Estimating Stock Beta: Market Proxy and Regression Window
Summary
The document discusses how Google Finance may estimate a stock’s beta, focusing on the market benchmark and historical observation window. One reported account says it regresses monthly closing prices against the S&P 500 using month-end data over five years. Another answer describes beta as a single-factor market regression, while a further response says Google’s exact method is not clearly documented.
The evidence is informal and partly speculative: commenters compare the suspected approach with Yahoo’s reported three years of monthly observations and Bloomberg’s reported five years. These figures should not be treated as a verified description of Google Finance’s calculation. The practical lesson is that beta depends on choices such as the benchmark and sample period. For a focused risk comparison, a researcher may prefer a benchmark that better matches the stock’s sector or exposure, then calculate beta using a clearly stated method.
Key ideas
- Beta is estimated by regressing a stock’s returns against a chosen market benchmark.
- The document reports a possible Google Finance method using month-end observations and the S&P 500.
- The calculation method attributed to Google is uncertain and is based on informal reports.
- Beta comparisons depend on the benchmark and observation window, so a custom estimate may better fit the research question.
Tags
Full text
# How google finance calculates beta of a stock # How google finance calculates beta of a stock How google finance calculates beta of a stock - What is the proxy for the market? - What is the time period it uses for regression? ## Answer by pyCthon (score 1) https://quant.stackexchange.com/a/18100 According to a comment on another post here: > ". It regresses against the SP500 using MONTH-END closing prices for the last five years." - @Dimitri ## Answer by Alex Bădoi (score 1) https://quant.stackexchange.com/a/22964 Google uses the 1 factor CAPM model developed by Fama French (1974). Its a simple linear regression with the stock as dependent variable and the market portfolio as independent ## Answer by consuli (score 0) https://quant.stackexchange.com/a/19544 There are some variants to calculate the beta of a stock. If not fully documented at Google, in doubt you have to validate yourself. You will find a help to do this in the linked website. However, the results of the different calculation variants are usually quite similar. ## Answer by Tim (score 0) https://quant.stackexchange.com/a/22981 Being a Google Finance user myself I was not able to figure out how it computes the beta. However, my best guess is that it's done in a way that is very similar to the methods used by Yahoo and Bloomberg. I.e. SP500 and 36 or 60 monthly observations. In general, I would say that it does not really matter how the beta is computed since the betas on Google and Yahoo are only used as a quick scan. For a closer examination, it always better to create your own definition of beta. For example, if you want to figure out how risky apple is compared to other tech companies it makes more sense to regress against the NASDAQ than the SP500. Yahoo: monthly 3 year observations (36 in total) against S&P 500 Bloomberg: monthly 5 year observations (60 in total) against S&P 500
Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.