| |
 May 2005 Beta Isn't Alpha
It's not that they're perplexed by Greek scholarship or a foreign alphabet. Truth be told, many investors aren't even aware of how alpha and beta affect their investments. Actually, that's the problem. Before buying a stock or bond, you probably have some idea of what sort of return you can expect. That's why you buy that particular security instead of others: You think its potential is better than that of the alternatives. Once you own it, you probably check periodically to see if it's behaving the way you anticipated. If it is (or if it's doing better) you probably pat yourself on the back for your astute judgment. If it's not doing as well as you'd hoped, this might be a reason to sell or at least to again consider the alternatives.
How do you make those decisions? Do you have an objective test or do you simply look at past returns and relative market performance? This is where alpha and beta can come in, and it's also where investors confuse them. In financial parlance, the Greek letters alpha and beta are used to represent statistical concepts denoting two distinct aspects of a security's return. We've discussed this before in different context, but it's worth another look.
But think back to the late 1990s: Many investors were beating the market -- especially if "the market" was defined as the Dow Industrials. It wasn't hard to do, they only had to load up on Tech stocks. Investors who did, were generally pleased with their returns -- at least until the bubble burst.
The point here is that simply looking at returns -- no matter how good they may appear on an absolute or relative basis -- doesn't provide an adequate evaluation of a security. Market-beating returns are great, but it's also important to know how they were achieved.
That's why we have academics. They make a living investigating, among other things, the workings of the financial markets. By applying descriptive statistical techniques, they are able to explain at least some returns. In addition, the same processes can be adapted to help project future returns as well.
All that sounds extremely complicated, but the end results really aren't. That's where the concepts of alpha and beta come in.
According to the widely accepted Capital Market Pricing Theory, a great deal of any security's return is determined by the market in which it trades. The rest is a function of the specific security itself. This can stated mathematically as:
Here, y represents the security return, x is market return, and α, β, and є are constants. Since market return is the same for each security, it's these constants that account for the differences in their returns.
The least important is є. It's simply an error constant that makes the regression equation complete. It's unpredictable and usually relatively small, having little impact on the equation.
The beta constant (β) measures the security's sensitivity to market movements. By definition, the market itself has a beta of 1. Betas greater than 1 or less than -1 indicate the security is more volatile than the market. Those between 1 or -1 mean it's less volatile. A positive beta means the security moves in the same direction as the market while a negative one indicates the two move in opposite directions.
For example, if the security has a beta of 2 and the market (x) goes up 5%, the second term in the regression equation (β * x) yields a 10% increase (2 * 5%). On the other hand, with a beta of .5, the result would be 2.5% (.5 * 5%), less than the market increase.
The greater the beta, the more volatile the stock. This is often used as a measure of risk -- the greater the volatility, the more the risk.
Again consider the security with a beta of 2: It will rise 10% if the market goes up 5%, but will also fall 10% if the market drops 5%. On the other hand, the security with the beta of .5 will move less than the market both on the upside and downside. That's why it's also considered to be less "risky" than the one with the beta of 2.
But the beta term isn't the only determinant of return, alpha plays a role, too. Think about what happens to the return equation if the market is flat, in other words, when x=0. In that case, the beta value really doesn't matter, the second term ((β * 0) will always be 0. When that happens, return (y) will be equal to α + є. Since the error term is usually negligible, return will essentially be equivalent to the alpha term.
Alphas aren't always positive. In fact, many (particularly those of equity mutual funds) are negative. When this is the case and the market's flat, the security will actually have a negative return regardless of its beta.
Obviously, the greater the alpha, the better the return in any given year. Unlike the beta term (β * x), alpha doesn't change with market return. (This isn't to say that alphas don't change over time, they do, but their annual contribution to return isn't so directly tied to the market return as beta's.)
In most instances, the alpha and beta terms will both contribute to return. Whereas greater alpha is always helpful, greater beta only enhances return when the market goes up. This is a critical difference.
The answer depends on the stock's alpha and beta. To see why, consider Chart 1 which shows four different stocks each with a different values for alpha and beta. If stock 1 or 2 was up 4% when the market gained 3% you'd be happy: Actual return exceeded projected return. On the other hand, stocks 3 and 4 are disappointments in that they were expected to climb 5.0% and 4.5%, respectively.
In this case, a 4% gain isn't enough for stocks 3 and 4 while it exceeds expectations for stocks 1 and 2. A simple comparison of security and market returns isn't enough to determine if you've been adequately compensated for the risk you've taken. To do that, you need to know alpha and beta to come up with expected returns.
And no, just knowing expected returns isn't enough, either. It works for single time-periods, but isn't sufficient when making investment decisions for the future.
Charts 2 and 3 illustrate why. The first columns of Chart 2 show six different combinations of alpha and beta. They all have expected returns of 4% when the market is up 3% (column 8). But the other columns show how dramatically expected returns differ at other levels of market performance. This is most obvious at the extremes, but is still noticeable at all other levels or market returns.
Chart 3 is a graphical representation of this data. As you'll notice, all six combinations converge when the market is up 3%, but diverge at all other levels of market performance.
Looking closer, you'll notice the combinations with the biggest fluctuations are those with the highest betas (combinations 1-3). Although they have the highest returns when the market is up, they also fall much harder when it's down.
On the other hand, those with the highest alphas (combinations 4-6), are much more stable. With lower betas, they aren't nearly as susceptible to market swings.
It's here that the confusion arises. The best time to buy stocks is when nobody wants to, when they've been beaten down and are about to turn back up. At that point, stocks with high alphas have held up the best while those with high betas have fallen harder than the market. Think of this as the left extremes on Chart 3.
When the market does head back up, high beta stocks lead the pack. It's then that investors finally feel comfortable enough to return to equities, but it's also when high alpha stocks begin to pale. Think of this as the right side of Chart 3.
When the market is rising, investors gravitate toward high beta issues. Since they're the ones with the highest returns at the point, investors think they're "beating the market".
That's one of the primary reasons why Tech shares were in such demand in the late 1990s. As you can see from Chart 4, Tech and Telecom shares have the highest betas as well as the lowest -- actually negative -- alphas. That's why they ran so far ahead of the market during the run-up of the 1990s and also why they fell so hard when the bubble burst.
When investors loaded up on these high beta shares, they thought they were beating the market but instead their portfolios were simply more market sensitive. The fact that their returns surpassed those of the benchmarks was the result of their above-average betas, not positive alphas. Many investors were confusing beta with alpha.
But there's two important points here. First, it's critical to know what's generating your portfolio's return. Is it alpha or beta? If it's alpha, returns will be more stable although not as great as a high beta alternative when the market's hot. On the other hand, it won't have as much downside risk when the overall market is falling.
Which brings up the second point: High alpha portfolios are much better suited to long-term investors while high beta works best for traders. Given the latter's high volatility, it's best to maximize portfolio beta in up markets while minimizing it when stocks are falling. This is a short-term trading approach, not long-term investing.
Most taxable investors try to build portfolios that minimize transaction and tax costs. Many take this to the extreme, utilizing a "buy-and-hold" approach. Those who did this with high beta stocks in the late-1990s were in for a rough ride. To the extent that the equity market is cyclical, it's hard to be a long-term investor with high beta stocks.
Ideally, the most appropriate portfolio for a long-term investor would have near market beta and positive alpha. The former would allow the portfolio to keep up with the benchmark when stocks were rising while the latter would provide additional return even when the overall market was in decline.
Although Tech and Telecom, the two highest beta sectors had the lowest alphas, this inverse relationship doesn't hold for the other eight S&P sectors. As you'll notice from Chart 4 which lists the sectors in ascending beta order, Industrials and Utilities have considerably lower alphas than sectors with lower betas.
Nevertheless, we would suggest that long-term investors -- especially buy-and-hold investors -- focus on Financials, Healthcare, and Consumer Staples. The first two have near-market betas and positive alphas. They'll do well when stocks are climbing.
The Consumer Staples sector has the highest alpha while its beta is only about half that of the market. This allows it to provide above-average return when the market is down.
You've probably heard this before, just in a different context. Financial advisors are always encouraging their clients to diversify their portfolios to include stocks that perform well in any given part of the market cycle. Alpha and beta simply provide a quantitative explanation of why this works.
So the next time you're patting yourself on the back for "beating the market", take a closer look at why that's the case. You don't want to confuse alpha with beta.
Search this site! Just enter you key word or words:
Get current quotes or follow your own custom portfolio,
courtesy of E-Line Financials:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
HE FIRST TWO LETTERS OF Greek the alphabet are alpha and beta. Their names don't sound alike and their symbols are distinctly different. So why do so many investors confuse the two?
