Is it normal for (individual stock) betas to fluctuate vastly from month to month?

Question

using the Covariance (stock, market)/ Varp (market) to compute beta in excel, whereby stock and market returns were first subtracted with the risk free rate, then ln-ed (natural log) to get the ln real return of the stock and market returns.

we found that the beta values differ vastly from month to month. eg. January's beta for XYY stock is 2.53 but February's is -1.56.

Is this normal?

If it is, why does the beta fluctuate so much?

Does this mean that if investors were to use the CAPM to calculate the required return, they would have to compute the stock's beta monthly?

Answer 1

probably a daily basis would lead to greater variane the asset specific risk, beta is calculated on a three monthly or greater basis, if the returns u foind were that adverse it is due to forces in the market but when taken on an avg basis last quarter last 2 quarters, etc. they dont fluctuate thatg much...

the socks r risky if they do so.....

Answer 2

The CAPM is useful for writing papers that win Nobel prizes, not for explaining real world activity.  This is why, when you take two Nobel prize winners and give them real money to invest, you get the Long Term Capital Management disaster.

CAPM formula has a problem with negative returns?  In the real world, stock prices do go down.

Beta is not a constant.  Look at the real world. If stock ZZZ had a constant beta of 2 then if the market went up 3%, ZZZ would go up 6%.  If the market went down 2% ZZZ would go down 4%.  The movements of the stock would go in lock-step with the movements of the market.  Simple obseration shows that stocks don't act this way.

To more directly address your question: Studies show that price changes are explained about 50% by market factors and about 50% by company specific factors.  Market factors are the CAPM's so-called "systematic risk".  This is over a long time period.  In short periods, there will be an absence of company specific news and price movements will more closely parallel the market.  In other short periods, there will be important company specific news and price movements will happen more independently.  Thus beta will change based on the appearance of company or industry specific news (or the lack thereof).

Many attempts to mathematically model the stock market use the normal distribution because it's convenient.  One of the biggest flaws is that markets don't follow a normal distribution; dats is frequently skewed and outliers occur more frequently than predicted by a normal distribution.  (This last bit is what brought down LTCM and regularly kills of traders who follow regression strategies.)

Feel free to email me if I have not completely answered your question.

Answer 3

yes very normal.  Beta is the risk vs the S&P 500 which changes daily.  Beta is also a measurement of the past, so it could be calculate on a daily basis if yo really wanted it to.