ࡱ> 7967 bjbjUU .7|7|@]l~~~~~~~$P  ($ $ Lp p p y y y 6888888$ \~y W "y y y \ ~~p p q y ~p ~p 6 y 6  ~~p  ŏ 40  X  ~~~~What Has Happened to the Predictive Power of Dividend Yield? BA453: International Investments Professor Campbell Harvey March 1, 2000 Submitted By: Big Alpha Asset Management ("BAAM") Scott Chan Derrick Roth Ray Tong Goodloe White Frank Young  Introduction Historically, dividend yield has predicted future market returns; however, since 1980, this relationship has deteriorated. As the market capitalization of growth stocks have blossomed, stocks that pay no or little dividends have become a larger part of the market. Accordingly, the dividend yield of the market as a whole has declined dramatically, ultimately reducing the effectiveness of the dividend yield as a predictor of the overall market return. In fact, the correlation between dividend yield and market return has turned negative. Our hypothesis is that dividend yield and market returns still possess a strong positive correlation. In order to uncover this relationship, we obtained the largest and most consistent dividend paying stocks in the S&P 100 index via removing any stocks that do not pay dividends or distribute a very low dividend. We will test whether dividend yields of this portfolio can predict the portfolios returns. Method We first determined the constituents of the S&P 100 index as of December 31, 1999 as a fair proxy for the market return. We then collected data for the outstanding shares, year-end prices and cash dividends for each of these constituents from 1971 to 1999. Unfortunately, data in Compustat was only available post 1970 for a majority of the existing index. Therefore, only the firms with data since 1970 were included in the sample. In addition, we could not determine the changes in the index post 1970 and hence only the stocks as of the December 31, 1999 S&P 100 index were considered. We then filtered out the stocks that fell below our dividend yield requirement each year. We varied this dividend yield requirement from .5% to 5%. With this filtered list of stocks, we regressed the lagged dividend yield, dividend and 1/(Market Capitalization) against the five-year return of the market. We used 1/(Market Capitalization) versus the standard 1/Price due to the potential effect of unadjusted stock splits in the data. Results Our results again showed that dividend yield is not an accurate predictor of market returns, even when you exclude non-dividend paying stocks. Our regressions showed no correlation between dividend yield and market return and a negative correlation for dividends. In fact the data in the regression appeared to be scattered with no clear linear pattern compared to the negative linear relationship without the removal of low dividend yielding stocks. It is interesting to note that the dividend yield coefficient began to turn negative around 1980 and showed a consistent movement towards this sign since 1971. The strong correlation between the 1/(Market Capitalization) and market return (t-stat = 7.25 and r-square = 70%) seems to explain this phenomenon. In other words, the growth in the market capitalization of these stocks appears to explain the weakness in the dividend yield as a predictor. Next Steps It is possible that the simplifications we made in building a proxy for the market return biased our results. A more complete analysis would include the changes in the S&P 100 index for each year versus only gathering the current S&P list. In addition, we recommend gathering more data for the sample (i.e., on a quarterly basis) because our model includes only fifteen data points. Correcting these issues with the data may produce a more positive result. PAGE i  FILENAME ass1 |  SAVEDATE \* MERGEFORMAT 2/28/00 8:58 PM |  PAGE 2 >@@AGHIJKMNXY]^ab|}hmHnHuh jUh0J!mHnHu0J! j0J!Uj6UmHnHu6:5CJ 5:CJ$">?@a{|} -. it !$a$@t@KLM  !H$&`#$ !*00&PP/ =!"#$% &P/ =!"#$%nqGU rݮ{APNG  IHDRRRsPLTEƽ{{{ssscccZZZRRRBJJ֜ތBksޔcBck!)Bs9k{BZcBZRk1BBZJkcs{{kc!RkJcBc1Jƌ{kc9sBcZJs֌ΜJcsk{R{ZRBk1Zs9k)Rk1cZBk諒Υބބ֌cR1k)sR{JsZR1R!9)眽{Z{Z9Zs1Rk9cR!ZBcZ)BZJBs9cssJsRBkZJ{RBsB{)Rs1c)Z)c!R{!Z9ZZ1R9cJRB{9k)J֜RcskJc{祵k{ΔƄ֌ΥƜZsksZ{kcJk1JcZJsJ{Bs!Bc)Z!Js9Z!R1RR1Z9kB{JB1c)R!B1!ZB{BB9{֌sZ{kBk1R{!R9k)R1k)Z!Jc{k1R1{cs)k)c{)9Z)RBR{1R1J)JΌ1JZc罽ꁜItRNS@fbKGDH?IDATxՙ\g۪ho/=!+ԫ_{&kT=LC ZVyUJJ.-VfLs!v Zk<]Ͼ@[ Lfy>;<>C$x/aƀ-CQ/:LbI?* KH ]:kra~) ayp>T6|<. `}f8 +EVg"&_c}iR ~XF+&EL =& g|f<|>ntoNOah$xoeE?baϽ n_d4Q傐ݿ@Zw}8=B\o4_(qYV8Yb~9aPCl<&&1 hס?F![[Zl8k`oC ׫ b[23Qr+uQDϊ(vEwA+QuXHdqodf}qn^oL̯~@gѤ(1?:{⚰FG yNg9?Ak)ud8F#g~,?1@-jݍȽDc'_oOy bU \x#}Ƞ"y=n8ٱl?K$hDdvVK85 4׿afaM mŽ6ɌA$eIo?lʚhqrzdGB#O\"clLD/jsr4 5,53s<oEHܗ6$4bq5RQڮE B-L˟ YX+ݝP+o.Y_z8 :;v!$_VaDV1DllLUz, jIKȝC 'zB^_ZpԯaH&uA #i,;g^(?|C[WHCH]n"._!;k0ϒ "1]K 1d*tيz(򡘁FZ W߯)MWH75c~wu~, \.X~ >su 2|0^=VTY=}d4}raq8^(.TgVgVUUxY]~ZCR˻}̊s  N,(l83+1++ qf}}R6H)IHR*Zl B"2>(p]NV`H8?'_XAg0I7 y`yxwMj]xF1Qo8jn]mNt_ֈ8pCﱆ^((kR\kj*?|Yҳ? g˛SY[nVJz.U&V"/A$0f[|h#iӭMU1uoЯ/2;I hCDg u8$hfl-"k նSXTD.H %~{DAI3$gm6j6p4-.3漼_Y,cK[r莨.@v:>-kETMGa lo(0"?yi=:Db,Pfy_w'}[g?bc a$p @ֲfIMPjT`Yw0gPZ0107~T~1e|>,y}DL_Ţm#k#?`u\*3V}%疱t{ cjV?1^a$F{@͖^%#?sWǯml`R@01\ ˌ9͔|JV K!jEvւWiKkc!6CQzǞ,)ܬ'&ic3 Z -aI}qJT/ A`˜/)J%4|,OCVu\K8uIL[ǘ{ |pU^AWYte']Ѫ~򥅅6E)sV^ՔŌN'B_C#zwOHӨT+sOT5wm%cssڛbr{j-e^C܆d*?h˯ (KAGo*ysޕMNg{sۓouO衎Nx)ϪHq2CrUNѶ!SJ-JQtP@d9sIT:8jE!"(ߒbnSѸl鎒uaQy߼d]L9s]w>ӉI 'yENjqݪv*JsQڹR3JzܶbonˊKU{rW\dsmKP&lIo|#e)lN30:Xt%(4}a9HB5춉S>?&b~|--y7/ۙcs=Re\Czj{galX6[`[o`IW<>&5\]`x< y<уSĈS#1u\Sf 1#&ȿDN '=ƓԔ|<*3L~׏05aF1_>zRjg7ꡏNPIENDB`# i4`4 NormalCJ_HmH sH tH J`J Heading 1$ & F <@&5;CJKHTT Heading 2)$ & F <@&^5:RR Heading 3)$ & F 8<@&^6<A@< Default Paragraph Font,@, Header  !0 @0 Footer  !CJ."@. Caption xx5@"@ Formula Line <^626 Bullet & F h^01B0 Box & F h^8YR8 Document Map-D OJQJ00 TOC 1 xx 5;CJ.. TOC 2 ^:CJ.. TOC 3 ^6CJ** TOC 4 ^CJ** TOC 5 ^CJ** TOC 6 ^CJ** TOC 7 ^CJ** TOC 8 ^CJ** TOC 9 ^CJL#L Table of Figures ^` ;CJ TOC1 &)@& Page NumberDT"D Block Text":0]:^`0CJ  , >?@a{|} -.i t @KLM000000000000000000`0 0000 00 00 00i @0@0@0@0@0 \\\_ t   !<LPWY_!!l,b$f UmbUzN@j(    C AhC:\Files\Fuqua\Year1\t1y1\Ba395\WebPage\FSBLogo.gifC"B S  ?A$ @@?@JKKMGoodloe H. WhiteMC:\WINDOWS\Application Data\Microsoft\Word\AutoRecovery save of Document1.asdGoodloe H. White$C:\Files\Fuqua\T3\II\AS1\AS1BAAM.docGoodloe H. White$C:\Files\Fuqua\T3\II\AS1\AS1BAAM.doc8.F #24ȽE~- f24:Zd-8~ `CI9M9aY. hh^h`OJQJo(hh^h`.P^`P..^`...x^`x.... ^` .....  X ^ `X ......   ^ `....... 8^`8........ `^``......... hh^h`OJQJo( hh^h`OJQJo( hh^h`OJQJo(hh^h`.P^`P..^`...x^`x.... ^` .....  X ^ `X ......   ^ `....... 8^`8........ `^``.........hh^h`CJOJQJo( E~-d-88.F9aYI9M#d-8d-8d-8f24##?@M@@rL@UnknownG:Times New Roman5Symbol3& :Arial5& :Tahoma;Wingdings"hx Cf Cf2  $r0n@2?C:\WINDOWS\Application Data\Microsoft\Templates\Case-Report.dotGTAA - Assignment 1Goodloe H. WhiteGoodloe H. White Oh+'0 $ @ L X dpxGTAA - Assignment 1TAAGoodloe H. White 1ood Case-ReportGoodloe H. White 12odMicrosoft Word 9.0@e@.~/`@@A@K  ՜.+,D՜.+,@ hp|  ln GTAA - Assignment 1 Title 8@ _PID_HLINKSAnX4C:\Files\Fuqua\Year1\t1y1\Ba395\WebPage\FSBLogo.gif  !"#$%'()*+,-/0123458Root Entry F@9ŏ:1Table WordDocument.SummaryInformation(&DocumentSummaryInformation8.CompObjjObjectPool@9ŏ@9ŏy,@  FMicrosoft Word Document MSWordDocWord.Document.89q