Markowitz [12] and Tobin [19] pioneered in the development of a portfolio selection model resting on the assumptions that the investor

1. Chooses among alternative investment opportunities solely on the basis of expected return (E) and standard deviation of return 〈σ〉, and

2. Prefers more expected return to less but will refuse to incur additional risk (measured by standard deviation) unless compensated by increased expected return.

Recently, Cohen and Gujarati [2] have suggested that when multicollinearity is present there is “ …danger involved in mechanically dropping variables from multiple regression equations by t tests because t values of the regression coefficients may not be significantly different from zero when the true (population) values of these coefficients are in fact not zero…” The problem they discuss is not a new one and has been extensively treated in the existing literature. However, their approach is straightforward and will certainly aid the practitioner in his understanding of the problems associated with multicollinearity.

A recent article in this journal [1] described a model for the computation of taxadjusted true yields to maturity on discount bonds and explained the use of a computer routine implementing this model. Unfortunately, the translation of the computer program into equation form contained a number of notational errors. In addition, there was an equals sign missing from the third equation [1, page 267]. As a result, the reader, in attempting to implement the model as it was formulated in the original article, will probably fail.

During the “credit crunch” of 1966, starts of single-family residences fell 19.2 per cent, from 964,000 to 779,000. At that time, 10 states had usury laws limiting the maximum nominal interest rate that could be charged to individuals on residential mortgages to 6 per cent. When interest rates rose to high levels in 1966, there were widespread complaints that this artificial restriction caused residential construction to decline by even larger amounts in those states having 6 per cent usury laws as lenders shifted funds to other states not having this restriction. However, when nominal interest rates rise to the legal maximum, mortgages trade at a discount and the effective yield rises above the legal rate. Presumably, it is the effective yield, not the nominal rate, that is relevant for directing the flow of funds into alternative investments. Thus, there is good reason to doubt that the usury laws did restrict residential construction in spite of the contention by those in the mortgage and construction industries. The purpose of this study is to determine if—and if so, to what extent—state usury laws caused a decline in residential construction in 1966.

Investment literature, particularly materials made available to investors by financial magazines, brokerage houses, and investment services, places a great deal of stress upon analysis of investment opportunities by industry groups. Examples are industry analyses, such as those of Forbes and Financial Analysts Journal, as well as the industry segregations of popular services such as Value Line, Standard & Poor's and Moody's.

In this paper, a risk-analysis simulation procedure was utiliijed to incorporate both a cash-flow liquidity concept and uncertainty in a liquidity-planning simulation model. The components of cash flow were specified. The model was implemented with the assistance of a large savings bank. The results indicate that a substantial dispersion in probable outcomes exists, from a $1 million outflow to $10 million inflow. The expected net flow, $5 million, greatly exceeds the point estimate derived by simply summing the individual point estimates. In fact, there is a 50 percent chance that the net flow will exceed the point estimate by more than $1.5 million. Such results from the liquidity planning model clearly give the banker a basis for determining the adequacy of his present liquidity position and therefore his cash management policy, as well as the optimum strategy in terms of various adjustment policies.

Professor Tysseland's paper, “Further Tests of the Validity of the Industry Approach to Investment Analysis,” calls into question the validity of the industry approach to investment analysis. His examination of inter- and intra-industry investment results through time leads him to conclude that “the lack of homogeneity within industries is great and industry consistency in performance over time is unimpressive.” Furthermore, he states that a prospective investor who chooses investments based on past industry performance records is on tenuous grounds and that “the usual types of industry data, analyses and forecasts generally made available by financial magazines, brokerage houses, and investment services, may be of doubtful value.”

The number of empirical studies aimed at examination of the relationship between risk and return to securities portfolios has increased dramatically over the last five years. There are basically two features of Nancy Jacob's paper, “The Measurement of Systematic Risk for Securities and Portfolios: Some Empirical Results,” that I feel contribute significantly to this area of study. First, the empirical analysis is based on an axiomatic system of characterizing the securities investment decision; this departs substantially from the assumptions underlying the more familiar mean-variance approach. A comparison of the conclusions reached while empirically investigating different axiomatic systems of investor behavior may play an important role in the validation of positive models of capital markets. The second contribution stems from the depth of the study as it relates to the effects of changes in the observation interval, the horizon time, and the portfolio selection procedures. Determining the appropriate horizon time and observation interval is always a major problem, when working with historical price and dividend data. The author's grouping of data into one-, five-, and ten-year horizon periods - along with varying the observation interval over monthly, quarterly, and annual data - illuminates many of the problems associated with empirical studies using a fixed holding period and observation interval.

This paper has shown that the models developed to select common stock port-folios can be adapted to the selection of real estate portfolios and mixed asset portfolios. The concepts are all identical, and as long as return and risk can be quantified, the problems are soluble.

The portfolios identified using a small sample indicate that real estate portfolios can have more return and less risk than do common stock portfolios. When the two assets are combined, the real estate assets dominate the resultant portfolios. On an after-tax basis these results are more apparent. The local aspect of real estate versus the national aspect of common stocks is primarily responsible for these results.

Careful examination of the behavior of financial variables over time uncovers an important distinction: variables expressed in dollars, such as earnings per share, behave very differently from percentage changes in those same variables. Similarly, financial ratios, such as the price/earnings ratio, behave quite differently from percentage changes in financial ratios. Variables expressed in dollars and financial ratios appear comparatively stable and predictable over time. Successive values are fair approximations of one another. Percentage changes in dollar and financial ratio variables-i.e., growth variables on the other hand, tend to be erratic, volatile, and unpredictable over time. Successive values bear little relation to one another. This distinction between dollar and ratio variables, on the one hand, and percentage changes in these variables-i.e., growth variables-on the other, serves as a useful basis for a financial model.

Harry C. Friedman in his paper, “Real Estate Investment and Portfolio Theory,” has put his mind to an increasingly important area of financial concern, the relationship of real estate as a security in developing portfolio theory. Between now and the year 2000, it has been estimated that $1,500 billion will be spent on building and remodeling nonfarm housing. There will be an estimated $1,000 billion spent on commercial, industrial, and utility construction. In addition, $1,000 billion will probably be spent on public utilities and service institutions, plus $30 billion annually on community facilities. It is anticipated that each year in coming generations we will add to our existing inventory the equivalent of fifteen cities of 200,000 persons each. Predictions are that there will be a need for an additional 2 million dwelling units per year during the 1970's and that this need will climb steadily thereafter.

Murphy and Nelson in their article, “Random and Nonrandom Relationships Among Financial Variables: A Financial Model,” develop three postulates which deal with the temporal behavior of financial variables. They suggest that the three postulates constitute a useful financial model. The basis for the model is the distinction between dollar or ratio variables on the one hand and percentage change or growth variables on the other hand.