/* This program tries to replicate the regression equation discussed in a Business Week article entitled "Reading the Future? An equation may reveal your stocks' five-year return" by Carol Marie Cropper, April 15, 2002, page 116. The data from this regression was obtained from the 2006 Economic Report of the President, Table B-95 "Historical stock prices and yields, 1949 - 2003," page 392. In that table we have sp = "Standard & Poors composite index (1941 - 1943 = 100)" from the 9th column of table and ep = Earnings-price ratio (Common stock yields (S&P) (percent)) from the 12th column of the table. */ data bw; input year sp ep; datalines; 1949 15.23 15.48 1950 18.4 13.99 1951 22.34 11.82 1952 24.5 9.47 1953 24.73 10.26 1954 29.69 8.57 1955 40.49 7.95 1956 46.62 7.55 1957 44.38 7.89 1958 46.24 6.23 1959 57.38 5.78 1960 55.85 5.9 1961 66.27 4.62 1962 62.38 5.82 1963 69.87 5.5 1964 81.37 5.32 1965 88.17 5.59 1966 85.26 6.63 1967 91.93 5.73 1968 98.7 5.67 1969 97.84 6.08 1970 83.22 6.45 1971 98.29 5.41 1972 109.2 5.5 1973 107.43 7.12 1974 82.85 11.59 1975 86.16 9.15 1976 102.01 8.9 1977 98.2 10.79 1978 96.02 12.03 1979 103.01 13.46 1980 118.78 12.66 1981 128.05 11.96 1982 119.71 11.6 1983 160.41 8.03 1984 160.46 10.02 1985 186.84 8.12 1986 236.34 6.09 1987 286.83 5.48 1988 265.79 8.01 1989 322.84 7.42 1990 334.59 6.47 1991 376.18 4.79 1992 415.74 4.22 1993 451.41 4.46 1994 460.42 5.83 1995 541.72 6.09 1996 670.5 5.24 1997 873.43 4.57 1998 1085.5 3.46 1999 1327.33 3.17 2000 1427.22 3.63 2001 1194.18 2.95 2002 993.94 2.92 2003 965.23 3.84 ; data bw; set bw; y = ((sp/lag5(sp))**0.2 - 1)*100; x = 100/ep; lx = lag5(x); proc print data=bw; run; proc reg data = bw; model y = lx/ dw dwprob; run; proc autoreg data = bw; model y = lx/ nlag = 4 method=ml dw=4 backstep slstay = 0.05; run;