We make it easy and fast to calculate many betas and many beta cases. But why should you?
There are many reasons to calculate your own betas. For example ...
- You want betas as of a date in the past, and only current or recent betas are available from public sources.
Analysts, corporate finance professionals, valuation practitioners, investors, and researchers sometimes need betas as of historical dates. They may be hard to find, and it takes time to gather the requisite information and correctly calculate and evaluate historical betas "by hand".
- You want betas which are calculated using total returns for both individual stocks and indexes.
Public beta sources, and much public stock and index data, often omit dividend reinvestment and therefore do not reflect total returns. Sometimes, some data is total return and other data is not. Sometimes, price data is not adjusted even for splits. Beta calculations are best performed using total returns for both the stock and the index.
- You want betas which reflect both a meaningful look-back period and enough observations for statistical relevance.
Say a company you're analyzing made a major acquisition in a new line of business last year, doubling its revenue. A beta calculated over a 36-month or 60-month period (both used by both free and paid sources of pre-calculated betas) therefore reflects, to some degree, a different company than the one you're looking at today. You may want to consider, instead, calculating a beta over the past year, using weekly data. That way you will have both a more relevant look-back period, and 52 observations, instead of just 12.
Many factors may affect your view of the relevant look-back period. These include not just company-specific changes in leverage or operations, but also events with an industry-wide impact and your view of the future. For example, as of June 30, 2009, should an analysis of betas and the cost of equity in the home-building industry be dominated by, or largely exclude, the impact of the sub-prime mortgage crisis? The answer is likely "it depends". With CalculateBeta™, you can readily construct your analysis either way, depending on how you calculate your betas, whether and how you consider prior period rolling betas, etc.
You don't avoid such choices by using pre-calculated betas. You just make them implicitly, and perhaps unknowingly.
The sub-prime crisis; 9/11; wars abroad; the release of the iPhone; a recent election; the decline of print media; an individual company's decision to buy or sell a major line of business, or to borrow at a whole new level: For your purpose, are such events relevant? Do you think they're non-recurring, rare, the new normal, or something else? We don't know, and that's the point. No provider of pre-calculated betas knows either. Only you can, or should, decide what is meaningful for purposes of the analyses you perform. With CalculateBeta you can easily explore alternatives and evaluate results.
- You want to see if beta is stable, or if it has been changing over time.
Beta is supposed to be forward-looking, and we usually calculate historical betas as proxies for expected future betas. But beta is not always stable over time. Is it in the case and time frame that matters to you? If it's moving, why, and is there a trend? Answering these questions requires considering rolling betas over a relevant time frame. With CalculateBeta, that's easy to do.
- You want to evaluate the quality of the betas you've calculated.
Betas' statistical quality varies - do you know the quality of the betas you use? Free and paid sources of pre-calculated betas frequently don't provide information such as standard errors and t-statistics. These statistics can be important in judging how much to rely on individual betas, or on a peer group beta mean or median.
These are among the reasons for calculating your own betas. Whether current or historical, beta is often treated as a black box. Not knowing what's inside can lead to poor estimates of risk and the cost of equity. Instead, you can use CalculateBeta to consciously choose what's in the box, without spending the substantial time and effort required to calculate and evaluate different betas and beta cases "by hand".