I understand the beta of a stock to be the correlation between the stock's closing price and the closing value of a market index ETF (SPX), calculated over a period of time, (let us say 3 months).
So if a stock X has a beta = 0.8, as calculated by the closing prices over the past 3 months, it means for every point move in SPX, stock X moves 0.8$.

Is this correct? If yes, then does this relation hold on a day to day basis, or over a period of time? i.e if SPX advances 1 point in a day, will X advance 0.8 the same day, or is it: if SPX advances 1 point over a week, then X will advance 0.8 over the week?

What I am ultimately hoping is to find a stock that lags the market in a predictable way and hopefully the lag is more than a microsecond

  • R-squared may be more useful first before looking at beta, I'd suspect. – JB King Feb 10 '15 at 2:07

Just to be clear to start, beta is a statistical property. So if your beta is 0.8 over a period of time. Stock X moved on average 0.8 for a point move in the index. We might hope this property is persistent and it seems to be fairly persistent (predictable) but it doesn't have to be.

Also it is important to note this is not a lag in time. Beta is a measure of the average size of a move in the stock at the same time as a move in the index. In your example both the stock and index are measured at end of day. You can say that the stock "lags" behind the index because it doesn't grow as quickly as the market when the market is growing, but this is not a lag in time just a lag in magnitude.

People do occasionally calculate betas between a stock and lagged in time market prices, but this is not the commonly used meaning of beta. This might actually be a more useful measure as then you could bet on the future of the stock given what happened today in the market, but these "betas" tend to be much more unstable than the synchronized version and hard to trade on.

When you calculated beta you choose a time scale, in this case daily. So if your calculation is on a day-to-day basis then you have only tested the relationship on a day-to-day basis not, for instance, on a week-to-week basis. Now day-to-day and week-to-week betas are often related and are generally reasonably close but they do not have to be. There can be longer term effects only picked up on the longer scale. Stock X could day-to-day with a (average) beta of 1 to the stock market, but could have even a negative beta year-to-year with the market if the stock is counter-cyclical to longer scale trends on the market. So beta can change with the time scale used in the calculation.

| improve this answer | |

The beta of a stock can be interpreted as the average relative movement of a stock with respect to the movement of a market index. In your case, the stock will move on average by 0.8. Thus over a longer time horizon, not on a daily, weekly basis.

| improve this answer | |
  • Thanks. Can you elaborate how long is long term in this context? – Victor123 Feb 9 '15 at 18:25
  • I'd say at least five years. It also depends on the stock, most stocks have long-term stable betas. However, others can have a varying one due to strategy or business model changes. – QuantK Feb 10 '15 at 7:23

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.