Voice from the Past
Chapter 19

One of the exercises we used to perform from time to time was the amendment of resolutions securing outstanding revenue bond issues. Almost every time the purpose was to permit the issuance of additional bonds on a parity with the outstanding issue when the issuer could not satisfy an existing parity test. The original resolution typically had a provision for amending any section with the consent of the holders of three-quarters, or, later, two-thirds of the outstanding bonds, provided that certain actions against a minority of the bondholders were prohibited. Thus the holders of bonds numbered 250 to 1000 could not consent to an amendment preventing the holders of bonds numbered 1 to 249 from receiving interest or principal when due. Though modern book entry bonds are not numbered, the principle remains the same.

The procedure was initiated by a bond dealer who would ask us to prepare the documents for the new bond issue as well as the amendatory proceedings. Early on, the original resolutions called for approval at a meeting of the bondholders, but the meeting requirement was dropped after a while. I suppose that the drafters of the documents had practiced in the Depression when many issuers were unable to pay their bonds, and meetings of bondholders were appropriate measures to determine how to make the best of a bad situation. This was no longer the case after World War II, but changes in boilerplate come slowly.

I remember one such meeting that took place in my office, which was then pretty small. John Shields, the head of Barcus, Kindred and Company, turned up about 10:00 in the morning and displayed the proxies he had received from three or four insurance companies, and these were sufficient to consent to the proposed amendment. Exercising those proxies, John elected himself chairman and me secretary of the meeting. We then proceeded to verify the fact that we already knew: that the proxies were sufficient to carry a vote in favor of the proposed amendment. This fact I duly noted in the minutes of the meeting that I, as secretary, was taking. When the meeting was over, the minutes typed up, and John and I signed them, there was still time for us both to go back to work before lunch.

In most of the amendments, it was sufficient that the necessary consents be filed with the city clerk and no meeting was required. Typically, at the meeting when the governing body determined that the consents had been obtained, the old resolution was duly amended, and the new bond resolution was adopted. I was a little nervous about possible sales of the underlying old bonds by the holders between the time they consented and the time the resolution was adopted. Sometimes we required stamping the bonds, but this was not welcomed by the bondholders, and there was no feasible way to make sure that non- consenting bondholders either stamped their bonds or knew of the amendment. As the bonds were unregistered and passed by delivery, we may have taken the position that disclosure on the face of the bond that the resolution authorizing it could be amended put subsequent holders on notice of the fact that it might have been amended before they took possession.

Cumbersome though the amendatory process was, it was better for an issuer than the alternative, which was to issue subordinate lien bonds if the old bonds were not immediately callable for redemption and refunding. Investors were not at all happy to buy them, so that the principal market for these obligations was the holders of the first lien bonds. Knowing that, such holders could pretty much set interest rates independent of the market.

In time, I learned that the insurance companies or other investors that consented to these amendments generally were the ones who bought the new bonds. The underwriters knew who the old bondholders were because they had sold them the prior issue. This was well and good when the bondholders were a few insurance companies, but if the bonds were spread over scores or hundreds of smaller holders, tracking them down was an expensive chore. Of course, the holders of the old bonds insisted on an opportunity to buy the new bonds at a price with which they were well satisfied.

The practice of amending resolutions authorizing outstanding bonds stopped all of a sudden. Not because of a lesser need to issue revenue bonds that didn't meet a parity test, but because of the advent of advance refunding. When the economic advantages of advance refunding (prior to rules against arbitrage) became known, failing to refund the old bonds made no sense. The old resolutions could stay on the books, and investors cared not that the new bonds bore junior liens because the first lien bonds were defeased. Until the resolutions authorizing the old bonds included appropriate defeasance clauses, our opinions on advance refunding revenue bonds referred to the old bonds, and stated that we had examined evidence that government bonds sufficient, with interest, to pay principal of and interest on the old bonds when due, had been deposited in escrow for that purpose. By the time that the arbitrage rules made it difficult for an issuer to make much money from an advance refunding, the practice of amending outstanding bond resolutions had been forgotten, and advance refunding became the standard procedure for dealing with the problem of parity tests that could not be met. Even without arbitrage benefits, issuers and underwriters benefit: the lack of any need for consent from the holders of the outstanding bonds gives issuers a much larger potential market for the new bonds, and gives dealers many more bonds to sell.

Manly W. Mumford