Assets selection in portfolio management

Assets selection in portfolio management

An investor, having decided on the proportions of the portfolio that he or she would like to invest in stocks, bonds and real assets, has to decide on exactly what stocks he or she will hold in the stock portion of the portfolio, what bonds in the bond portion and what real assets in the real asset portion. This asset selection decision, like the asset allocation decision, can be an active one, where the investor attempts to buy undervalued assets in each asset class (or sell overvalued ones) or a passive one, where the investor invests across assets in an asset class, without attempting to make judgments on under or over valuation.


For example, according to the SAA a portfolio should invest, say, 10% of its assets in US equities, while their benchmark is the S&P 500 Index. The investor can select from:

  • Individual constituency stocks of the S&P 500 (e.g. Apple, Exxon Mobil, Microsoft, IBM and so on)
  • ETFs or passive funds tracking the S&P 500 Index
  • Actively-managed funds aiming to outperform the S&P 500 Index
  • A combination of funds benchmarked against the Russell 1000 Value Index and Russell 1000 Growth Index
  • A long futures contract on the S&P 500 Index
  • A segregated portfolio managed by a portfolio manager picking stocks
  • A combination of some or all of these

The question that investment selection aims to answer is how to choose from among all these different options. Investment selection must be linked to the investment strategy. Capital Market Assumptions for SAA are normally derived from historic returns of passive indices, each representing an asset class. The assumption is that the investments under each asset class either have similar return and risk characteristics to those of the passive index used for deriving CMAs. Careful investment selection is necessary to keep the link between the portfolio and the investment objectives.

Investment selection has three main objectives:

  • Identifying investments that match as closely as possible the desired exposures as per the asset allocation (beta)
  • Identifying investments that potentially add outperformance (alpha)
  • Evaluating whether investments are likely to meet their expected return and risk

Beta exposures align portfolios with their investment strategy so actual portfolios closely reflect the asset allocation, which reflects the investment objectives. Alpha adds to investment returns, either across asset classes (TAA) or within asset classes (investment selection).

The selection of a portfolio of securities can be thought of as a multistage process:

  • The first stage consists of studying the economic and social environment and the characteristics of individual companies to produce a set of forecasts of individual company variables
  • The second stage consists of turning these forecasts of fundamental data into a set of forecasts of security prices and/or returns and risk measures (valuation process)
  • The third and last stage consists of forming portfolios of securities based on the forecast of security returns.

Some individuals and institutions may be able to outperform historical extrapolation and the consensus forecasts. The theory of efficient markets tells us there is no simple mechanical way to pick winners in the stock market or at least none that will recover its cost of operation. Yet people continue to spend a disproportionate amount of time on both of these endeavors.

Assets selection with valuation models

However, the determinants of common stock prices are quite easy to specify in general terms. The price of common stock is a function of the level of a company’s earnings, dividends, risk, the cost of money, and future growth rate. While it is easy to specify these broad influences, the implementation of a system that uses these concepts to successfully value or select common stocks is a difficult task.

Discounted cash flow models are based on the concept that the value of a share of stock is equal to the present value of the cash flow that the stockholder expects to receive from it. This is equivalent to the present value of all future dividends. A number of different assumptions about growth-rate patterns have been made and embodied in valuation models.

  • One-period, constant growth over an infinite amount of time assume the firm will maintain a stable dividend policy (keep its retention rate constant) and earn a stable return.
  • Two-period, growth for a finite number of years at a constant rate, then growth at the same rate as a typical firm in the economy from that point on. Dividends and earnings had two distinct growth rates, g1 and g2. As the period of high growth draws to an end, more of the contribution comes from dividends and less from capital gains. The growth in price declines each period until the period of extraordinary growth is over and the growth in price equals the long-term growth rate in earnings and dividends. Once steady state occurs, the single-period growth model is appropriate, and earnings, dividends, and the price will grow at the same rate.
  • Three-period, growth for a finite number of years at a constant rate (first period), followed by a transitional period during which growth declines to a steady-state level; growth is then assumed to continue at the steady-state level into the indefinite future (third and final period).

The majority of security analysts still value common stocks by applying some sort of earnings multiple (price-earnings ratio) to either present earnings, normalized earnings, or forecasted earnings. Another approach is the use of cross-sectional regression analysis to define the weights the market places on a set of hypothesized determinants of common stock prices earnings, growth, risk, time value of money, and dividend policy and to define the average relationship between each of these variables and price-earnings ratios to form an estimate of the theoretical P/E ratio. This equation represents the estimate at a point in time of the simultaneous impact of the variables on the price-earnings ratio. It is not uncommon for these models to explain more than 80% of the difference in stock prices at a point in time. The theory behind their use in finding under and overvalued securities is that the market price will converge to the theoretical price before the theoretical price itself changes. There is no doubt that cross-sectional regression models are helpful in understanding what has happened in the market over time. However, the evidence they may prove of some use in selecting stocks at this time is not conclusive.

The value of earnings

Several studies have shown that knowledge about past and future earnings can lead to investors earning superior returns despite ambiguity in measuring earnings. Kormendi and Lipe (1987) construct a measure of earnings persistence. Persistence captures the permanence of an earnings change, that is, how much the earnings change continues to the future. The more persistent is an earnings change, the greater is its impact on the entire future earnings series, and thus the greater its stock price impact should be. Kormendi and Lipe also estimated the stock market’s response to a firm’s annual earnings news, often referred to as the earnings response coefficient (ERC). ERC is positively correlated with earnings persistence; that is, the stock market responds more to the earnings news of firms whose earnings changes are more persistent. This indicates that the market understands the time-series properties of a firm’s earnings, and stock prices adjust accordingly to earnings information. Collins and Kothari show that the ERC is also positively correlated with firm growth (since greater growth leads to greater future earnings) and negatively correlated with a firm’s risk (beta) and market interest rates (since higher risk and interest rates mean a greater discounting of future earnings). Elton, Gruber, and Gultekin divided firms into deciles by the change in the forecast of earnings. While it was profitable to forecast earnings, it was even more profitable to forecast the change in expectations about future earnings. A number of mutual funds have a strategy of buying high-growth firms. By itself, this should not be a useful strategy. What is important is to find high-growth firms that the market believes will be low-growth firms. Even more valuable would be to forecast changes in the market’s belief about the future growth of a firm.

A recent series of articles suggests that when earnings are extreme, the concept of reversion to the mean might provide useful information when forecasting earnings changes. The economy is highly competitive. The earnings of a company are subject to a large number of uncertainties not under management control (strikes, mineral discoveries, regulatory changes, foreign competition, changing tastes) that are the dominant influence on a company’s results year on year. The counterargument is that there are a number of companies with monopoly control of the markets, with patent protection on unique products, or with superior management, and these companies are able to sustain a high level of growth over a long period of time. These results have led researchers to speculate that earnings changes might be independent from period to period (independence in earnings) because earnings are determined by a physical process, while stock prices are determined by expectations. It is reasonable to assume that changes in expectations cannot be predicted from past data or they would already be incorporated in the expectations.

Ou and Penman (1989) and Levy and Thiagarajan (1993) show that past accounting data can forecast future earnings. Ou and Penman estimate a logit model that explains the sign of next year’s change in earnings using a large set of financial statement items (accounting information). The model is primarily data-driven. Since the model estimates the probability of a positive earnings change, they call the fitted prediction PR. Moreover not only can PR predict the sign of next periods earnings change, it can also predict future stock returns. Specifically, Ou and Penman show that a portfolio that is long in stocks with Pr 0.6 (firms with a relatively high probability of a positive earnings change) and short in stocks with a Pr 0.4 (firms with a relatively low probability of a positive earnings change), earns positive future risk-adjusted stock returns over the 36 months subsequent to the portfolio formation date. The combination of both earnings and return predictability implies both that fundamental analysis is useful for forecasting accounting earnings, and that the stock market does not appear to appreciate the forecasting value of the accounting data. The market is not semi-strong form efficient with respect to accounting information.


Elton, E., Gruber, M., Brown, S., Goetzmann, W. (2007). Modern Portfolio Theory and Investment Analysis, 11th edition.

Leave a Reply

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

This site uses Akismet to reduce spam. Learn how your comment data is processed.

error: Hey, drop me a line if you want some content!!