What Is Attribution Analysis?

Attribution analysis is a sophisticated method for evaluating the performance of a portfolio or fund manager. Also known as “return attribution” or “performance attribution,” it attempts to quantitatively analyze aspects of an active fund manager’s investment selections and decisions—and to identify sources of excess returns, especially as compared to an index or other benchmark.

For portfolio managers and investment firms, attribution analysis can be an effective tool to assess strategies. For investors, attribution analysis works as a way to assess the performance of fund or money managers.

How Attribution Analysis Works

Attribution analysis focuses on three factors: the manager’s investment picks and asset allocation, their investment style, and the market timing of their decisions and trades.

The method begins by identifying the asset class in which a fund manager chooses to invest. An asset class generally describes the type of investments that a manager chooses; within that, it can also get more specific, describing a geographical marketplace in which they originate and/or an industry sector. European fixed income debt or U.S. technology equities could both be examples.

Then, there is the allocation of the different assets—that is, what percentage of the portfolio is weighted to specific segments, sectors, or industries. 

Specifying the type of assets will help identify a general benchmark for the comparison of performance. Often, this benchmark will take the form of a market index, a basket of comparable assets.

Analyzing Investment Style

The next step in attribution analysis is to determine the manager’s investment style. Like the class identification discussed above, a style will provide a benchmark against which to gauge the manager’s performance.

The first method of style analysis concentrates on the nature of the manager’s holdings. If they are equities, for example, are they the stocks of large-cap or small-cap companies? Value- or growth-oriented?

American economist Bill Sharpe introduced the second type of style analysis in 1988. Returns-based style analysis (RBSA) charts a fund’s returns and seeks an index with comparable performance history. Sharpe refined this method with a technique that he called quadratic optimization, which allowed him to assign a blend of indices that correlated most closely to a manager’s returns.

Explaining Alpha

Once an attribution analyst identifies that blend, they can formulate a customized benchmark of returns against which they can evaluate the manager’s performance. Such an analysis should shine a light on the excess returns, or alpha, that the manager enjoys over those benchmarks.

The next step in attribution analysis attempts to explain that alpha. Is it due to the manager’s stock picks, selection of sectors, or market timing? To determine the alpha generated by their stock picks, an analyst must identify and subtract the portion of the alpha attributable to sector and timing. Again, this can be done by developing customize benchmarks based on the manager’s selected blend of sectors and the timing of their trades. If the alpha of the fund is 13%, it is possible to assign a certain slice of that 13% to sector selection and timing of entry and exit from those sectors. The remainder will be stock selection alpha.

Market Timing and Attribution Analysis

Though some managers employ a buy-and-hold strategy, most are constantly trading, making buy and sell decisions throughout a given period. Segmenting returns by activity can be useful, telling you if a manager’s decisions to add or subtract positions from the portfolio helped or hurt the final return—vis-à-vis a more passive buy-and-hold approach.

Enter market timing, the third big factor that goes into attribution analysis. A fair amount of debate exists on its importance, though.

Certainly, this is the most difficult part of attribute analysis to put into quantitative terms. To the extent that market timing can be measured, scholars point out the importance of gauging a manager’s returns against benchmarks reflective of upturns and downturns. Ideally, the fund will go up in bullish times and will decline less than the market in bearish periods.

Even so, some scholars note that a significant portion of a manager’s performance with respect to timing is random, or luck. As a result, in general, most analysts attribute less significance to market timing than asset selection and investment style.

What Is Alpha?

Alpha (α) is a term used in investing to describe an investment strategy’s ability to beat the market, or its “edge.” Alpha is thus also often referred to as “excess return” or “abnormal rate of return,” which refers to the idea that markets are efficient, and so there is no way to systematically earn returns that exceed the broad market as a whole. Alpha is often used in conjunction with beta (the Greek letter β), which measures the broad market’s overall volatility or risk, known as systematic market risk.

Alpha is used in finance as a measure of performance, indicating when a strategy, trader, or portfolio manager has managed to beat the market return over some period. Alpha, often considered the active return on an investment, gauges the performance of an investment against a market index or benchmark that is considered to represent the market’s movement as a whole.

The excess return of an investment relative to the return of a benchmark index is the investment’s alpha. Alpha may be positive or negative and is the result of active investing. Beta, on the other hand, can be earned through passive index investing.

Understanding Alpha

Alpha is one of five popular technical investment risk ratios. The others are beta, standard deviation, R-squared, and the Sharpe ratio. These are all statistical measurements used in modern portfolio theory (MPT). All of these indicators are intended to help investors determine the risk-return profile of an investment.

Active portfolio managers seek to generate alpha in diversified portfolios, with diversification intended to eliminate unsystematic risk. Because alpha represents the performance of a portfolio relative to a benchmark, it is often considered to represent the value that a portfolio manager adds to or subtracts from a fund’s return.

In other words, alpha is the return on an investment that is not a result of a general movement in the greater market. As such, an alpha of zero would indicate that the portfolio or fund is tracking perfectly with the benchmark index and that the manager has not added or lost any additional value compared to the broad market.

The concept of alpha became more popular with the advent of smart beta index funds tied to indexes like the Standard & Poor’s 500 index and the Wilshire 5000 Total Market Index. These funds attempt to enhance the performance of a portfolio that tracks a targeted subset of the market.

Despite the considerable desirability of alpha in a portfolio, many index benchmarks manage to beat asset managers the vast majority of the time. Due in part to a growing lack of faith in traditional financial advising brought about by this trend, more and more investors are switching to low-cost, passive online advisors (often called roboadvisors​) who exclusively or almost exclusively invest clients’ capital into index-tracking funds, the rationale being that if they cannot beat the market they may as well join it.

Moreover, because most “traditional” financial advisors charge a fee, when one manages a portfolio and nets an alpha of zero, it actually represents a slight net loss for the investor. For example, suppose that Jim, a financial advisor, charges 1% of a portfolio’s value for his services and that during a 12-month period Jim managed to produce an alpha of 0.75 for the portfolio of one of his clients, Frank. While Jim has indeed helped the performance of Frank’s portfolio, the fee that Jim charges is in excess of the alpha he has generated, so Frank’s portfolio has experienced a net loss. For investors, the example highlights the importance of considering fees in conjunction with performance returns and alpha.

The Efficient Market Hypothesis (EMH) postulates that market prices incorporate all available information at all times, and so securities are always properly priced (the market is efficient.) Therefore, according to the EMH, there is no way to systematically identify and take advantage of mispricings in the market because they do not exist.

If mispricings are identified, they are quickly arbitraged away and so persistent patterns of market anomalies that can be taken advantage of tend to be few and far between.

Empirical evidence comparing historical returns of active mutual funds relative to their passive benchmarks indicates that fewer than 10% of all active funds are able to earn a positive alpha over a 10-plus year time period, and this percentage falls once taxes and fees are taken into consideration. In other words, alpha is hard to come by, especially after taxes and fees.

Because beta risk can be isolated by diversifying and hedging various risks (which comes with various transaction costs), some have proposed that alpha does not really exist, but that it simply represents the compensation for taking some un-hedged risk that hadn’t been identified or was overlooked.

Seeking Investment Alpha

Alpha is commonly used to rank active mutual funds as well as all other types of investments. It is often represented as a single number (like +3.0 or -5.0), and this typically refers to a percentage measuring how the portfolio or fund performed compared to the referenced benchmark index (i.e., 3% better or 5% worse).

Deeper analysis of alpha may also include “Jensen’s alpha.” Jensen’s alpha takes into consideration the capital asset pricing model (CAPM) market theory and includes a risk-adjusted component in its calculation. Beta (or the beta coefficient) is used in the CAPM, which calculates the expected return of an asset based on its own particular beta and the expected market returns. Alpha and beta are used together by investment managers to calculate, compare, and analyze returns.

The entire investing universe offers a broad range of securities, investment products, and advisory options for investors to consider. Different market cycles also have an influence on the alpha of investments across different asset classes. This is why risk-return metrics are important to consider in conjunction with alpha.

Examples

This is illustrated in the following two historical examples for a fixed income ETF and an equity ETF:

The iShares Convertible Bond ETF (ICVT) is a fixed income investment with low risk. It tracks a customized index called the Bloomberg U.S. Convertible Cash Pay Bond > $250MM Index. The 3-year standard deviation was 18.94%, as of Feb. 28, 2022. The year-to-date return, as of Feb. 28, 2022, was -6.67%. The Bloomberg U.S. Convertible Cash Pay Bond > $250MM Index had a return of -13.17% over the same period. Therefore, the alpha for ICVT was -0 12% in comparison to the Bloomberg U.S. Aggregate Index and a 3-year standard deviation of 18.97%.

However, since the aggregate bond index is not the proper benchmark for ICVT (it should be the Bloomberg Convertible index), this alpha may not be as large as initially thought; and in fact, may be misattributed since convertible bonds have far riskier profiles than plain vanilla bonds.

The WisdomTree U.S. Quality Dividend Growth Fund (DGRW) is an equity investment with higher market risk that seeks to invest in dividend growth equities. Its holdings track a customized index called the WisdomTree U.S. Quality Dividend Growth Index. It had a three-year annualized standard deviation of 10.58%, higher than ICVT.

As of Feb. 28, 2022, DGRW annualized return was 18.1%, which was also higher than the S&P 500 at 16.4%, so it had an alpha of 1.7% in comparison to the S&P 500. But again, the S&P 500 may not be the correct benchmark for this ETF, since dividend-paying growth stocks are a very particular subset of the overall stock market, and may not even be inclusive of the 500 most valuable stocks in America.

Alpha Considerations

While alpha has been called the “holy grail” of investing, and as such, receives a lot of attention from investors and advisors alike, there are a couple of important considerations that one should take into account when using alpha.

1. A basic calculation of alpha subtracts the total return of an investment from a comparable benchmark in its asset category. This alpha calculation is primarily only used against a comparable asset category benchmark, as noted in the examples above. Therefore, it does not measure the outperformance of an equity ETF versus a fixed income benchmark. This alpha is also best used when comparing the performance of similar asset investments. Thus, the alpha of the equity ETF, DGRW, is not relatively comparable to the alpha of the fixed income ETF, ICVT.
2. Some references to alpha may refer to a more advanced technique. Jensen’s alpha takes into consideration CAPM theory and risk-adjusted measures by utilizing the risk-free rate and beta.

When using a generated alpha calculation it is important to understand the calculations involved. Alpha can be calculated using various different index benchmarks within an asset class. In some cases, there might not be a suitable pre-existing index, in which case advisors may use algorithms and other models to simulate an index for comparative alpha calculation purposes.

Alpha can also refer to the abnormal rate of return on a security or portfolio in excess of what would be predicted by an equilibrium model like CAPM. In this instance, a CAPM model might aim to estimate returns for investors at various points along an efficient frontier. The CAPM analysis might estimate that a portfolio should earn 10% based on the portfolio’s risk profile. If the portfolio actually earns 15%, the portfolio’s alpha would be 5.0, or +5% over what was predicted in the CAPM model.