Let’s Assume Taxes

In his career at the Bell System, Salomon Brothers, and his own firms, Andy Kalotay examined matters others ignored. He also developed a critical view of the public finance industry.

Bell Labs

I graduated in 1968 with a math PhD from the University of Toronto and joined AT&T’s Bell Labs in New Jersey. I was excited about going to work there. Bell Lab scientists synced movie motion and sound in 1926, developed a computer in 1937, launched satellite communications in the 1960s, and did pioneering work on sonar, lasers, and solar cells. They won Nobel prizes for their work on electrons, transistors, and cosmic microwaves. But between the time I interviewed and the time I arrived, there was a big re-organization. The job I was promised didn’t exist and I was disappointed in my new assignment in the Picturephone group.

My job was to figure out the locations of potential Picturephone buyers so AT&T could start wiring cities. There were supposedly going to be one million Picturephones in use, but the origin of that estimate was accidental. In the early 1960s, in preparation for a press conference, AT&T’s chairman asked Bell Labs for a sales estimate. They faxed a memo estimating 600,000 Picturephones by 1980. But when the fax arrived at AT&T’s New York headquarters, the smudged six looked like an eight and the chairman decided to make it an even million. The Picturephone group embraced the number and “One Million Picturephones by 1980” became our slogan.

I could not figure out to whom or where AT&T would sell one million Picturephones. When I visited sales managers at New York headquarters, I casually said, “You know, I just can’t see one million Picturephones. Maybe 200 or 300 thousand.” The sales guys were ecstatic because they didn’t believe they could sell one million Picturephones. But the million number had been endorsed by Bell Labs and the sales managers couldn’t argue with the PhDs.

Of course, Bell Labs’ managers didn’t like that I had contradicted them. But meanwhile, the AT&T headquarters guys told my bosses, “We love Andy’s research. Give him all the support he needs.” This created tension that I didn’t ameliorate by continually complaining about Bell Labs’ endemic bureaucracy. In 1973 I took a promotion to work at AT&T headquarters. The Picturephone? A new AT&T chairman killed it after sales stalled at 485 units. Now we have Zoom.

My first job at AT&T was in the prestigious Corporate Planning Group. I was excited because the group was composed mostly of former Bell Labs scientists, and it reported directly to AT&T’s CEO. And again, I was disappointed. I learned a valuable lesson: if you are interested in planning, never work for an entity with “planning” in its name. Planning groups are typically isolated entities with no practical experience and no power to implement their ideas.

But I got an offer to work in AT&T Treasury’s financial research group and joined in the fall of 1974. I wanted a more formal finance education than what I had picked up at various projects at Bell Labs and AT&T, so I enrolled in the executive MBA program held onsite at AT&T. Although the program was under the aegis of Pace University, AT&T brought in professors from Princeton and Harvard to teach. My very first finance class was taught by Burt Malkiel, and it had an incredible effect on my career. Professor Malkiel is famous for his work on the efficient-market hypothesis and his book for ordinary investors, A Random Walk Down Wall Street. But he did his PhD thesis on the term structure of interest rates and wrote a book on that subject.

Assume No Taxes

In his first class, Malkiel spoke on forward interest rates. I would learn later that forward interest rates were important for risk management and predicting future interest rates, but Malkiel motivated this lecture with an investment question. Is it better to buy a two-year note or to instead buy a one-year note, wait for it to mature in a year, and then buy another one-year note? Which will give you the best return over two years? That depends on what the one-year rate is in one year (the one-year rate one year forward). What rate would make the two investment outcomes equal each other?

Malkiel wrote this formula on the blackboard:

(1+ R01) * (1+ R11) = (1 + R02)^2

Where: R01 = the one-year rate now (time 0, 1-year rate)

              R11 = the one-year rate one year forward (time 1, 1-year rate)

  R02 = the two-year rate now (time 0, 2-year rate)

If you know R01 and R02, simple algebra solves for R11. For example, if the one-year rate is 4.0% and the two-year rate is 4.4%, the break-even one-year rate one year forward is 4.8%.

This was the first thing that he showed the class because we were complete beginners. But I knew a little bit about taxes and at the end of the lecture I went up to Malkiel and asked “What if I pay tax on the interest? What’s my break-even then?” And he looked at me and he said, “I don’t know.” He wrote a book on this stuff, but he never considered taxes. That’s typical of academics. Finance professors often make the simplification “assume no taxes” before they start their analysis. Malkiel finally said, “I don’t think it makes any difference.” That was his guess.

On the train home I looked at Malkiel’s equation and in about five minutes I saw that taxes affect the break-even rate. I thought, “Heck, if this guy doesn’t know it, this could be important.” And I was right. Thinking carefully about taxes turned out to be very important for my career.

Look at any finance textbook. It talks about Modigliani & Miller and that’s about the extent of the textbook’s tax discussion. Taxes don’t get the attention they deserve from academics, considering their importance. Transactions are analyzed pre-tax. For example, a 1975 Journal of Finance[1] article addressed bond refunds. It showed why it makes sense to call bonds and issue new bonds with a lower coupon if rates go down far enough. This is the standard bond refund. But then the article talked favorably about repurchasing bonds when rates go higher and the bonds trade at a discount. But the paper ignored the fact that the discount below par is taxable income to the issuer buying back its debt. If a corporation’s tax rate is 40%, the after-tax cost of repurchasing its own bond at 70 is 82; the 70 price of the bond plus 12 in taxes: 40% times the 30-point discount below par.

I sent a short note to the Journal of Finance explaining how the article had overlooked taxes, but the editor rejected it. Being a naive youngster, I wrote to him again, saying, “If you reject my paper, you should tell me why you rejected it.” And he wrote back, “I don’t like taxes and I won’t publish your paper.” I eventually got my note published in Financial Management.[2]

Another mistake academics make is to assume that everyone’s tax rate is the same. A 1979 Journal of Finance article[3] correctly showed that original issue discount bonds (OIDs) provide a tax benefit to the issuer, while investors holding such bonds in taxable accounts experience an off-setting disadvantage. But then, assuming that bonds are held only in taxable accounts, it concluded that OID bonds would never be issued. It was a bold and badly mistaken prediction. Within two years, OID bonds, including zero-coupon bonds, were issued in the billions, purchased by tax-exempt entities such as pension funds.

Even commonly used risk-management software makes fatal simplifying assumptions regarding taxes. For example, they estimate the duration of a 10-year 2.75% coupon municipal bond priced to yield 3% at 9 years. But that is only because the analytics ignore the fact that the muni’s discount below par is taxable to the purchaser as it accrues to par. The bond’s price is really more sensitive to interest rate moves. To yield more than 3% after tax, the bond’s price must decline sufficiently to pay tax on the additional discount below par. And then the bond’s price must decline again to pay taxes on the additional discount! When all the discounting and tax-paying is done, after-tax duration is 13 years, longer than the bond’s maturity! A one percentage point increase in rates results in a 13% price decline.[4]

Calling Bell System Bonds

AT&T sent me to the annual conference put on by the Center for Research in Security Prices at the University of Chicago. I heard a talk on options pricing theory – a hot new topic, since Black-Scholes’ seminal paper had only been published a few years earlier, in 1973. The speaker was Edward O. Thorp, mathematics professor and hedge fund manager. Outside finance, he is known for his book, Beat the Dealer, which proved that card counting at the blackjack table works. After the conference, we were both stuck at O’Hare waiting for our delayed flights home and I had the chance to ask him, “These pricing models only address stock options. What about options on bonds?” Thorp replied, “I know nothing about that. It would be a great question to study.”

This was the second time I’d stumped a distinguished professor. And he was right, it was a great question to study, particularly for me at AT&T. The telephone companies within the Bell system had continuously issued 40-year bonds callable after five years. If Bell system’s borrowing rate had declined since issuance, making it practical to the call our bonds, should we wait for the rate to decline further? But our borrowing rate could rise, instead! And while we were trying to decide whether to call our bonds, they were still requiring debt service at their high coupons.

In 1979, my Bell Labs colleague Bill Boyce and I published “Optimum Bond Calling and Refunding.”[5] Later that same year we published “Tax Differentials and Callable Bonds[6]” in The Journal of Finance, which now had a new, tax-friendly, editor. In the second article, we showed that it’s better for taxable corporations to issue callable bonds than optionless bonds. And using essentially the same analytical method, last year, 44 years later, I published “Callable Tax-Exempt Bonds Are Too Costly”[7] showing that tax-exempt issuers should issue noncallable bullet maturity bonds. These papers, and much of my work between them, was a consequence of thinking about taxes after Burt Malkiel’s lecture and thinking about option pricing after Edward Thorp’s lecture.

Going to the Street

I thought the head of financial planning at AT&T treasury was inept, and still being a naïve youngster, told the treasurer, “Look, we have important, interesting problems that he doesn’t understand because he doesn’t know enough finance.” I was fortunate because he might have reprimanded me for speaking ill of my boss or jumping the chain of command. Instead, the treasurer’s response was, “I agree with you a hundred percent, but I don’t know what to do with him. I can’t fire or demote him; he’s a senior person.” This was how things worked at AT&T; seniority counted for a lot. The treasurer’s solution? “Why don’t you go to New York Telephone for a couple of years until this guy retires and then come back and do your thing.”

I did go to New York Telephone, but I realized I didn’t have to stay at AT&T. I could work at a financial services firm instead of working in AT&T’s finance division. Quitting Bell was about as unheard of as AT&T firing a senior person, but that’s what I did. I think I was the first person from Bell Laboratories to move to Wall Street, but others followed. For example, Emanuel Derman joined Goldman Sachs and I recruited Bill Boyce to Salomon Brothers.

My first stop on Wall Street was Dillon Read in 1979. It was one of the premier old-line investment houses. Douglas Dillon, son of the founder, would still visit the office from time to time. He had been chairman after the war; leaving the firm to become Ambassador to France under Eisenhower and Treasury Secretary under Kennedy. Nicolas Brady was now head of the bank, and he would become a US Senator and Treasury Secretary under Reagan and Bush I.

The people at Dillon Read were very wealthy and had fantastic connections, but they weren’t very technical and didn’t have much interest in bond analytics, which is what I did for them. So, in 1981, when Marty Leibowitz asked if I would be interested in joining Salomon Brothers’ Bond Portfolio Analysis Group, I jumped ship again.

One can only guess what the patricians thought of Dillon Read’s trajectory over the next 26 years. Sold to Travelers in 1986, then sold to Barings in 1991, then sold to Swiss Bank in 1997, merged into UBS in 1998, until finally the Dillon Read name was abandoned. But then the name was revived in 2005 to christen UBS’ new proprietary trading entity. And ignominiously scrapped a second and final time when that subsidiary lost $3 billion in 2007 subprime mortgage meltdown and was shut down.

Salomon Brothers

I knew Marty Leibowitz from my days at AT&T when Salomon Brothers pitched us financing deals. Marty wanted me to lead a group dedicated to debt management strategies for corporate bond issuers. This contrasted with most people in Bond Portfolio Analysis who took an investors’ perspective. I named my group “Special Studies” instead of the more descriptive “Debt Management,” which made the group rather mysterious inside and outside of Salomon. Unlike a lot of the research analysts, my group tried to bring some humor to the job. We came up with a promotional handout titled “The Critics Rave” about our computer program called “DebtCost.” It had 15 obviously made-up quotes attributed to our clients such as “I never realized scenario analysis could be so exciting” and “DebtCost has changed my life.”

Much of what my group and I did at Salomon had to do with tax-driven transactions. I mentioned previously the problem when an issuer buys back its own debt at a discount: the discount below par is taxable income. If a corporation’s tax rate is 40%, the after-tax cost of repurchasing its bond at 70 is 82; the 70 price of the bond plus 12 in taxes: 40% times the 30-point discount below par.

When I started at Salomon in 1981, the 30-year Treasury traded over 15%, up from 8% four years earlier, so there were a lot of corporate bonds trading at discount. We came up with a scheme if an issuer wanted to buy back its debt: equity-for-debt swaps. Normally these swaps are done to help a distressed company having trouble meeting its interest expense. The company issues equity in exchange for its hard-to-service debt. But our idea had nothing to do with that. We did these swaps to keep the issuer from paying taxes on the discount below par.

In a typical deal, Salomon would accumulate the company’s discounted bonds through piecemeal trades in the market or in larger privately arranged sales. When we acquired the amount of bonds the issuer wanted to retire, we would exchange the bonds for the issuer’s stock, which we would then sell. The reason the issuer avoided the tax on the discount below par is that under the US tax code, transactions entered into by corporates on their own stock are neither taxable nor tax-deductible. We did the first equity-debt swap for Quaker Oats and one of the largest, $125 million, for Duke Power.

By 1986 the 30-year Treasury had gone down to 8% and bonds issued when rates were higher (1979 - 1985) were trading at a premium to par. Now the issuer could pay cash if it wanted to buy back its debt and the premium above par was an immediate tax deduction. I think Salomon did 48 of the first 50 transactions; buying back debt at a premium and issuing new debt to replace it.

I mentioned zero-coupon bonds and the misconception academics had that these would never be issued, because the professors erroneously assumed issuers and investors had the same tax rate. But with our help, corporations started issuing zero-coupon bonds around 1979, and they came in volume in 1981-82. One contributing fact was that the US tax code allowed the issuer to deduct the bond’s accretion to par in a straight line. Say a corporation issued a 30-year zero for proceeds of 10% of par. It could expense 1/30th of that 90 discount each year. This reduces taxes without incurring a cash expense. Of course, the sensible way to accrete to par is not in a straight line producing the same dollar amount each year. The accretion should be on a constant yield basis, producing a small initial annual expense that grows over time. Eventually the tax code was changed in this regard from straight line to constant yield and zero-coupon bonds became less attractive.

After Salomon

I left Salomon in 1990 and formed my own advisory and software company. Part of my work was in public finance, advising both issuers and investors. The tax situation is very different, of course. Municipal bond issuers, unlike corporate bond issuers, are tax exempt. Municipal bond investors avoid tax on their interest income, but can still incur taxable gains and deductible losses when they sell a municipal bond before maturity. Again, the implications of these differences are woefully understood or ignored by academics.

In 2014 I published a paper[8] extending option-adjusted spread analysis to munis, by introducing the tax treatment of the gain resulting from purchasing bonds at a discount. I applied for a patent for this idea, and eventually received it in 2023. That year I also published “Callable Tax-Exempt Bonds Are Too Costly”[9] and “Tax-Loss Harvesting Municipal Bonds: A Primer.”[10] The first showed that tax-exempt issuers should issue noncallable bullet maturity bonds. The second called attention to the surprising but poorly understood fact that a tax-beneficial sale of a muni might sacrifice value.

This year Public Finance Journal will publish my article on how municipalities wasted billions of dollars between 2018 and 2021 by engaging in taxable advance refundings, when they should have waited until their bonds were callable and refunded them with tax-exempt debt. In July, I will present a paper about another wasteful transaction currently in vogue, tendering for tax-exempt bonds, at the Brookings Municipal Finance Conference. Based on my experience, my findings will go unchallenged, yet short of a major shake-up of the municipal finance industry, the waste will continue unabated.

It troubles me that municipalities often practice poor debt management. They agree to bad deals, and the damage adds up to billions of dollars in needless taxpayer expense. Independent advisors are supposed to help municipalities in their negotiations with security firms, but they are often ineffectual. Advisors often endorse whatever bond underwriters recommend. Until 2010, there was no certification process to be a municipal advisor and many of them were incompetent. I was on the Municipal Securities Rulemaking Board’s committee to come up with certification standards for municipal advisors, but I couldn’t push through what I consider to be adequate qualifications.

Meanwhile, public officials have other priorities than getting the best financing deal for their entities. I was advisor on a deal for Chicago when the city’s director of finance called and asked if I could attend a meeting the very next day. So I dropped everything, got on a plane, and made it in time to city hall. The mostly black city officials had called in everyone who was working on their bond deal for an emergency meeting on short notice; bankers, lawyers, and advisors; because they wanted some answers. But they weren’t interested in our qualifications, or in the maturity schedule of their debt, or whether it was priced optimally. Instead, the whole purpose of the emergency meeting, the only thing discussed, was how many minorities were working on their deal and whether they were being adequately compensated.

Andy sold Andrew Kalotay Analytics to ICE Data Services in 2021. He is now president of Kalotay Advisors and continues to be active in municipal finance. He is the author of Interest Rate Risk Management of Municipal Bonds.

[1] Ang, James, “The Two Faces of Bond Refunding,” Journal of Finance, June 1975, 30(3), pp. 869-874.

[2] Kalotay, Andrew, “On the Advanced Refunding of Discounted Debt,” Financial Management, Summer 1978, 7(2), pp. 14-18.

[3] Livingston, Miles, “A Note on the Issuance of Long-Term Pure Discount Bonds,” Journal of Finance, March 1979, 34(1), pp. 241-246.

[4] Kalotay, Andrew, “Interest Rate Sensitivity of Tax-Exempt Bonds Under Tax-Neutral Valuation,” Journal of Investment Management, First Quarter 2014, 12(1), pp. 62-68.

[5] Boyce, W. M. and A. J. Kalotay, “Optimum Bond Calling and Refunding,” INTERFACES, November 1979, 9(5), pp. 36-49.

[6] Boyce, William and Andrew Kalotay, “Tax Differentials and Callable Bonds,” Journal of Finance, September 1979, 34(4) pp. 825-838.

[7] Kalotay, Andrew, “Callable Tax-Exempt Bonds Are Too Costly,” The Journal of Fixed Income, Spring 2023, 32(4) pp. 77-82.

[8] Kalotay, Andrew, “The Interest Rate Sensitivity of Tax-Exempt Bonds Under Tax-Neutral Valuation,” Journal of Investment Management, first quarter 2014, 12(1).

[9] Kalotay, Andrew, “Callable Tax-Exempt Bonds Are Too Costly,” The Journal of Fixed Income, Spring 2023, 32(4) pp. 77-82.

[10] Kalotay, Andrew, “Tax-Loss Harvesting Municipal Bonds: A Primer,” The Journal of Wealth Management, Fall 2003.

To comment on a story or offer a story of your own, email Doug.Lucas@Stories.Finance

Copyright © 2024 Andrew Kalotay. All rights reserved. Used here by permission. Short excerpts may be republished if Stories.Finance is credited or linked.