenues (mio) |
% total revenue |
% total EBIT |
EBIT margin |
| Newspaper publishing |
$99.8 |
46% |
47% |
10.2% |
| Magazine and book publishing |
$93.8 |
43% |
26% |
6.0% |
| Broadcasting |
$24.3 |
11% |
27% |
24.4% |
| Total |
$217.9 |
100% |
100% |
10% |
Source: The Washington Post Company, 1972 Annual Report, 2.
As we can see, at the time the newspaper division was clearly still the most significant division of the business, accounting for approximately half of total company earnings. However, the other two divisions were certainly financially relevant.
Reading the annual report more closely, one quickly realizes that it is more informative than most reports, full of discussions of operational metrics that matter. To me it indicates a management team that was focused and knew what it was doing.
The report starts with John S. Prescott’s discussion of the newspaper business. Specifically, he details a structural change that occurred in the market that year when the Washington Daily News ceased publication in July 1972. For an investor interested in the company, this is an extremely significant event. Foremost, it removed one out of three daily papers in the D.C. market, which automatically forced customers who still wanted to receive a paper to turn to one of the remaining two papers. In addition, a duopoly is a much preferred situation for a company to be in than an oligopoly with three players—at least from my own experience observing competitive dynamics. Usually competitive action is much more rational, resulting in fewer price wars and higher margins.
As for operational metrics, the factual data are overwhelmingly positive. Lines of advertising increased by 6.5 million—an increase of nearly nine percent based on prior year total lines of about 73 million. Quoting market share, this represented 63 percent of all advertising lines placed in metropolitan papers in its area. Circulation had increased by 6,000 for the daily paper (an increase of about one percent) and by 15,000 for the Sunday edition (an increase of about two percent). While these numbers are not astounding as percentages, the report points out that they are significantly more than the numbers posted by its main remaining competitor, the Star-News. This implies that the Washington Post was gaining market share. Prescott goes on to reference a market study conducted by an independent agency, W.R. Simmons & Associates, that determined that three out of five—or about 60 percent—of all adults in the Washington Post’s market read the Post. He goes on to calibrate this number compared to other newspapers with leading market positions in the country. I find this exercise very compelling; not only does John Prescott show that the Washington Post Company is indeed a great franchise, but he also demonstrates that he is a manager who steers by key metrics and is attuned to his competition. In sum, the core newspaper business of the Washington Post Company clearly was a great business.
The second business division discussed is the magazine and book publishing division. Its primary asset was Newsweek magazine, which at the time ranked fourth among all magazines in advertising revenue. Osborn Elliott, the divisional head, says that 1972 was a record year for Newsweek. Revenues rose eight percent year-on-year, and this was accompanied by a circulation growth of 125,000 per week (about five percent) from 2,600,000 per week to 2,750,000 per week. As is seen elsewhere in the report, an independent third-party source is cited when presenting these data; in this case, it was the Publishers Information Bureau.
Operationally, 1972 marked the year when the veteran Newsweek management team, which had been very successful during the sixties, resumed management responsibility of the paper. This strengthening of the divisional management included Elliott himself taking over the editorship, while Gibson McCabe and Robert Campbell were placed to lead operations and publishing, respectively. Their focus on providing content that their readers valued resulted in Newsweek winning 11 major journalism awards during the year, including the American Bar Association Gavel, the G.M. Leob Award, and the Overseas Press Club Award. Along with a final discussion on the opportunity that lay ahead for Newsweek in terms of continued expansion abroad, the Newsweek business, like the Post, seemed to be churning on all cylinders.
The last business within the Washington Post Company was the Post–Newsweek Stations (broadcasting). Larry Israel, who was responsible for this division, had less specific market share data and third-party accolades to present. Nevertheless, again one receives a very positive picture of the business. Israel speaks about one successfully completed acquisition (of WTIC-TV, Hartford, Connecticut) but focuses the discussion on the excellence of the stations’ local area generation of media content.
However, in this division there was also some negative news among all the positives. There were two competitive challenges to license renewals for two of the division’s TV stations (WJXT-TV in Jacksonville and WPLG-TV in Miami). Israel tried thoughtfully to dismiss the real risk of this challenge by citing that both previous examples of such challenges and the legal principles involved in judging such cases pointed toward a positive resolution for the group. Had I been I an investor, unaware of the inner workings of broadcasting regulation, I would not have been completely convinced by this explanation. This would be an area where I would have to assess the risk.
Aside from this potential structural issue, we can quantify the success of the division by looking at the divisional numbers on page two of the 1972 annual report. The broadcasting division in 1972 increased its revenues to $24.3 million from $20.8 million year-on-year (an increase of 17 percent) and its reported operating income to $5.9 million from $3.8 million year-on-year (an increase of 55 percent). This is an astounding financial performance for a business one might otherwise consider relatively stable with the same set of TV channels and radio stations. Although one large acquisition of WTIC-TV was discussed, this was being finalized in January 2013, so I am assuming that since no other acquisition was discussed in the 1972 divisional review of broadcasting the aforementioned divisional financial growth is mostly if not entirely organic.
When synthesizing the performance at the overall group level, the total operational performance seems so positive that, as a contemporary investor, I would frankly have had a hard time believing it. If it were not for the very factual and independent market data presented in the newspaper and magazine divisions, I would indeed be feeling very suspicious that the management was being overly optimistic in their reporting and would check this by turning to the hard numbers.
The first item of note is the 10-year review of the Washington Post presented on pages 22 and 23 of the annual report. In every one of the years, the revenue figures increased from the previous year. Over the total 10-year period, revenues increased to $217.8 million from $85.5 million, which comes to a figure of about 11 percent per annum. On the reported operating income, the picture is a bit more volatile, though also realistic because not every single year is a positive. Here, if we take the 10 years in total, operating income grew to $21.8 million from $9.4 million. The corresponding annual rate of growth is about 10 percent, slightly behind the top line growth. Although this means that margins have not increased despite the obvious gain in scale from a growing business, I would not see this as negative. If one of my concerns was that 1972 was a year of peak margins (operating margin grew to 10 percent from 7.9 percent in 1971) I would be reassured to see that during the 10-year period, operating margin varied between 7.8 percent and 12.2 percent, placing the 10 percent margin in 1972 right in the middle of the range. This means that if there is a capable management in place—and there are some indications in the report that this is the case—there is potential for a margin increase with higher operational efficiency. In fact, because the business in 1972 had even more scale than in 1968, it could be reasonably argued that the potential margin under a good management should be even higher than the 12.2% achieved in 1968.
Turning to the financial review and financial statements of the Washington Post, two items require further scrutiny and would be areas of concern for me as a potential investor. First, there is the section on retirement accounts. Like other media companies at the time, The Washington Post had defined benefit retirement liabilities to its employees. This, which is still a large issue especially with European companies, creates uncertainty and could result in repeated demands on cash if not carefully managed. On page 14 of the annual report, the reader is told that in general “the accrued costs and liabilities of these plans are fully funded.” But later in the section, there is one set of plans associated with the newspaper dealer incentive plans that is unfunded and whose costs are charged directly to current expenses. The expense under this set of plans was $3.6 million in 1972. This is not completely insignificant, but overall, without further inspection, the retirement plans seem to be a contained issue.
Second, there are capital stock and stock options, which are discussed on page 16 of the 1972 annual report. Here we see that in 1971, along with the initial public offering, the company adopted a stock option plan, for which 365,000 shares of common stock were reserved, out of which 279,650 were subject to options outstanding and 64,175 were available to be allocated at a later date. Although there were already options outstanding prior to 1971 and there were also two classes of common shares to complicate things further, to simplify we should look only at the 279,650 options outstanding at the end of 1972, and when they became exercisable.
Table 8.2.
Analysis of stock options (1972)
| Date exercisable |
No. of shares |
Avg. option price |
Effect of dilution (% of outstanding shares)* |
| 1972 EOY |
65.625 |
$26 |
1.4% |
| 1973 EOY |
71.300 |
$26 |
1.5% |
| 1974 EOY |
69.800 |
$26 |
1.4% |
| 1975 EOY |
69.050 |
$26 |
1.4% |
| … |
… |
… |
… |
| 1960 EOY |
3.875 |
$26 |
0.1% |
| Total |
214.025 |
~$26 per share |
5.8% |
*Calculation is based on using the weighted average number of common shares outstanding in 1972 or 4,806,802.
As defined by the Association of British Insurers (ABI),3 general good practice is not to allocate more than 10 percent of share capital in 10 years, that is, one percent per year. Although the figure between 1972 and 1975 exceeds this general rule of thumb, which would make me a bit wary as a potential investor, it is not unusual to give a few more options with an IPO, which is the case here. What is reassuring is that the IPO price was $25.18 per share,4 and the average option price is struck at about $26.00 per share, slightly higher than the IPO price and around or above the market price at the time. So overall, these observations would ease my concerns.
Finally, I turn to the financial statements, which can be found on pages 17 to 21 of the 1972 annual report. To assess the quality of the overall Washington Post business, an investor could calculate the returns on tangible capital employed (ROTCE) because this is a proxy for the ability of the business to generate returns above the cost of comparable businesses.
Table 8.3.
Analysis of total capital employed (1972)
| Category |
Capital employed (mio) |
As a % of revenues |
| PPE |
$46.2 |
21.2% |
| Intangibles excluding goodwill |
$0* |
0.0% |
| Inventories |
$3.8 |
1.7% |
| Accounts receivable |
$25.8 |
11.6% |
| Accounts payable |
−$19.4 |
−8.9% |
| Total capital employed (TCE) |
$56.4 |
25.6% |
*Based on The Washington Post’s definition of its balance sheet item “Goodwill and other intangibles” as “the excess of acquiring subsidiary companies over the related fair values of tangible assets at the dates of acquisition” on page 14, I have assumed all goodwill and other intangibles to be goodwill and none to be intangible assets that should be amortized.
If we take net operating profit after tax (NOPAT) but before exceptional items of $10 million as the basis for the calculation of ROTCE, we come to a ROTCE of 17.9 percent based on a total capital employed (TCE) of $55.8 million. This indicates a fairly good business with the ability to compound returns internally with growth. Because the business grew at 11 percent per annum in the last 10 years, we can be fairly certain that it is a good compounder.
Finally, coming to the valuation of the business, a sensible assessment would be to look at the conventional valuation metrics of EV/EBIT and PER. On page three of Buffett’s 1975 year-end annual letter to Berkshire Hathaway shareholders, Buffett describes Berkshire’s “largest equity investment [at the time] is 467,150 shares of Washington Post ‘B’ stock with a cost of $10.6 million.” Calculating the per share price basis, Buffett paid roughly $22.69 per share for the stock. Note that the 467,150 shares would have represented approximately 10 percent of the total common shares outstanding of The Washington Post at the time. From other accounts of the stock price movement, one knows that while the Washington Post stock was first quoted on the stock exchange after going public at about $26 per share, it had fallen to as low as $16 per share during 1973.5 Although this is based on third-party accounts of what was happening in the stock market at the time, I believe that the share price drop was a result of both macroeconomic concerns and bad press for the Washington Post. The bad press related to the Post’s investigation into the Watergate scandal that led to President Richard Nixon’s resignation and the presumably Nixon-led challenges to the Post’s television licenses mentioned earlier. In fact, alluding to this situation is the following quote from Beebe and Graham on page five of the 1972 Washington Post annual report: “There is nothing surprising in the adversary relationship between government and the press. It is as old as the Republic. Yet for those of us in the news business, the intense hostility evident during the recent past has been disquieting.”
Looking at the stock price at $22.69 per share, the average price at which Buffett accumulated his stake, the backward looking EV/EBIT and PER multiples would have appeared as listed in tables 8.4. and 8.5.6 Note that this price would have been approximately 40 percent above the 52-week low price of the share in 1973.
Table 8.4.
| EV/EBIT |
1971 actual |
1972 actual |
| Revenues |
$192.7m |
$217.8m |
| EBIT |
$15.2m |
$21.8m |
| EBIT Margin |
7.9% |
10.0% |
| EV/EBIT |
7.7× |
5.3× |
At a last full year EV/EBIT of 5.3×, a good quality business like the Washington Post would seem very cheap by today’s standards. However, this is a bit misleading. Looking at the income statement of the Washington Post, we note that the total tax rate paid in 1971 and 1972 by the company amounted to 50.2 percent and 49.5 percent respectively. This compares quite unfavorably with the approximate 30 percent total tax rate that most companies pay today. In fact, as we see in table 8.5, this was a result of the prevailing federal corporate tax policy at the time.
In effect, the higher tax rate in 1973 means that for the same amount of EBIT, a shareholder would get less in cash earnings compared to someone who owned the business today. If we were to adjust for this factor, a comparable EV/EBIT at today’s 30 percent tax rate would be 7.5×. When I say comparable, I specifically mean an EV/EBIT figure based on an EBIT that would have resulted in a comparable NOPAT to what the Washington Post made in 1973. The EV/EBIT of 7.5× is still cheap for such a good business that is growing, but not extraordinarily so.
Looking at the PER, we also see a moderate valuation.
Table 8.5.
| PER |
1971 actual |
1972 actual |
| EPS (adjusted) |
$1.52 |
$2.08 |
| PER |
14.9× |
10.9× |
Note that I have used adjusted EPS, which does not include extraordinary items or special credits, which is reported on page 17 of the annual report under the consolidated income statement. I have done so to make a valuation judgement based only on income generated from the business. At a 10.9× 1972 fiscal year PER, again, the Washington Post looks fairly cheap given its high-quality business and growth at a high ROTCE (compounding ability).
Table 8.6.
Historical corporate top tax rate and bracket (1909–2010)
Source: Office of Tax Policy Research, 1909–2001: World Tax Database. http://www.wtdb.org/index.html. Accessed October 17, 2002. 2002–2010: Internal Revenue Service, Instructions for Form 1120.
To summarize the valuation, the Washington Post seems to be a case where Buffett paid a decent multiple for a great growing business with an inherent ability to compound. By earnings power standards, based on the information an investor had at the beginning of 1973, it would not have looked like rock-bottom valuation. That said, it is certainly true that when Buffett began buying, it was at a lower price than his average, and he continued buying and was intent on accumulating a 10 percent stake of the company, which is fundamentally much more difficult than if a small investor acquires a small amount of stock.
There are two learning points I draw here. First, it seems that the Washington Post, like American Express, was a very high-quality business, with a historical 10-year track record of growth and ability to generate a superior ROTCE. In this case, the business was selling at a cheap price, but not ridiculously so. Second, given all the detailed circulation and competitive data given in the annual reports, there was clearly a very high quality of information available to the investor who looked carefully.
The lifelong friendship between Buffett and Graham developed from this point on. Buffett became a director of the Washington Post in the fall of 1974 and over time became a trusted advisor to Graham, instilling in her the conservatism in capital allocation that would make her a CEO who rarely overspent on acquisitions, but who also at times missed development opportunities that might have been stellar. The largest impact Buffett had on Graham, however, was instilling a shareholder-friendly mindset and a focus on operating the business efficiently. By 1985 operating profit margin had increased to 19 percent from 10 percent in 1974. The Washington Post also had used its excess cash to retire almost 40 percent of total shares in the interim. Net earnings grew seven times over, but earnings per share grew by a factor of ten.7
Table 8.7.
Income statement (1971–1972)
| |
1972 |
1971 |
| Operating revenues |
| Advertising |
$166,100,000 |
$147,633,000 |
| Circulation |
$47,421,000 |
$42,397,000 |
| Other |
$4,323,000 |
$2,719,000 |
| Total operating revenues |
$217,844,000 |
$192,749,000 |
| Costs and expenses |
|
|
| Operating |
$146,644,000 |
$133,869,000 |
| Selling, general, and administrative |
$46,254,000 |
$41,250,000 |
| Depreciation and amortization |
$3,140,000 |
$2,436,000 |
| Total costs and expenses |
$196,038,000 |
$177,555,000 |
| Income from operations |
$21,806,000 |
$15,194,000 |
| Other income/deductions |
|
|
| Other income (incl. interest of $804,000 and $845,000) |
$1,143,000 |
$1,091,000 |
| Other deductions (incl. interest of $2,484,000 and $2,774,000) |
−$3,240,000 |
−$3,275,000 |
| Equity in earnings of affiliates |
$512,000 |
$509,000 |
| Total other income/deductions |
−$1,585,000 |
−$1,675,000 |
| Income before income taxes, extraordinary items, and special credits |
$20,221,000 |
$13,519,000 |
| Income taxes |
|
|
| Currently payable |
$7,485,000 |
$5,698,000 |
| Deferred |
$2,721,000 |
$1,037,000 |
| Total income taxes |
$10,206,000 |
$6,735,000 |
| Income before extraordinary items and special credit |
$10,015,000 |
$6,784,000 |
| Extraordinary items |
−$283,000 |
$387,000 |
| Special credit* |
— |
$4,586,000 |
| Net income |
$9,732,000 |
$11,757,000 |
| Earnings per common share and common equivalent share |
|
|
| Income before extraordinary items and special credit |
$2.08 |
$1.52 |
| Extraordinary items |
$0.06 |
$0.09 |
| Special credit |
— |
$1.04 |
| Net income |
$2.02 |
$2.65 |
Source: Washington Post, 1972 Annual Report, 17.
*Cumulative effect on years prior to 1971 of changes in accounting methods for magazine subscription procurement and book promotion costs.
Table 8.8.
Balance sheet (January and December 1972)
| Assets |
Dec. 1972 |
Jan. 1972 |
| Current assets |
|
|
| Cash and time deposits |
$10,215,000 |
$10,268,000 |
| Commercial promissory notes at cost which approximates market value |
$19,635,000 |
$15,224,000 |
| Accounts receivable, less estimated returns, doubtful accounts and allowances of $2,663,000 and $2,342,000 |
$25,195,000 |
$19,992,000 |
| Inventories at lower of average cost or market |
$3,801,000 |
$4,641,000 |
| Prepaid expenses and other |
$2,908,000 |
$2,012,000 |
| |
$61,754,000 |
$52,137,000 |
| Investments in affiliates |
|
|
| Bowaters Mersey Paper Company Limited |
$8,649,000 |
$8,834,000 |
| Other |
$2,679,000 |
$1,736,000 |
| |
$11,328,000 |
$10,570,000 |
| Plant assets, at cost |
|
|
| Buildings |
$30,185,000 |
$16,258,000 |
| Machinery, equipment, and fixtures |
$34,412,000 |
$25,549,000 |
| Leasehold improvements |
$2,473,000 |
$2,378,000 |
| |
$67,070,000 |
$44,185,000 |
| Less accumulated depreciation and amortization |
($27,625,000) |
($25,796,000) |
| |
$39,445,000 |
$18,389,000 |
| Land |
$6,403,000 |
$6,403,000 |
| Construction in progress |
$323,000 |
$16,323,000 |
| |
$46,171,000 |
$41,115,000 |
| Goodwill and other intangibles |
$36,860,000 |
$37,517,000 |
| Deferred charges and other assets |
$4,918,000 |
$4,353,000 |
| |
$161,031,000 |
$145,692,000 |
| Liabilities and shareholders’ equity |
|
|
| Current liabilities |
|
|
| Accounts payable and accrued expenses |
$19,437,000 |
$17,368,000 |
| Federal and state income taxes |
$3,142,000 |
$735,000 |
| Contributions due to employee benefit trust funds |
$1,316,000 |
$837,000 |
| Current portion of long-term debt |
$1,734,000 |
$797,000 |
| |
$25,629,000 |
$19,737,000 |
| Other liabilities |
$5,529,000 |
$5,467,000 |
| Long-term debt |
$35,436,000 |
$38,033,000 |
| Deferred subscription income less related magazine subscription procurement costs of $11,998,000 and $10,496,000 |
$8,973,000 |
$7,900,000 |
| Deferred income taxes |
$6,077,000 |
$3,891,000 |
| Minority interest in subsidiary company |
$356,000 |
$313,000 |
| Shareholders’ equity |
|
|
| Preferred stock, $1 par value, authorized 1,000,000 shares |
|
|
| Common stock: |
|
|
| Class A common stock, $1 par value, authorized 1,000,000 shares; 763,440 shares issued and outstanding |
763,000 |
763,000 |
| Class B common stock, $1 par value, authorized 10,000,000 shares; 4,304,040 shares issued; 3,982,888 and 3,993,257 shares outstanding |
4,304,000 |
4,304,000 |
| Capital in excess of pay value |
$10,149,000 |
$10,079,000 |
| Retained earnings |
$68,835,000 |
$60,052,000 |
| Less: cost of 321,152 and 310,783 shares of Class B common stock held in treasury |
($5,020,000) |
($4,847,000) |
| Total shareholders’ equity |
$79,031,000 |
$70,351,000 |
| |
$161,031,000 |
$145,692,000 |
Source: Washington Post, 1972 Annual Report, 18–19.
Table 8.9.
Cash flow (1971–1972)
| Fiscal year |
1972 |
1971 |
| Financial resources were provided by |
|
|
| Operations |
|
|
| Net income |
$9,732,000 |
$11,757,000 |
| Less: portion of 1971 special credit not affecting working capital |
— |
($4,300,000) |
| |
$9,732,000 |
$7,457,000 |
| Depreciation and amortization of plant assets |
$3,140,000 |
$2,436,000 |
| Amortization of deferred film costs |
$1,661,000 |
$1,306,000 |
| Income tax timing differences |
$2,186,000 |
$808,000 |
| Sale of Art News goodwill |
$650,000 |
— |
| Other |
$386,000 |
$296,000 |
| |
$17,755,000 |
$12,303,000 |
| Increase in long-term debt |
— |
$8,222,000 |
| Increase in deferred subscription income |
$2,575,000 |
$875,000 |
| Proceeds from issuance of Class B common stock |
|
|
| Stock options |
$161,000 |
$929,000 |
| Public offering and sales to employees |
— |
$15,025,000 |
| Newsweek employees saving plan trust |
— |
$58,000 |
| Other |
$375,000 |
$118,000 |
| |
$20,866,000 |
$37,530,000 |
| Financial resources were used for purchases of |
|
|
| Plant assets |
$8,820,000 |
$13,748,000 |
| Television film rights |
$2,232,000 |
$1,449,000 |
| Treasury stock |
$307,000 |
$530,000 |
| Reduction of long-term debt |
$2,597,000 |
$10,061,000 |
| Increase in deferred magazine subscription procurement costs |
$1,502,000 |
$1,128,000 |
| Dividends on common stock |
$949,000 |
$871,000 |
| Increase in other investments |
$700,000 |
— |
| Other |
$34,000 |
$319,000 |
| |
$17,141,000 |
$28,106,000 |
| Net increase in working capital |
$3,725,000 |
$9,424,000 |
| |
|
|
| Changes in composition of working capital |
|
|
| Increase (decrease) in current assets |
|
|
| Cash and time deposits |
($53,000) |
$1,231,000 |
| Commercial and promissory notes |
$4,411,000 |
$3,815,000 |
| Accounts receivable |
$5,203,000 |
$99,000 |
| Inventories |
($840,000) |
$922,000 |
| Prepaid expenses and other |
$896,000 |
$348,000 |
| |
$9,617,000 |
$6,415,000 |
| (Increase) decrease in current liabilities |
|
|
| Accounts payable and accrued expenses |
($2,069,000) |
($1,980,000) |
| Dividends payable |
— |
$200,000 |
| Federal and state income taxes |
($2,407,000) |
$1,116,000 |
| Contributions due to employee benefit trust funds |
($479,000) |
1,157,000 |
| Current portion of long-term debt |
($937,000) |
$2,516,000 |
| |
($5,592,000) |
$3,009,000 |
| Net increase in working capital |
$3,725,000 |
$9,424,000 |
Source: Washington Post, 1972 Annual Report, 20.
