[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [amibroker] r-squared and regression analysis



PureBytes Links

Trading Reference Links




Dominick:
 
R2 (r-squared) is a statistical term that defines the degree of correlation 
between 2 variables over a given number of data points. As Anthony correctly 
stated, the closer the R2 is to 1, the better the correlation is. You stated 
that if a fund had an R2 of 80 (actually, 0.80), that meant 80% of the fund's 
movement is due to the index. This is technically not correct. It really means 
that 80% of the variability around the data can be explained or accounted for by 
the index. The other 20% of the variability is explained either by random noise 
or other factors that you have not evaluated. The less the variability is, i.e., 
the closer the data are to the regression line, the higher the r-squared term 
is, and therefore the greater the degree of variability that can 
be explained by the independent variable, in this case, the index. It does 
not necessarily establish cause and effect, although you may infer such 
causality if you want to. You just have to be careful when you do so. For 
example, if you conduct a regression of the closing price of INTC with the price 
of Japanese silk cocoons and find an R2 of 0.95, would you conclude that Intel's 
stock price is determined by the price of Japanese silk cocoons? However, if you 
compare the NDX with the QQQ and find an R2 of 0.95, it is likely safe to say 
that such a correlation has causality because they both derive their variation 
from the same population of data. I cannot comment on the beta that you are 
referring to because I'm not that familiar with it.
 
Al V.
<BLOCKQUOTE 
>
  ----- Original Message ----- 
  <DIV 
  >From: 
  dom1_1998 
  <dominick@xxxxxxx> 
  To: <A title=amibroker@xxxxxxxxxxxxxxx 
  href="">amibroker@xxxxxxxxxxxxxxx 
  Sent: Saturday, February 15, 2003 10:00 
  AM
  Subject: Re: [amibroker] r-squared and 
  regression analysis
  Hi Anthony:When I used to subscribe to Morningstar 
  which rated mutual funds, themeaing of R2 referred to how much of the 
  mutual fund point movementwas due to its comparative index, usually the 
  S&P 500.  If the fundhad an R2 of 80, that meant 80 percent of 
  the funds movement will bedue to the index.     The 
  other 20% was attributed to the individualcharacteristics of that 
  fund.You would think the above would have been beta but it's 
  not.Is your R2 and the one above the same?  If not could a code be 
  writtento determine the R2 for a set number of 
  periods?TIA,Dominick  Post 
  AmiQuote-related messages ONLY to: amiquote@xxxxxxxxxxxxxxx (Web page: <A 
  href="">http://groups.yahoo.com/group/amiquote/messages/)Check 
  group FAQ at: <A 
  href="">http://groups.yahoo.com/group/amibroker/files/groupfaq.html 
  Your use of Yahoo! Groups is subject to the <A 
  href="">Yahoo! Terms of Service. 







Yahoo! Groups Sponsor


  ADVERTISEMENT









Post AmiQuote-related messages ONLY to: amiquote@xxxxxxxxxxxxxxx 
(Web page: http://groups.yahoo.com/group/amiquote/messages/)

Check group FAQ at: http://groups.yahoo.com/group/amibroker/files/groupfaq.html



Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.