Bond Battles

From the all-time high of 106 in December 1936, the Dow-Jones bond average fell to a low of 83 in March. Then the Government's desterilization program and the fall in commercial loans, gave bonds a rally quite unlike anything stocks have enjoyed, and the average jumped to 88, has since steadied at 85. Despite this pleasant development, the annual frolic at the Sleepy Hollow Country Club which Wall Street's bond traders enjoyed last week, cost them no sacrifice; for bond buying, like stock buying, is in the doldrums.

One day last week total bond trading on the New York Stock Exchange was a pee-wee $3,270,000. smallest five-hour day in 20 years. The same day, trading in bonds on the over-the-counter market, which has been grabbing more & more of the bond business, was estimated at $18,000,000. Over-the-counter bond trading amounts to wholesaling such as that of banks and insurance companies buying and selling huge blocks at a time ($5,000,000 in one deal is not unusual). Deals this big are virtually impossible on the Exchange because the attendant publicity would cause great price fluctuation. Exchange trading amounts to retail buying, suitable for individuals since the price is determined by open competition.

To consider how to increase bond-trading on the floor of the Exchange, SEC Chairman William O. Douglas last week sat down with a round table of Exchange members, potent Wall Street investment bankers and representatives of the big insurance companies. Ideas broached to aid the Exchange included: 1) lowering the commission charged from one-quarter of a point to the one-eighth generally charged OTC; 2) admitting big bond dealers and institutions to associate membership on the Exchange.

Ratings was another prime topic oi conversation among last week's frolicking bond men. For many years banks buying bonds have generally relied on the ratings published by the four big statistical services--Standard Statistics, Moody's, Poor's and Fitch's.* In 1936 the Comptroller of the Currency made this custom a requirement in cases when bonds are of doubtful value. Last January a research economist at the University of Chicago with the resounding name of Melchior Palyi took it upon himself to denounce this setup. Said he: "The ruling of the Comptroller has put the 'recognized' agencies into a strategic position which may permit them virtually to control the market. The agencies are under no governmental or other supervision . . . are extremely secretive about the technique at which they arrive at their judgment. ..." Mr. Palyi also had a long list of highly technical criticisms of the accuracy and permanence of the ratings. He was presently supported in his stand by Statistician J. Harvie Wilkinson Jr. of the State-Planters Bank & Trust of Richmond in a book called Investment Policies for Commercial Banks.

Rating agency officials retorted that they did not suggest the Comptroller's action, that the record of the 9,000 bond ratings is top-notch and their best defense. Only one of the four to put these ideas into print is Fitch's, now preparing a book. Meanwhile, Associate Professor Gilbert Harold of the University of Oklahoma produced a book called Bond Ratings as an Investment Guide, concluded: "The ratings operate quite effectively to protect the investor against loss. . . . The record is not perfect . . . but it is certainly beyond reasonable criticism."

Last week Wall Street was still chuckling over an incident which somewhat supports criticism of the setup: Fitch's rated the recent $100,000,000 issue of U. S. Steel debentures as AAA (highest), Standard as A1 (second rank), Poor's and Moody's as A (third rank). Last week, too, Secretary of the Treasury Morgenthau revealed that the group which has been devising a uniform bank examination was also brewing substantial modifications of bond ratings and eligibility.

*Each has different symbols, but a fair sample is Standard's: Sound bank quality--A1+, A1 and A; borderline bank quality--B1 + ; speculative--B1, B, C1+, C1 and C; defaulted bonds--D1+, D1 and D.

This file is automatically generated by a robot program, so reader's discretion is required.