SOURCE CAPITAL /DE/ — Interim / Quarterly Report 1995

Nov 16, 1995

1 [SOURCE CAPITAL, INC. LOGO] SOURCE CAPITAL, INC. THIRD QUARTER REPORT September 30, 1995 2 OFFICERS AND DIRECTORS [SOURCE CAPITAL, INC. LOGO] SOURCE CAPITAL, INC.

3 LETTER TO SHAREHOLDERS TO OUR SHAREHOLDERS: As we recall our high school physics, Newton's law relating to momentum went something like "Bodies in motion remain in motion; bodies at rest remain at rest." This principle can be paraphrased as "You can't turn a moving battleship on a dime." Supported by rising profits and falling interest rates, this year's bull market in equities offers an apt example of this "moving battleship" phenomenon as prices continued their advance in the most recent quarter producing the best nine months returns since 1987 and the best trailing twelve months results since calendar 1991. During the most recent quarter, Source Capital's total net assets increased to $359,678,348 from $350,032,835 at mid-year. Net asset value per Common Share increased to $42.25 at September 30, 1995 from $41.09 at June 30, 1995. This increase is after payment of a $0.90 cash distribution during the quarter. INVESTMENT RESULTS In the most recent quarter, Source Capital's net asset value per share of Common Stock increased 5.1% while total net assets gained 4.6% with both comparisons reflecting reinvestment of dividends and distributions paid during the period. In comparison, the Standard & Poor's 500 Stock Average increased 8.0% during the quarter, also on a reinvested basis. For the nine months ended September 30, 1995, the net asset value of Source Capital Common Stock increased by 17.2% including reinvestment of distributions paid during the period, while total net assets gained 15.4%. These increases compare with a total positive return of 29.8% on the Standard & Poor's 500 Stock Average. NET INVESTMENT INCOME Net investment income totalled $2,371,074 for the third quarter and $7,773,799 for the nine months, decreases of 16.4% and 7.9%, respectively, from the corresponding periods of 1994. After providing for Preferred dividends, net investment income per Common Share amounted to $0.17 and $0.59 for the quarter and nine months, respectively, compared with the $0.24 and $0.72 earned in the corresponding periods of 1994. DISTRIBUTIONS TO COMMON SHAREHOLDERS A regular quarterly distribution at the rate of $0.90 per share was paid on September 15, 1995 to shareholders of record on August 25, 1995. Source's 10% Distribution Policy, adopted in 1976, calls for payments to Common shareholders approximating 10% of the Common Stock's ongoing net asset value. Shareholders are reminded that these payments substantially exceed the Company's net investment income and thus represent a continuing payment of a portion of the Company's capital. As we repeatedly point out, maintenance of the current $3.60 Common distribution rate is dependent upon achieving investment results which will sustain a net asset value of approximately $36.00. PREFERRED DIVIDENDS The regular Preferred dividend of $0.60 per share was paid on September 15, 1995 to shareholders of record on August 25, 1995. Asset coverage on the Preferred shares was 664% on September 30, 1995, compared with 646% at June 30, 1995 and 608% at year-end 1994. Net investment income provided Preferred dividend coverage of 201% in the third quarter and 219% for the nine months of the current year, compared with 240% and 238%, respectively, in the corresponding periods of 1994. 4 MARKET PRICE OF SOURCE CAPITAL SHARES After increasing from $37 to $40 1/2 in the first half of 1995, the market price of Source Capital Common Stock increased to $41 1/4 at September 30, 1995. This $0.75 increase in market price was less than the $1.16 gain in net asset value during the quarter. As a result, the market discount from net asset value increased to 2.4% at September 30, 1995 from 1.4% at mid-year. The market price of Source Capital Preferred Stock increased from $27 3/4 to $27 7/8 during the quarter. COMMENTARY Four and a half years ago in the spring of 1991, John C. Bogel, the CEO of The Vanguard Group, published an initial article in the Journal of Portfolio Management entitled "Investing in the 1990s: Remembrance of Things Past and Things Yet to Come." In that article, Bogel set out "a simple methodology for analyzing long-term total returns on financial market assets." For stocks, he concluded that "total returns could be explained largely by the interrelationship of three components: (a) the initial dividend yield, (b) the rate of earnings growth during the period, and (c) the annualized impact on total return that results from any change in the price/earnings multiple during the period." A second article followed in the fall of 1991 in which Bogel actually used this model to forecast future returns for the decade of the 1990s (1990-1999). His analysis only came to our attention recently by reading his follow-up article, "The 1990s at the Halfway Mark," published in the summer 1995 issue of the Journal of Portfolio Management. However, his approach of analyzing long-term future equity returns in terms of the contribution from each of the three factors -- yield, growth and change in valuation -- is exactly parallel to that initially explored in our 1994 Semi-Annual and Annual Report Commentaries and revisited in the 1995 Semi-Annual Report. In combining yield, growth and revaluation to make his forecasts in 1991, Bogel decided to eliminate all subjectivity from the process. Yield is of course of known quantity, and growth over long periods has not varied widely. The third factor, predicting valuation, is inherently speculative because the historical range of valuation has been so wide. Bogel's approach is ingenious, simple and logical. For growth, he utilizes the average of the prior thirty years, and deals with valuation by assuming that the price/earnings ratio at the end of his ten year forecast will regress to the average of the prior thirty years. The application of this methodology produced the following forecast for equity market returns for the decade of the 1990s.

While we will have to wait another four years to render a final judgment on the ultimate accuracy of this 1991 forecast, it is notable that returns for the first half of the period (1990-1994) were exactly 8.7%! Bogel acknowledges "that luck, pure and simple, played a role in the virtually exact correspondence of the forecast and actual returns." However, examining the results of all thirty-four moving ten-year periods from 1948 through 1990 produced a correlation of +0.78 between forecasted and actual returns (note: a correlation of +1.0 would indicate a perfect forecast record). Examination of the actual data presented in the current article reveals that the correlation of actual with forecast returns 5 seems notably higher in the most recent fifteen years than previously. One might expect that the thirty year averages of growth and valuation data would be more relevant and reliable once the data of the 1930s and 1940s is excluded. In any event, it is impossible to ignore the inherent logic and the outstanding forecasting record of this model. Bogel has updated his 1990-1999 forecast based on data as of December 31, 1994 which yields the following result:

This year's sharp market advance has driven the initial yield down to a current level of 2.5%. Since neither the trailing thirty year averages of earnings growth nor price earnings ratios are likely to have changed much with the passage of nine months, a current forecast based on Bogel's methodology would likely result in an estimated total return of about 8.0%. As we initially pointed out, this approach to analyzing long-term equity returns in terms of the three components of yield, growth and change in valuation is identical to that which we presented in last year's Annual Report. While Bogel uses long-term historical data to project future growth and future valuation levels, we attempted to use judgment in supplementing pure historical extrapolation of past growth rates. Further we ignored the tendency of valuation levels to regress to the mean and simply assumed current valuation levels would be sustained. Despite these differences, our conclusions are, not surprisingly quite similar to his. As we expressed in the recent Semi-Annual Report, "the high end of likely equity returns seems realistically limited to 8.5% to 9.0%." Bogel's original articles forecasting 8.7% stock market returns over the 1990s were published in 1991. In that year, equity returns exceeded 30%. Nevertheless, his 8.7% forecast turned out to be right over the first half of his forecast decade. Now he has updated his forecast to 8.4% over the decade 1995-2004, and the first year again is producing results well over three times his forecast. Just as the notably above-average returns in the year 1991 did not invalidate the ultimate accuracy of the long-term forecasts Bogel made in that year, we doubt that this year's sharp gains will prove sufficient to derail the ultimate accuracy of his updated forecast. The ultimate point of these articles is not to simply make predictions, but to point out that if reasonably reliable long-term forecasts are possible, they should help focus investors attention on appropriate long-term investment strategies. For us, that message translates into continued focus on businesses with enduring competitive advantages capable of earning above-average returns on capital over the long-term. Respectfully submitted, /s/ GEORGE H. MICHAELIS ----------------------------- George H. Michaelis President October 24, 1995 6 MAJOR PORTFOLIO CHANGES Quarter Ended September 30, 1995

7 COMPOSITION OF TOTAL NET ASSETS* September 30, 1995

SUMMARY FINANCIAL INFORMATION*

8 SOURCE CAPITAL, INC. 11400 West Olympic Boulevard, Suite 1200 Los Angeles, California 90064