Beta of a Stock is a Meaningless Measure

For most investors who have grown up on the diet of high stock beta = high risk, this statement will come as a surprise. Beta means nothing for a stock. And beta explains nothing about the investment merits of a stock.

Let me start with two thought experiments to drive this point home.

Scenario 1: Let’s say you are interested in a stock that has declined  significantly in the recent months. So much so that at this time, after you have done your research and due diligence, you come to a conclusion that the stock presents a compelling value. In fact, you conclude, that this investment is as sure a bet as any one is ever likely to get. The market has over done its selling, as it generally does when there is a investor panic.

Let’s also spice it up a little and inject some specifics. This company, even though its stock has taken a beating, continues to generate more in free cash flow per year, than its entire market value, even after servicing its debt. You look deeper and find that after you clear away all the accounting clutter and GAAP mandated non sense, the stock is priced at about 1 times net earnings per share.

In short, this is a bargain unlike any other.

Heck, if this level of undervaluation continues, the company can just decide to buy itself out and go private. That would be a very wise decision on their part. Alternatively, it is a very good acquisition for a smart corporate buyer. There is no way this kind of undervaluation will be allowed to continue. Sanity eventually returns.

Now the question is, if you decide to invest in this stock, are you taking enormous amounts of risk? The market thinks you are. The recent falling over the cliff of the stock price has pushed the stock’s beta to beyond the threshold a rational investor would contemplate investing at. When the price starts moving up, it is likely going to be very volatile as well.

But you know that the risk here is close to zero. Just like Buffett thought the risk of investing in Washington Post was close to zero.

What do you think?

Related reading: Risk is not the same as volatility

Scenario 2: Let’s say you found a stock that no one cares about. It is a small company in the middle of nowhere in a boring industry. There is no analyst coverage, they have never been part of the Inc 500 and CEO owns so much of the company stock that it is almost like his family business. Which it was at one time.

In a quaint departure from the norm, the company management makes decisions that are decidedly long term. Some of these decisions involve capital expenses that will kill their next quarter’s earnings, were it not for the fact that they do not worry about issuing guidance and there is no analyst breathing down their necks.

The company’s balance sheet is like Fort Knox. Their margins lead the industry. It is a real gem. And the dividend is to die for. The management takes care of its shareholders.

You realize that the stock price today is much below where it ought to be given the company fundamentals. It is a great buy. And when some of the past capital investments start paying off, the company and the stock will be noticed. For the past few years, though, the stock has stayed sleepy and its beta is less than half of the market.

So you buy the stock.

You know your risk is low, and the beta is low, so it all matches up. Right? But that is not why you bought the stock. You bought this stock because you expect the stock to provide a return that is significantly better than the market. And as the stock gets discovered, this return will come to you in fits and starts. The price may become volatile, but you know in the long run you will make handsome profits.

You bought the stock because you know you are buying it at a low and your downside is protected, but you want the beta to be high on the upside and you want to capture that upside.

Just to rile up the Efficient Market Hypothesis proponents a little more, I will lay out the 2 key rules of investing. Especially value investing.

  1. Price =/= Value, and
  2. Volatility (beta) =/= Risk

Besides, stock prices over the short term are essentially random and over long term are dictated by the fundamentals and the company performance in the future. The beta measures the past volatility of the stock and has no bearing on what the stock does in the future. A stock is not born with a beta assigned to it. Every stock moves through periods of high beta and low beta. The trick is to focus on the business fundamentals and find great opportunities to invest.

Where Does Beta Make Sense?

Beta makes sense in the context of a portfolio. You want to have enough diversity in your stock picks that your overall portfolio becomes less volatile by reducing correlation between your holdings. One way to do this is to get exposure to different industries that are not tied to each other. However, unless you are buying a mutual fund, you will still actively manage your investments based on their own individual investment merit. And if you are a value investor, you have already taken pains to not over pay and have therefore cut down a good chunk of your downside risk on each stock that you buy.

As Warren Buffett points out, cash and other “currency” investments are some of the lowest beta investments available, and they are the riskiest ones for preserving or enhancing your future purchasing power. What is more important is to judge the risk of capital loss before making any investment.

This article is quoted and referenced in the upcoming book Mezzanine Financing: Tools, Applications and Total Performance by Professor Luc Nijs and published by Wiley Finance. Look for it in Chapter 3 which deals with the Intrinsic Cost of the Mezzanine Finance Products. The book is set to be released in Oct 2013.

Comments

  1. DanielMurray says

    I have to respectfully disagree with some of what you wrote. 
     In Scenario 1, you write, ” The recent falling over the cliff of the stock price has pushed the stock’s beta to beyond the threshold a rational investor would contemplate http://valuestockguide.com/ at.”
    A stock declining over a given period of time does not mean that its beta will be high.  It’s beta will only be high if this decline is correlated with a decline in the stock market and/or risk assets in general.
    Also, volatility(beta) does not equal risk.  Beta only represents that part of a stock’s risk that cannot be “removed” through diversification.  A stock’s total risk equals its “idiosyncratic” and “systematic” (beta) risk.

    • Shailesh Kumar says

      Daniel,

      My whole contention is that beta =/= risk. We agree on that.

      A stock declining consistently over a period of time does not mean the beta is high. I agree with this too. Apple stock is a case in point and last I checked sports a beta of 0.66.

      Take away the consistency, if the volatility goes up (high beta), your run of the mill investment advice counsels that the stock has become too risky. I disagree that short term gyrations in the stock price reflects well on the risk of investing in the stock. It may, or it may not, and you will not know until you understand the investment. When you do this, the beta is irrelevant.

      There are countless investors looking for low-beta stocks to invest in. These may not be good investments.

      Further more, beta on a stock does not stay constant over time. It only measures the past volatility. The future performance and volatility of the stock depends on the current/future set of economics and the choices that management makes in the future. Boiling it down to a co-efficient (or even a set of) does not do justice to the business of value creation.

      Most of the MPT is an attempt to shoe-horn math to describe how the capital markets function. It comes close but not close enough.

      Regards,
      Shailesh

  2. DanielMurray says

    I guess my main point is that Beta does not equal (total) risk or volatility. Anyways…As a trader, I also don’t agree with “run of the mill investment advice,” so we are on the same page and in agreement in general.

  3. Peter says

    Shailesh,

    I directionally agree with you statements, but I disagree with extent of the claim. Beta has many problems, many of which you’ve outlined. However, to claim beta is “meaningless” implies that no useful information whatsoever can be extracted through the knowledge of this metric and that is an exaggeration.

    Of course beta is based on historical data and is inherently unstable, but then again, what indicator isn’t? Your free cash flow model was based on historical information, and so is a chartist’s trend following strategy. And there is no assurance that next quarter’s results will be inline with last quarter, nor that a pattern will complete and price will breakout to the upside.

    Beta can be adjusted to make it more forward looking (e.g. adjust for changes in financial leverage, influence of DB pension plan IPS on stock price, mean-reverting tendency, sector rotations and regime changes), although it then becomes more subjective, just like your next quarter FCF model.

    Beta is a useful measure regardless of your belief in the EMH – indeed CAPM predates the EMH by a decade. And the EMH does not assume that stock prices returns are random (which of course they are not), only that they can be reasonably modelled as if they were random.

    The truth is that CAPM has had some poor out of sample results. Fama & French discovered that constructing a model that includes and systematic risk premium similar to Beta with additional risk premiums that reflect P/BV and a Small/Mid Cap premium improved out of sample results. And this, I believe, summarizes the one universal principle of investing: If you combine an increasing number of uncorrelated, multi-disciplinary indicators (that have some, but not perfect, predictive value) you will have progressively better results on an ever decreasing set of investable opportunities.

    The big takeaway from your article I believe is your last paragraph, which I couldn’t agree more with. Clearly Beta becomes much more meaningful at the diversified portfolio level when the firm specific risks have canceled each other out. It is at this level that Beta needs to be monitored and controlled. However, Beta still needs to be understood at the security level so that an investor can make appropriate decisions regarding a marginal security: Will the additional security to the portfolio positively or negatively affect the Treynor ratio of the portfolio as a whole?

    • Shailesh Kumar says

      Peter,

      Thank you for your more balanced treatment of the subject than mine certainly was :-)

      I want investors focus on beta at the portfolio level and business fundamentals and valuations at the individual security level. I often encounter investors who want to invest in low beta stocks and to me a low beta by itself is not a good reason to base your investment decision on.

      Like you say, these indicators become more meaningful when the sample size is larger, both in terms of number of securities and the combination of indicators used. At the individual security and single indicator level the variance could be large enough to render it useless for meaningful decisions (I will class your example of marginal security in terms of the portfolio as a whole, the make or break decision to include a new security should be based on the investment merit of the security)

      I am also a EMH skeptic and believe that the markets are influenced by both rationality and emotional responses. EMH precludes emotional triggers while behavioral investors tend to pay less attention to value/price equation.

      Regards,
      Shailesh

  4. Rohit Kapoor says

    You do realize that the Beta of a stock is its covariance with the market, and not its volatility, right? That is the beta is the measure of its undiversifiable systematic risk, not its volatility which could be simply due to idiosyncratic risk – beta may be not reliable but your arguments are not based on valid definitions of the beta (volatility is not equal to beta!).

    • Shailesh Kumar says

      Rohit, beta is the volatility over and above (or under and below, as the case maybe) the market volatility. As such, it is a measure of volatility. Whether it is the excess of, or less than the market is represented by the value of the beta coefficient.

      Beta of 0.5 = half as volatile as the market
      Beta of 2 = twice as volatile as the market

      Remember, beta is a coefficient applied to the market volatility, it has no meaning in by itself and derives its meaning from the market volatility.

      I am much more interested in practical implications for a value investor. Definitions are nice and it might pay to be pedantic in some circles (academic?), but I would rather know the meaning and implication.

      I generally do not like to use wikipedia as a reference but if you wish, here it is http://en.wikipedia.org/wiki/Beta_(finance)

  5. TW Koh says

    We can’t wave away definitions and assign our own meanings to terms/words. Come up with new words if you wish, but if you misinterpret commonly used terms/words, investors could potentially get confused when they read about these measures in other contexts.

    Let’s look at the wikipedia link: It states beta_a = cov(r_a,r_b)/var(r_b) where b is the reference index. That’s the commonly understood definition. Your paragraph:

    “beta is the volatility over and above (or under and below, as the case maybe) the market volatility. As such, it is a measure of volatility. Whether it is the excess of, or less than the market is represented by the value of the beta coefficient.

    Beta of 0.5 = half as volatile as the market
    Beta of 2 = twice as volatile as the market”

    doesn’t follow from the definition. Let’s think about this in 2 ways:

    1. Plot a scattergraph of returns, r_a against r_b (r_b on the x-axis). Do a least-squares best fit using a straight line through the points. The slope of the line is Beta.
    2. I suppose we all (think) we understand the concept of correlation (assuming normality). By definition, corr_a,b = cov(r_a,r_b)/(vol(r_a) * vol(r_b)) (in terms of returns r_a and r_b as above). Now, you can think about beta, mathematically, as correlation * vol(r_a)/vol(r_b) .

    From (1), you can see that there are many reasons why a low beta stock (to the market index) could exhibit very high volatility. For example, consider a stock that has very low statistical relationship to the movements of the market but it has huge amount of idiosyncratic risks. Plotting returns of this stock against that of the returns of the market would show a big uncorrelated ball of points. Thus your statement that beta is a measurement of volatility is inaccurate and your example that 0.5 beta means half the volatility of the market is incorrect. For example, MCD (MacDonald’s) (daily) beta to the SPX is 0.512 according to Bloomberg, and the 30 day historical realized volatility of MCD is about 14.667% compared against SPX of 9.097% according to Bloomberg.

    Finally, from (2), and the example above, you can also think about beta as expressing two components: (a) correlation and (b) relative volatility against the market index. A stock with low beta could be due to a low correlation to the market index (ie. it has more of a idiosyncratic risk component), and/or it has lower volatility as a ratio against the market index volatility.

Leave a Reply

Your email address will not be published. Required fields are marked *