The Average Cost of Debt for Companies with Multiple Bonds

Outstanding

Robin Grieves*
Clinical Faculty
Department of Finance
Darla Moore School of Business
University of South Carolina
Columbia, South Carolina 29208
1-251-223-3474
robin.grieves@moore.sc.edu

Mark D. Griffiths
Jack Anderson Professor of Finance
Department of Finance
Miami University
Oxford, Ohio
1-602-882-0092
mark.griffiths@muohio.edu

*Corresponding author

November 2012

0
 The Average Cost of Debt for Companies with Multiple Bonds
Outstanding

1. Introduction

As a first approximation, every company, everywhere in the world that reports

the average interest rate on its debt gets it wrong, every time. Why is this

approximation correct? We looked at S&P 500 companies and selected eight that

report average interest on their long term debt. They all do it wrong.

Those eight companies are audited by all of the Big Four auditing firms. If one

auditing firm made one of its customers report the correct number, it would be

reasonable to suspect that they would make all of their customers report the correct

number. In addition, if one auditing firm made one or all of its customers report the

correct number, the other auditing firms, upon seeing the correct solution, would

almost certainly require all of their customers to use the correct calculation.

If US companies reported the correct number, much of the rest of the world

would follow. In this case, the US is not leading.

If major companies abroad were reporting the correct calculation, one or more

of the US auditing firms would almost certainly observe it and adopt it.

Average interest on long-term debt is one of the important inputs for analyzing

a company and its prospects. Not all companies with multiple bonds outstanding

report average interest on long-term debt, but they should. Even though every

company gets it wrong, divisions within companies may well get it right. For

example, the investments desks of major banks almost certainly use the correct

average interest rate in their decisions, even if the company discloses a mistaken

number.

1
 For some debt structures and during some yield curve scenarios, explained

below, the reporting error is material, which implies that financial analysts should

either transform the company data that are provided or gather their own information.

A recent CFA Institute Centre for Financial Market Integrity publication, A

Comprehensive Business Reporting Model (2007), says, “We must make clear that

our proposed additions and other amendments to the framework are not considered

lightly but are based upon the changes to the current financial statements that

investors routinely must make in order to try to generate the data they need to analyze

and value securities. Perhaps the most compelling argument for requiring that the

reporting changes be made is that if investors must transform financial statements,

and the information they contain, into a different form so that they can use the

information in their decision making, then the statements and information should be

presented in that form in the first place. Such a change would promote both market

efficiency and effectiveness.”1

Currently, there is no requirement that companies report the average yield on

their long-term debt. Consequently, there is no standard for reporting its value.

Nevertheless, many companies do report their average yield presumably because they

believe that the information is useful to investors and other stakeholders. The purpose

of this paper is to demonstrate that the current calculations are incorrect and to present

the correct calculation.

The correct calculation of the average yield on a company’s long-term debt is

important not only for reporting purposes, but for managing the company’s business.

Investment decisions that depend on the company’s weighted average cost of capital

(WACC) include the cost of debt. Companies that misreport the cost of debt are

1
CFA Institute Centre for Financial Market Integrity, A Comprehensive Business Reporting Model
(2007), p. 5

2
likely to be using the same, erroneous value for investment decisions, which can lead

to errors in deciding which projects to pursue and which ones to drop.

For companies to persist in this error is something of a mystery because the

correct answer to bond portfolio yield has been known and published for more than

twenty years for those with access to Wall Street research.2 Fixed income portfolio

managers know that the correct measure of their portfolio yield is the duration-dollar

weighted average of the individual security yields. Duration-dollar weights are

calculated by multiplying the market value of each issue times the bond’s modified

duration or, where applicable, its effective duration. Each value divided by the sum

of the duration-dollars is that issue’s weight in the yield calculation.

Ambrose and Warga (1996) explored average interest measures on fixed

income target portfolios (index ‘bogeys’), which almost always used market value

weighted yield to maturity. The value being sought, the yield to maturity on a

collection of bonds, is the internal rate of return of the collection’s cash flows. Yield

to maturity is the discount rate that makes the present value of the cash flows equal to

the full price of the bonds (quoted prices plus accrued interest). When a company has

two or more bonds outstanding, especially when their maturities are quite different as

in 5-years and 30-years, the current accounting practice of reporting market weighted

average yield of the debt outstanding can be far from the internal rate of return on the

combined cash flows and prices of those bonds.

2
Ambrose and Warga (1996) and Garbade (1988) in Garbade (1996).

3
 2. How Do Companies Report Average Interest Rates and

what is the Correct Calculation?

Let’s consider a simple numerical example. Suppose the XYZ Corp. has $20

million each of a 2-year note and a 30-year bond outstanding. Further suppose that

the 2-year has a 4% coupon priced at par to yield 4% and the 30-year has an 8%

coupon, priced at par to yield 8%. While this sounds extreme, the steepest 2s/30s in

the past thirty years was 401bp in on February 1, 2011. For the eight months,

November 30, 2010 to July 31, 2011, the average slope (both mean and median) was

383bp.

The market weighted average yield to maturity on that (two bond) portfolio of

liabilities is 6%. The duration-dollar weighted average yield to maturity is 7.42%.

That difference is material.

Which is correct? The internal rate of return on the $20 semi-annual coupon

payments for the two year, $40 semi-annual coupon payments on the 30-year, the

$1000 face value of the two year two years hence, the $1000 face value of the 30-year

thirty years hence and the $2000 required to buy one of each of the bonds today is

7.47%. The duration-dollar weighted average yield is quite close while the market

value weighted average is not close at all. 3

The difference between market weighted average and duration-dollar weighted

average depends on how different the maturities (actually, durations) of the

company’s issues are and how steep the yield curve is. If the bond’s maturities (and,

consequently, their durations) are close, duration-dollar weights and market weights

3
The correct calculations are covered in many fixed income texts. For example, see Fabozzi and Mann
(2001) and Grieves and Griffiths (2005).

4
are also close. If the bond’s yields are the same or they are similar, the weights in a

weighted average do not matter very much.

Does this matter? Does anyone actually get it wrong? We have chosen eight

large US companies that disclose the average cost of their outstanding debt. They all

get it wrong.

Table 1 presents our eight companies, sorted by credit rating. We display the

market value weighted average interest rate and the duration-dollar average interest

rate. The difference between the measures ranges from 47bp to 140bp. Getting the

interest rate wrong by 1.4% on a 5.45% level is material.4

[Table 1 about here.]

Table 2 displays the mapping of major auditing firms to our selected

companies. All are represented.

[Table 2 about here.]

3. How Did this Appalling Circumstance Arise?

There are two culprits in this matter, the accounting profession and finance

professors. The accounting profession gets blame because there is no standard for

reporting average interest, leaving the calculation methodology to companies to

decide. And lemming-like, they have chosen a mistaken calculation.

Finance professors are to blame because no introductory finance text and no

intermediate corporate finance text include the correct calculation. This is all the

more surprising because when they teach fixed income, finance professor teach proper

technique for portfolio yield to maturity.

Table 3 lists the introductory and intermediate finance texts that we have

examined in search for the correct calculation. The flaw almost always comes down
4
Table 4 in Appendix A contains two examples of the detailed calculations presented in Table 1.

5
to assuming a single debt issue or simply announcing an average interest rate on a

company’s debt, without explaining how one should calculate that value.

[Figure 3 about here.]

If a corporate CFO had taken only corporate finance courses, he or she would

fall in line with current practice without hesitation. If a bond analyst looked at a

portfolio containing only one company’s bonds, he or she would calculate a different

average yield from the CFO on exactly the same set of bonds. That cannot be right.

4. Consequences

Does the error matter? As in so many cases, it depends. If the yield curve

happens to be flat, which occurs from time to time, then average interest on a

company’s debt would not depend on which weighting scheme is used. If companies

have several-to-many bond issues ‘laddered’ along the yield curve, the average

interest rate from the competing calculation methods would be quite similar. For

example, if a company’s outstanding debt comprised four bond issues with equal

market value and maturities 5-years, 10-years, 20-years, 30-years, their reported

average interest rate would be 6.04% with a yield curve that slopes linearly in

maturity from 4.5% to 8%. The duration dollar weighted average would be 6.43%.

The gap between market weights and duration-dollar weights is much smaller than for

the barbell we considered above.

If companies have no long term bank loans, then analysts need consider only

their traded debt, which is observable. In this case, the aggregation rule that the

company uses for its financial reports is not misleading. Investors can undertake the

correct calculation for themselves. In fact, CFA charter holders, who populate buy

side firms heavily, are taught the correct analysis, and they can choose the right

measure for their valuation estimations.

6
 When the company has a mix of bank loans and traded debt, analysts can

make an educated guess at the company’s average interest rate, but standard

disclosures come up short on the required information.

The main difficulty with reporting the wrong average interest rate on debt is

that companies may be deluding themselves on how much their debt costs, leading to

a mis-estimation of weighted average cost of capital and faulty investment decisions.

What does it take for the average interest rate calculation to matter? First, the

yield curve must be steep, as it is now. Second, the company’s debt must be a

‘barbell.’ For example, they have a 2-year bond outstanding and a 30-year bond

outstanding. Finally, debt must be a significant portion of the company’s capital

structure. In that case, companies can end up with the wrong WACC and the wrong

investment level or mix.

To be sure, companies’ measures and application of WACC most nearly

resembles, “measure with a micrometer, mark it with chalk and cut it with an axe.”

Still, there is no reason to use a broken micrometer willingly.

Even if their funding were barbells, some companies would still use the right

measure for average interest internally, while reporting the flawed measure. Financial

firms, in particular, have plenty of expertise in measuring portfolio yield, irrespective

of whether the positions are long (assets) or short (debt). Typically, they use the

correct calculation on both sides, consistently.

5. Conclusion

Current reporting practice nearly certainly has all companies reporting a mistaken

average yield on their debt. Analysts can undo the error when evaluation the

company, but the companies may be using the wrong interest rate in deciding on

capital budgeting decisions.

7
 References

Ambrose, Brent and Arthur Warga, 1996, Yield Bogeys, Financial Analysts Journal,
63-68.

Bodie, Zvi, Alex Kane and Alan J. Marcus, 2005, Investments, 6th Edition, McGraw-
Hill Irwin, New York, NY 10020.

Brealey, Richard A., Steward C. Myers, and Franklin Allen, 2006, Principles of
Corporate Finance, McGraw-Hill Irwin, New York, NY 10020.

Brigham, Eugene F., and Joel F. Houston, 2007, Fundamentals of Financial

Management, Concise Fifth Edition, South-Western College Publishing,
Cincinnati, OH.

Fama, Eugene, and Kenneth French, 1988, Permanent and Temporary Components of
Stock Prices, Journal of Political Economy 96, 246–273.

Fabozzi, Frank J., and Steven V. Mann, 2001, Introduction to Fixed Income
Analytics, Frank J. Fabozzi Associates, New Hope, PA.

Garbade, Kenneth D., 1996, Fixed Income Analytics, MIT Press, Boston MA.

Grieves, Robin and Mark D. Griffiths, 2005, A Wall Street Internship: Introduction to
Fixed Income Analytics, Thomson/South-Western, Mason, OH

Macaulay, F. R., 1938, Some Theoretical Problems suggested by The Movements of

Interest Rates, Bond Yields and Stock Prices in the United States Since 1856,
National Bureau of Economic Research, Inc., 1819 Broadway, New York,
New York.

McKinsey & Company Inc., Tom Copeland, Tim Koller and Jack Murrin, 2000,
McKinsey DCF Valuation 2000 Model (to accompany Valuation: Measuring
and Managing the Value of Companies, Third Edition) (CD-ROM), John
Wiley & Sons, New York, NY.

Moyer, R. Charles, James R. McGuigan and William J. Kretlow, 2001, Contemporary

Financial Management, Eighth Edition, South-Western College Publishing,
Cincinnati, OH.

Ross, Stephen A., Randolph W. Westerfield, and Bradford D. Jordan, 2006,

Fundamentals of Corporate Finance, (Seventh Edition), McGraw-Hill Irwin,
New York, NY 10020.

Stewart III, G. Bennett, 1999, The Quest for Value: A Guide for Senior Managers,
Harper Collins Publishers Inc, New York, NY 10020.

8
 Table 1

Market Traditional Dollar-Duration

Value Average Average
Name Rating of Debt Interest Interest Difference
($M) (bp)

Sysco Corp AA- 1238 3.621% 4.909% 128.8

The Chubb Company A 1841 4.544% 5.163% 61.9

H.J. Heinz Company A 3912 4.507% 5.165% 65.8

VF Corp A- 1002 4.028% 5.203% 117.5

Nordstrom Inc A- 1046 4.050% 5.452% 140.2
Baker Hughes International A- 1292 4.875% 5.342% 46.7

Electronic Data Systems BBB 3011 5.563% 6.500% 93.7

Motorola Inc BBB 7710 4.898% 5.690% 79.2

Table 2
Auditing Firms for Selected Companies

Baker Huges International Deloitte

Nordstrom, Inc. Deloitte
The Chubb Company Ernst and Young
Sysco Systems Ernst and Young
Electronic Data Systems KPMG
Motorola, Inc KPMG
H.J.Heinz Company PWC
VF Corp PWC

9
 Table 3
Survey of major academic texts 10/29/12
None of which calculate average yield correctly

Brigham, Houston Fundamentals of Financial Management, 7e

Damodaran Applied Corporate Finance, 3e
Ross, Westerfield, Jaffee Corporate Finance, 7e
Brealey, Myers Principles of Corporate Finance, 7e
Berk, DeMarzo Corporate Finance
Grinblatt, Titman Financial Markets and Corporate Strategy
Ross, Westerfield, Jordan Fundamentals of Corporate Finance, 7e
Pratt Cost of Capital, 2e
Eun, Resnick International Financial Management, 6e
Benninga Principles of Finance with Excel

10
 Appendix A

Table 4 Panel A
SYY: Sysco Corp
AA- Bonds

Yield-Traditional SYY Bonds

Face Mkt
Coupon ($M) ($M) wt K wk
6/15/2005 0.06500 150 160 0.1292 0.01895 0.002449
7/30/2005 0.04750 200 213 0.1721 0.0197 0.003389
5/1/2006 0.07000 200 223 0.1801 0.02509 0.004519
4/15/2007 0.07250 100 115 0.0929 0.02935 0.002726
6/1/2012 0.06100 200 219 0.1769 0.04811 0.008511
4/15/2027 0.07160 50 60 0.0485 0.05663 0.002745
8/1/2028 0.06500 225 248 0.2003 0.05924 0.011867
1,238 1.0000 3.621%

WACC-
Dollar Duration SYY Bonds t=0.4
Face Mkt Dollar
Coupon ($M) duration ($M) duration wt K wk
6/15/2005 0.06500 150 1.40 160 224 0.0343 0.01895 0.00065
7/30/2005 0.04750 200 1.51 213 321.63 0.0492 0.0197 0.00097
5/1/2006 0.07000 200 2.16 223 481.68 0.0738 0.02509 0.001851
4/15/2007 0.07250 100 2.95 115 339.25 0.0519 0.02935 0.001525
6/1/2012 0.06100 200 6.77 219 1482.63 0.2270 0.04811 0.010922
4/15/2027 0.07160 50 9.07 60 544.2 0.0833 0.05663 0.004719
8/1/2028 0.06500 225 12.65 248 3137.2 0.4804 0.05924 0.028458
1,238 6530.59 1.0000 4.909%

11
 Table 4 Panel B
CB: The Chubb Company
A Bonds

Yield-Traditional CB Bonds
Face Mkt
Coupon ($M) ($M) wt K wk
8/15/2005 0.06150 300 327 0.1776 0.02019 0.003586
12/15/2007 0.07125 75 85 0.0462 0.03551 0.00164
4/1/2008 0.03950 225 229 0.1244 0.0372 0.004627
11/15/2011 0.06000 400 434 0.2357 0.04811 0.011342
4/1/2013 0.05200 275 283 0.1537 0.0496 0.007625
8/15/2018 0.06600 100 112 0.0608 0.05636 0.003429
2/1/2027* 0.08675 125 151 0.0820 0.07161 0.005873
11/15/2031 0.06800 200 220 0.1195 0.06124 0.007318
1,841 1.0000 4.544%

WACC-
Dollar Duration CB Bonds
Face Mkt Dollar
Coupon ($M) duration ($M) duration wt K wk
8/15/2005 0.06150 300 1.53 327 500.31 0.0443 0.02019 0.000895
12/15/2007 0.07125 75 3.51 85 298.35 0.0264 0.03551 0.000939
4/1/2008 0.03950 225 3.90 229 893.1 0.0792 0.0372 0.002945
11/15/2011 0.06000 400 6.40 434 2777.6 0.2462 0.04811 0.011844
4/1/2013 0.05200 275 7.42 283 2099.86 0.1861 0.0496 0.009231
8/15/2018 0.06600 100 9.59 112 1074.08 0.0952 0.05636 0.005365
2/1/2027* 0.08675 125 4.74 151 715.74 0.0634 0.07161 0.004543
11/15/2031 0.06800 200 13.29 220 2923.8 0.2591 0.06124 0.01587
1,841 11282.84 1.0000 5.163%
* rated BBB+

12