Any bond can do well when the conditions are favorable; it is only under acid test of depression that the advantages of strong over weak issues become manifest and vitally important. Confidence in the ability of bond issue to weather depression may be based either on
a) Character of the industry and the particular business will be immune from a drastic shrinkage in earning power. The distinction is between those which are more and less affected by depression. The more stable the type of enterprise, the better suited it is to bond financing and the larger the portion of the supposed normal earning power which may be consumed by interest charges. If there is such a lack of inherent stability as to make survival of the enterprise doubtful under unfavorable conditions, then the bond issue cannot meet the requirements of fixed value investment, even though the margin of safety, measured by past performance, may be exceedingly large. Such a bond will meet the quantitative but not the qualitative test, but both are essential to our concept of investment.
b) The margin of safety is so large that it can undergo such shrinkage without resultant danger. The margin of safety would be dependent on the character of the industry.
Even though the conditions prevalent in the depression years may not be duplicated, the behavior of various types of securities at the time should throw a useful light on investment problems.
Proper Theory of Bond Financing: A reasonable amount of funded debt is of advantage to a prosperous business, because the stockholders can earn a profit about interest charges through the use of the bond holder’s capital. It is desirable for both the corporation and the investor that the borrowing is limited to an amount which can safely be taken care of under all conditions.
Graham suggests that an investor should reconcile himself to accepting an unattractive yield from the best bonds rather than risking his principle in a second grade issue for the sake of a large coupon.
In the traditional theory of bond investment a mathematical relationship is supposed to exist between the interest rate and the degree of risk incurred. The interest return is divided into two components, the first constituting of pure interest – the rate obtainable with no risk of loss – and the second representing the premium obtained to compensate for the risk assumed. This theory assumes that bond interest rates measure the degree of risk on some reasonable precise actuarial basis. It would follow that, by and large, the return from high and low yielding investments should tend to equalize, since what the former gains in income would be offset by their greater percentage of principal losses, and vice versa.
Graham does not believe such a mathematical relationship exists between yield and risk. Security prices and yields are not determined by any exact mathematical calculation but they rather depend upon the popularity of the issue. This popularity reflects in a general way the investor’s view as to the risk involved, but it is also influenced largely by other factors, such as the degree of familiarity of the public with the company and the issue and the ease with which the bond can be sold.
The relationship between different kinds of investments and the risk of loss is entirely too indefinite and to variable with changing conditions, to permit sound mathematical formulation. This is particularly true because investment losses are not distributed fairly evenly in point of time, but tend to be concentrated at intervals, i.e. during periods of general depression.
The main objections to sacrificing safety for yield is that (a) such a policy requires wide distribution of risk in order to minimize the influence of luck by holding a large number of different bonds and (b) more importantly the danger that many risky investments may collapse together in a depression period, so that the investor in high yielding issues will find a period of large income followed suddenly by a deluge of losses of principal.
The bond buyer is neither financially nor psychologically equipped to carry on extensive transactions involving setting up of reserves out of regular income to absorb losses in substantial amounts suffered at irregular intervals. Graham may not have this objection for say a fund manager who does not have these constraints.
Graham suggest that while risk of losing principal should not be accepted merely by a higher yielding coupon, he does not object to purchasing a bond at a substantial discount to par. While these are mathematically equivalent, the psychological difference is important. The purchaser of low priced bond is aware of the risk and is more likely to make a thorough investigation of the issue and carefully appraise the chance of profit and loss.
Graham suggests it would be sounder to start with definite standards of safety, which all bonds must be required to meet in order to be eligible for further consideration. Issuing failing to meet these minimum requirements should automatically be disqualified as straight investments regardless of high yield, attractive prospects, or other grounds. Essentially, bond selection should consist of working upward from definite minimum standards rather than working downward in haphazard fashion from some ideal but unacceptable level of maximum security.