chapter 7 to 9

Ratings and the Risk Structure of Interest Rates Chapter 7 I 163

These examples highlight the need to understand the differences among the many types of bonds that are sold and traded in financial markets. What was it about the movement in the prices of different bonds that Mr. McDonough found so informative? How did information about profitability affect investors' willingness to lend to GM and Ford? To answer these questions, we will study the differences among the multitude of bonds issued by governments and private corporations. As we will see, these bonds differ in two crucial respects: the identity of the issuer and the time to maturity. The purpose of this chapter is to examine how each of these affects the price of a bond, and then to use our knowledge to interpret fluctuations in a broad variety of bond prices.

Default is one of the most important risks a bondholder faces. Not surprisingly, the risk that an issuer will fail to make a bond's promised payments vaties substantially from one bonower to another. The risk of default is so important to potential investors that independent companies have come into existence to evaluate the creditworthiness of potential borrowers. These firms, sometimes called rating agencies, estimate the likeli-hood that a corporate or government borrower will make a bond's promised payments. The first such ratings began in the United States more than 100 years ago. Since 1975, the U.S. government has acknowledged a few firms as "nationally recognized statisti-cal rating organizations" (NRSROs), a designation that has encouraged investors and governments worldwide to rely on their ratings. In 2010, sweeping financial reform legislation (called the Dodd-Frank Wall Street Reform and Consumer Protection Act) included provisions to reduce reliance on ratings agencies. Let's look at these compa-nies and the information that they produce.

The best-known bond rating services are Moody's and Standard & Poor's? These companies monitor the status of individual bond issuers and assess the likelihood that a lender/bondholder will be repaid by a borrower/bond issuer. Companies with good credit-those with low levels of debt, high profitability, and sizable

STANDARD ---

&POOR'S ---

amounts of cash assets--earn high bond A high rating suggests that a SOURCE: Courtesy of Standard & Poor's.

bond issuer will have little problem meeting a bond's payment obligations. Table 7.1 reports the rating systems of Moody's and Standard & Poor's. As you can

see, they are very similar. Both systems are based on letters and bear a broad similarity to the rankings in minor-league baseball. Firms or governments with an exceptionally strong financial position carry the highest ratings and are able to issue the highest-rated bonds, Triple A. ExxonMobil, Microsoft, and the government of Canada are all examples of entities with Aaa bond ratings.3

2Fi tch is a third, less well-known bond rating company. In addition to the big three, seven small agencies also now enjoy the U.S. government's NRSRO designation. 'U.S. government debt is a standard of comparison for other dollar-denominated debt because it is widely perceived as among the world's safest financial instruments. In August 2011, however, reacting to the rising debt burden and what it called "political risks," Standard and Poor's downgraded long-term U.S. Treasury debt to AA +.In September 2012, Moody's warned that it may downgrade U.S. debt from Aaa to Aa1 (the highest Aa category) if U.S. budget negotiations in 2013 failed to put the ratio of federal debt to GDP on a declining path in the medium term.

164 I Chapter 7 The Risk and Term Structure of Interest Rates

A Guide to Bond Ratings

Standard Examples of Issuers with i Moody's & Poor's Description Bonds Outstanding in 2012 ! Investment Grade Aaa AAA Bonds of the best quality with the smallest Exxon Mobil

risk of default. Issuers are exceptionally Microsoft stable and dependable. Canada

..

A a AA Highest quality with slightly higher degree General Electric

of long-term risk. Procter & Gamble China

A A High-medium quality, with many strong JPMorgan Chase attributes but somewhat vulnerable to Oracle changing economic conditions. Israel

Baa BBB Medium quality, currently adequate but General Mills perhaps unreliable over the long term. Time Warner

Brazil Italy

Noninvestment, Ba BB Some speculative element, with moderate Goodyear Tire Speculative security but not well safeguarded. General Motors Grade Turkey

B B Able to pay now but at risk of default in the Hertz future. Office Depot

Venezuela

Highly Caa CCC Poor quality, clear danger of default. Cuba* Speculative Pakistan*

Ca cc Highly speculative quality, often in default. c c Lowest-rated, poor prospects of repayment Greece**

though may still be paying. D D In default.

,.

*S&P rated Pakistan B and did not rate Cuba. **S&P rated Greece CCC. For a more detailed definition of ratings see Moody's website, www.moodys.com, or Standard and Poor's website, www.standardandpoors.com.

The top four categories, Aaa down to Baa in the Moody's scheme, are considered meaning they have a very low risk of default. These rat-

ings are reserved for most government issuers as well as corporations that are among the most financially sound.4 Famously, these top ratings also were awarded to many risky mortgage-backed securities (see Lessons from the Crisis: Subprime Mortgages on page 165) that later plunged in value, triggering the financial crisis of 2007-2009 (see Lessons from the Crisis: Rating Agencies on page 166). Table 7.1 gives examples of governments and firms in selected rating levels above default in the autumn of 2012.

The distinction between investment-grade and speculative, noninvestment-grade bonds is an important one. A number of regulated institutional investors, among them some insurance companies, pension funds, and commercial banks, are not allowed to

4Government debt ratings are important, as they generally create a ceiling on the ratings for private companies in that country.

Ratings and the Risk Structure of Interest Rates Chapter 7 165

~ SUBPRIME MORTGAGES The financial crisis of 2007-2009 initially was known as the subprime crisis because of the mortgage-backed securities (MBS, see page 53) that helped trigger the crisis in the United States. What is a subprime mortgage? What role does it play in the mortgage market? And how did it contrib-ute to the financial crisis?

A residential mortgage is called subprime when it does not meet key standards of creditworthiness that apply to a conventional prime mortgage.* Conventional prime mort-gages are those that satisfy the rules for inclusion in a col-lection or pool of mortgages to be guaranteed by a U.S. government agency. For this reason, conventional prime mortgages also are called qualifying or conforming mort-gages.t Prime mortgages in which the loan amount exceeds a government-specified limit are called jumbo prime and do not qualify for the government guarantee.

The purpose of the creditworthiness standards is to in-crease the likelihood that the borrower will be able to repay the loan. Reducing borrower defaults lowers the cost to the government agency of its guarantee for the mortgage pools that back the MBS. The lending standards for qualifying mortgages typically include rules about the borrower's in-come, wealth, and credit score (see Your Financial World: Your Credit Rating on page 169). The standards also cover the size of mortgage, the price of the home, and the ratio between those two amounts-known as the Joan-to-value ratio, or LTV ratio.

Subprime loans may fail to meet some or all of these standards for a qualifying mortgage. Typically, a loan is sub-prime if it is made to someone with a low credit score or whose income may be low relative to the price of the home; or if the LTV is high; or if the borrower does not provide suffi-cient documentation of their ability to pay. All of these factors make the loan more risky. Put differently, a subprime loan has a higher probability of default than qualifying loans.

Like conventional mortgages, the subprime type comes in two forms: fixed-rate and adjustable-rate mortgages (the latter are called ARMs-Chapter 6). And like conventional ARMs, subprime ARMs typically provide a low interest

rate-known as a teaser rate-for the first two or three years. But once this initial period ends, the interest rate re-sets to a higher level-often, a much higher level-for the remaining term of the loan. This structure gives borrowers an incentive to replace their mortgage after the early years to obtain a new, low teaser rate or a fixed-rate loan. This process is called refinancing. Rising house prices make it possible to refinance even when a borrower's ability to pay is low because the lender can sell the house to recover the loan if the borrower fails to make timely payments.

During the housing bubble, starting in 2002, the volume of subprime loans surged as mortgage lenders relaxed their lending standards. Borrowers could obtain loans with lower down payments (high LTVs) and ever-poorer documenta-tion. A complacent belief that the rise in nationwide house prices that had continued since the 1930s would persist indefinitely encouraged lending to borrowers with progres-sively lower ability to pay. When house prices started to fall in 2006, home values began to sink "below water"-that is, the home price fell to less than the amount of the mortgage-and lenders became unwilling to refinance many subprime loans. A wave of defaults followed. And further house price declines meant that lenders would not be able to recover the loan amounts in the increasingly likely event that borrowers would not be able to pay.

Even at their peak, outstanding subprime mortgages probably accounted for less than 15 percent of overall resi-dential mortgages (which were $11 trillion in 2007)-and only a fraction of these were of really terrible quality. So why did subprime mortgage defaults trigger the financial disruptions of 2007-2009? The key reason is that some large, highly lev-eraged financial institutions held a sizable volume of MBS backed by subprime mortgages. Buying the MBS had al-lowed the institutions to increase their leverage and risk-tak-ing (see Chapter 5) at the same time that they earned fees for new MBS issuance. In effect, some financial institutions "bet the house" on subprime mortgage securities, and the price collapse of those securities threatened their existence.

*There is another category of nonprime mortgages called alt-A that fall between prime and subprime in their default probability. 1The guarantee on a conforming mortgage pool lowers the default risk on a security backed by the pool. As a result, investors will pay more for the secmity and borrowers can obtain a cheaper mortgage.

invest in bonds that are rated below investment grade, that is, below Baa on Moody's scale or BBB on Standard & Poor's scale.5

Bonds issued by Goodyear, Hertz, and the governments of Turkey and Venezuela are in the noninvestment, speculative grade. These companies and countries may have difficulty meeting their bond payments but are not at risk of immediate defau_lt. T~e final category in Table 7.1, highly speculative bonds, consists of debts that are m sen-ous risk of default.

'Restrictions on the investments of financial intennediaries, such as insurance companies, are a_ matter for government regulators. There is no comprehensive reference for all of the legal restrictions that force fi~a~cml firms to. sell bonds whose ratings fall below Baa. In many cases, such as those of bond mutual funds, the restnctions are self-Imposed.


b. RATING AGENCIES Rating mistakes contributed significantly to the financial cri-sis of 2007-2009. So, as governments sought to reform the financial system in the wake of the crisis, the role of rating agencies naturally attracted great attention.

Investors and regulators around the world relied on the high ratings that agencies had awarded to a large group of mortgage-backed securities (MBS, see page 53, and discussed in Chapter 6). When U.S. housing prices began to fall and MBS prices started to decline, rating agencies began a series of sharp downgrades, sometimes lowering ratings by four notches or more from the highest level (such as Moody's Aaa) directly to speculative grade (below Baa). Such downgrades reinforced the plunge in MBS prices. Losses on the highly rated assets, which were widely held by banks, diminished resources in the financial system suf-ficiently to help trigger (and later to intensify) the crisis.

What caused the rating errors? Agencies evaluating mortgage debt typically estimate default risks using mathe-matical models that rely on information about the borrowers, the houses they purchase, and historical patterns of default. However, the data used did not include a period in which U.S. housing prices fell significantly nationwide, because such a thing had not happened since the 1930s, during the Great Depression. As a result, the models severely underpredicted

the likelihood of default on individual mortgages originated during the housing bubble. When defaults began to surge beyond what their models anticipated, the rating agencies were compelled to reassess their ratings.

Other factors may have also contributed to rating errors. For example, in developing some MBS, issuers typically consulted rating agencies to determine which bond structure would lead to the highest rating (and, presumably, the high-est price). Payments for such consultation have the potential to create a conflict of interest when agencies are later called on to give the issue a rating in an objective, independent fashion. Some smaller rating agencies did not offer such consultations, and governments have already moved to pre-vent such potential conflicts from arising in the future.

Conflicts also can arise from the way rating agencies are compensated. Since the 1970s, the largest U.S. rating agencies have obtained their compensation from issuers (the issuer pays model). Some observers argue that such compensation produces an incentive for a favorable rating (ratings inflation). Accordingly, one proposed legislative re-form would require issuers to pay an independent authority that selects ratings firms to do their work free of influence from the issuer.

Another concern was that the agencies used a single rating scale to represent default probabilities regardless of the liquidity, transparency, and degree of standardization of an issue-a practice that may have led investors to underes-timate other important risks.

Bonds with ratings below investment grade are often refen-ed to as bonds or sometimes more politely as high-yield bonds (a reminder that to obtain a high yield, investors must take a large risk). 6 There are two types of junk bonds. The first type, called were once investment-grade bonds, but their issuers fell on hard times. The second are cases in which little is known about the risk of the issuer.

MCI WorldCom, the telecommunications giant purchased by Veri zan in early 2006, was one company whose bond rating fluctuated between investment grade and junk. When it began issuing bonds in 1997, the firm was below investment grade (Moody's Ba). MCI WorldCom saw its rating rise for several years, until it peaked as a Moody's A from mid-1999 to the end of2001. Taking advantage of this investment-grade rating, MCI World Com issued $11.8 billion worth of bonds in May 2001. Just one year later, MCI WorldCom's rating dropped back to where it started, Ba, and its 10-year bonds were trading for 44 cents on the dollar, less than half of their initial prices. By mid-2002, as the company filed for bankruptcy, its bonds had fallen one more notch to B.

Sovereigns can be "fallen angels," too. For example, bonds issued by Greece were rated A (investment grade) as recently as the spring of2010. As the financial crisis in the euro area intensified, the rating agencies repeatedly lowered Greece's credit rating (even-tually to the lowest grade, C) in the run-up to its March 2012 default. Frequent, sharp downgrades during such periods of market stress raise doubts about the usefulness of the ratings in anticipating default (see also Lessons from the Crisis: Rating Agencies above).

6Junk bond is an informal term used to mean a highly speculative security that has a low rating; it has no exact or formal definition.

Ratings and the Risli Structure of Interest Rates Chapter 7 I 167

Indeed, for investment-grade sovereign debt, Moody's puts the five-year default rate, weighted by issuer, at nearly 3 percent. The one-year-ahead record is better: 12 months before their default, no sovereign issuer was still rated as investment grade, but more than 1 percent of corporate defaults were.7

Material changes in a firm's or government's financial conditions precipitate changes in its debt ratings. The rating services are constantly monitoring events and announcing modifications to their views on the creditworthiness of borrowers. If a particular business or country encounters problems (as occurs with some frequency), Moody's and Standard & Poor's will lower that issuer's bond rating in what is called a Typically, an average of nearly 2 percent of bonds that begin a year in an investment-grade category-Aaa to Baa-have their ratings downgraded to one of the noninvestment grades. The MCI WorldCom downgrade in May 2002 re-flected the agencies' view that the company had too much debt and (given the dismal state of the telecommunications industry at the time) little opportunity to reduce it.

occur as well. Roughly 3 percent of Aa-rated bonds are upgraded to Aaa each year. 8

Commercial Paper Ratings paper is a short-term version of a bond. Both corporations and govern-

ments issue commercial paper. Because the borrower offers no collateral, this form of debt is unsecured. So only the most creditworthy companies can issue it.9 Moreover, the amount of commercial paper outstanding plunged dming the financial crisis of 2007-2009: After peaking in mid-2007 at $2.1 trillion, it dropped below $1 trillion by 2011. Nevertheless, financial companies, such as Bank of America and Goldman Sachs, still issued the vast majority of it.

Like a U.S. Treasury bill, commercial paper is issued on a discount basis, as a zero-coupon bond that specifies a single future payment with no associated coupon payments. For legal reasons, commercial paper usually has a maturity of less than 270 days. 10 More than one-third of all commercial paper is held by money-market mutual funds (MMMFs), which require ve1y short-term assets with immediate liquidity. Most commercial paper is issued with a maturity of 5 to 45 days and is used exclusively for short-term financing.

The rating agencies rate the creditwmthiness of commercial paper issuers in the same way as bond issuers. Again, Moody's and Standard & Poor's have parallel rating schemes that differ solely in their labeling (see Table 7 .2). By some estimates, 90 percent of issuers carry Moody's P-1 rating and another 9 percent are rated P-2-the P stands for Speculative-grade commercial paper does exist, but not because it was originally issued as such.

7For an in-depth assessment of sovereign credit ratings, see International Monetary Fund, Global Financial Stability Report, "The Uses and Abuses of Sovereign Credit Ratings" (October 2010, Chapter 3). Moody's issuer-weighted cumulative default rate estimates appear in Exhibit 10 of its report "Sovereign Default and Recovery Rates, 1983-2012," June 12,2013. 'Based on authors' estimates for the period 1983 to 2012 using Moody's Annual Default Study: Co17Jorate Default and Recovel)' Rates, 1920-2012, February 28, 2013, Exhibits 29 and 41. 'Recall that collateral is something of value pledged by the borrower that the lender could sell if the loan is not repaid. See also Lessons from the Crisis: Asset-Backed Commercial Paper on page 171. 10As described in detail by Thomas K. Hahn, "Commercial Paper," Instruments of the Money Market (Federal Reserve Bank of Richmond, 1998, Chapter 9), the Securities Act of 1933 generally requires registration of securities, a time-consuming and expensive process. But Section 3(a) (3) of the act exempts securities with less than 270 days to maturity as long as they meet certain requirements.


Commercial Paper Ratings

Examples of Issuers Standard with Commercial Paper

Moody's & Poor's Description Outstanding in 2012

Investment P-1 A-1+, A-1 Strong likelihood of timely Coca-Cola or Prime repayment. General Electric Grade Procter & Gamble

P-2 A-2 Satisfactory degree of General Mills safety for timely repayment. Time Warner

P-3 A-3 Adequate degree of safety Alcoa for timely repayment.

i Speculative, B,C Capacity for repayment below Prime is small relative to higher-Grade rated issuers. Defaulted D

SOURCE: Thomas K. Holm, "Commercial Pape1;" Instruments of the Money Market, Chapter 9, Federal ReseJ1'e Bank of Richmond, 1998; li'W\l'.JJJoodys.com; and lFWW.standardandpoors.com.

The of on Yields

..

~;

Bond ratings are designed to reflect default risk: The lower the rating, the higher the risk of default. We know investors require compensation for risk, so everything else held equal, the lower a bond's rating, the lower its price and the higher its yield. From Chapter 6 we know that we can think about changes in risk as shifts in the demand for bonds. Increases in 1isk will reduce investor demand for bonds at every price, shifting the demand curve to the left, decreasing the equilibrium price and increasing the yield (see Figure 7.1).

The easiest way to understand the quantitative impact of ratings on bond yields is to compare different bonds that are identical in every way except for the issuer's credit rating. U.S. Treasury issues have long served as a standard for compmison because they are viewed as having little default risk. This is why they are commonly referred to as

and the yields on other bonds are measured in terms of the (Remember from the definition in Chapter 5: Risk is measured relative

to a benchmark. For bonds, the most common benchmark is U.S. Treasury bonds.) We can think of any bond yield as the sum of two parts: the yield on the benchmark

U.S. Treasury bond plus a default-risk premium, sometimes called a

Bond yield = U.S. Treasury yield + Default risk premium (1) If bond ratings properly reflect the probability of default, then the lower the rating of the issuer, the higher the default-risk premium in equation (1). This way of thinking about bond yields provides us with a second insight: When Treasury yields move, all other yields move with them.

These two predictions-that interest rates on a variety of bonds will move together and that lower-rated bonds will have higher yields-are both borne out by the data. To see this, let's look at a plot of the Panel A of Figure 7.2 (on page 170) shows the yield to maturity for long-term bonds with three different

Ratings and the Risk Structure of Interest Rates Chapter 7 I 169

Companies aren't the only ones with credit ratings; you have one, too. Have you ever wondered how someone decides whether to give you a loan or a credit card? The answer is that there are companies keeping track of your financial information. They rate your creditworthiness, and they know more about you than you might think. Credit-rating companies know all about your credit cards, your car loan or mortgage (if you have one), and whether you pay your bills on time. All of this information is combined in something called a credit score. If you have low levels of debt and pay your bills on time, you have a high credit score.

might have an interest rate of 5 percent and monthly pay-ments of $230. But if your credit score was low, because you missed a payment on a credit card or paid your utility bill late, then the interest rate could be as high as 15 per-cent, which would mean monthly payments as much as $50 higher. The same principle applies to home mortgages; the better your credit score, the lower the interest rate. It pays to pay all your bills on time.*

You care about your credit score; here's why. Lenders use credit scores to calculate the interest rate they charge on a loan. With a top credit score, a four-year, $10,000 car loan

*Ironically, someone who has never had a credit card and never owed anyone any money has no credit history at all and so will have a low credit score. You cannot start too soon in creating a record as a good credit risk. And you are entitled to a free annual credit report from each of the credit-rating companies. To find out how to get it, go to

ratings: 10-yea:r U.S. Treasury, Moody's Aaa rated, and Moody's Baa long-term bonds. As you can see from the figure, all of these yields move together. When the U.S. Trea-sury yield goes up or down, theAaa and Baa yields do too. While the default-risk premi-ums do fluctuate-rising patticularly in periods of financial stress--changes in the U.S. Treasury yield account for most of the movement in the Aaa and Baa bond yields. Further-more, the yield on the higher-rated U.S. Treasury bond is con-sistently the lowest. In fact, over the years from 1971 to 2012, the 10-year U.S. Treasury bond yield averaged more than a full percentage point below the yield on Aaa bonds and two percentage points below the yield on Baa bonds.

How important is one or two percentage points in yield? To see, we can do a simple computation. At an interest rate of 5 percent, the present value of a $100 payment made 10 years from now is $61.39. If the interest rate rose to 7 per-cent, the value of this same promise would decline to $50.83. So a two-percentage point increase in the yield, from 5 per-cent to 7 percent, lowers the value of the promise of $100 in 10 years by $10.56, or 17 percent!

From the viewpoint of the bonower, an increase in the interest rate from 5 percent to 7 percent means paying $7 rather than $5 per year for each $100 borrowed. That is a 40 percent difference. Cleatly, ratings are cmcial to corpora-tions' ability to raise financing. Whenever a company's bond rating declines, the cost of funds goes up, impairing the com-pany's ability to finance new ventures. 11

"'g 0 ~

The Effect of an Increase in Risk on Equilibrium in the Bond Market

~ Pol---.---~---cX 0. pl 1---~----x 11) u

~ Quantity of Bonds

Increased risk reduces the demand for the bond at every price, shifting the demand curve to the left from D0 to D,. The result is a decline in the equilibrium price and quantity in the market. Importantly, the price falls from P 0 to P,, so the yield on the bond must rise.

11The same is true for individuals. Consider the impact on the monthly payments required to service a thirty-year, $100,000 mortgage. At an interest rate of 5 percent, payments would be approximately $530 per month. If the interest rate were to increase to 7 percent, the required monthly payments would rise to more than $650. You can compute these amounts using the formulas in the Appendix to Chapter 4.


The Risk Structure of Interest Rates

A. Comparing Long-Term Interest Rates 18 16 14

'"0 12 v 10 :::::: 8

6 4 2 OL---~--~---L--~----L---~---L--~----L---~---L--~----L-~ 1971 1974 1977 1980 1983 1986 1989 1992 1995 1998 2001 2004 2007 2010 2013

~ 10-Year Treasury "~""Moody's Aaa -Moody's Baa

B. Comparing Short-Term Interest Rates 18 16 14

'"0 12 v 10 :::::: 8

6 A 4 2

OL_--~--~---L--~----L_--~---L--~----L---~---L--~--~--~ 1971 1974 1977 1980 1983 1986 1989 1992 1995 1998 2001 2004 2007 2010 2013

j ~~ Treasury Bills """"" Commercial Paper j

SOURCE: Board of Governors of the Federal Rese~w System. (FRED data codes: GSJO, AAA, BAA, CPNJM [since 1997], and TB3MS).

What is true for long-term bond yields is true for short-term bond yields; they move together, and lower ratings imply higher yields. Compare the yields on three-month U.S. Treasury bills with those on A-l/P-1 commercial paper of the same maturity (see Panel B of Figure 7.2). The two yields clearly move together, and the U.S. Treasury bill yield is always lower than the yield on commercial paper. From 1971 to 2012, the spread of commercial paper over U.S. Treasury bills aver-aged nearly six-tenths of one percentage point, or roughly 60 basis points. (Recall from Chapter 6 that a basis point is one hundredth of a percentage point, or 0.01 percent.)

The lesson is clear; investors must be compensated for assuming risk. The less creditworthy the borrower, the higher the risk of default, the lower the borrower's rating, and the higher the cost of borrowing. And the lower the rating of the bond or conm1ercial paper, the higher the yield.

Differences in Tax Status and Municipal Bonds Chapter 7 I 171

~ ASSET-BACKED COMMERCIAl PAPER Asset-backed commercial paper (ABCP) is a short-term li-ability with a maturity of up to 270 days. Unlike most com-mercial paper, which is unsecured, ABCP is collateralized by assets that financial institutions place in a special portfolio. As we saw in Chapter 3, collateral is something of value that is pledged to pay a loan (in this case, CP) in the event that the borrower does not make the required payments. ABCP has existed for decades, but it played a special role in the housing boom that preceded the financial crisis of 2007-2009.

To lower their costs and limit their own asset holding, some large banks created firms (a form of shadow bank) that issued ABCP and used the money to buy mortgages and other loans (see Chapter 3, Lessons from the Crisis: Shadow Banks). The payment stream generated by the loans was used to compensate the holders of the ABCP. This off-balance-sheet financing also allowed banks to boost le-verage and take more risk. When mortgage volume surged in the housing bubble, these shadow banks issued more ABCP to finance their rapid expansion.

The mismatch between the long-term maturity of the as-sets (mortgages) and the short-term maturity of the liabilities

(ABCP) posed an underappreciated threat to the ABCP is-suers. When the ABCP matures, the issuers have to borrow (or to sell the underlying assets) to be able to return the principal to the ABCP holders. The risk was that the issuers would be unable to borrow-that is, they faced rollover risk. If they were also unable to sell the long-term assets easily, these shadow banks would face failure.

ABCP rollover risk played an early role in the financial crisis of 2007-2009. When the value of some mortgages became highly uncertain in 2007, purchasers of ABCP grew anxious that the assets backing their commercial paper might plunge in value. Because they had erroneously viewed ABCP as a very low risk security, the sudden awareness of risk caused a virtual halt to ABCP purchases. Outstanding commercial paper from ABCP issuers began to plunge in the third quarter of 2007, sinking from a peak of more than $900 billion to less than $100 billion by 2012.

Firms that had issued ABCP faced an immediate threat to their survival. Unable to sell their assets or to obtain other funding, some failed. In other cases, banks chose to res-cue their shadow banks to limit legal risks and reputational damage. As a result, those banks faced heightened liquidity needs and pressures to sell assets precisely when the cost of funds had surged and asset prices were plunging. The risk that they had sought to shift off balance sheet returned at the worst possible time-in the midst of a crisis.

Default risk is not the only factor that affects the return on a bond. The second important factor is taxes. Bondholders must pay income tax on the interest income they receive from owning privately issued bonds. These are bonds. In contrast, the coupon payments on bonds issued by state and local governments, called or

are specifically exempt from taxation. 12 The general rule in the United States is that the interest income from bonds issued

by one government is not taxed by another government, although the issuing govern-ment may tax it. The interest income from U.S. Treasury securities is taxed by the fed-eral government, which issued them, but not by state or local governments. In the same way, the federal government is precluded from taxing interest on municipal bonds. In an effort to make their bonds even more attractive to investors, however, state and local governments usually choose not to tax the interest on their own bonds, exempting it from all income taxes.

How does a tax exemption affect a bond's yield? Bondholders care about the return they actually receive, after tax authorities have taken their cut. If investors expect to receive a payment of $6 for holding a bond but know they will lose $1.80 of it in taxes, they will act as if the return on the investment is only $4.20. That is, investors base their decisions on the after-tax yield.

12Municipal bonds come in two varieties. Some are general-obligation bonds backed by the taxing power of the governmental issuer. Others are revenue bonds issued to fund specific projects; these are backed by revenues from the project or operator.


Calculating the tax implications for bond yields is straightforward. Consider a one-year $100 face value taxable bond with a coupon rate of 6 percent. This is a promise to pay $106 in one year. If the bond is selling at par, at a price of $100, then the yield to maturity is 6 percent. From the point of view of the government issuers, the bond-holder receives $6 in taxable income at maturity. If the tax rate is 30 percent, the tax on that income is $1.80, so the $100 bond yields $104.20 after taxes. In other words, at a 30 percent tax rate, a 6 percent taxable bond yields the equivalent of 4.2 percent.

This same calculation works for any interest rate and any bond, which allows us to derive a relationship between the yields on taxable and tax-exempt bonds. The rnle is that the yield on a tax-exempt bond equals the taxable bond yield times one minus the tax rate:

Tax-exempt bond yield= (Taxable bond yield) X (1 -Tax rate) (2) For an investor with a 30 percent tax rate, then, we can compute the tax-exempt yield on a 10 percent bond by multiplying 10 percent times (1 - 0.3), or 7 percent. Overall, the higher the tax rate, the wider the gap between the yields on taxable and tax -exempt bonds.

Structure of Interest Rates A bond's tax status and rating aren't the only factors that affect its yield. In fact, bonds with the same default rate and tax status but different maturity dates usually have different yields. Why? The answer is that long-tenn bonds are like a composite of a series of short-term bonds, so their yield depends on what people expect to happen in years to come. In this section, we will develop a framework for thinking aboutfitture interest rates.

The relationship among bonds with the same risk characteristics but different ma-turities is called the of rates.

In studying the term strncture of interest rates, we will focus our attention on Trea-sury yields; see Figure 7.3. Comparing information on 3-month (the blue line) and 10-year (the green line) Treasury issues, we can draw three conclusions:

1. Interest rates of different maturities tend to move togetheJ: The bulk of the varia-tion in short- and long-term interest rates is in the same direction. That is, the blue and green lines clearly move together.

2. Yields on short-term bonds are more volatile than yields on long-term bonds. The blue line moves over a broader range than the green line.

3. Long-term yields tend to be higher than short-term yields. The green line usually, but not always, lies above the blue line.

Default risk and tax differences cannot explain these relationships. What can? We will examine two explanations, the expectations hypothesis and the liquidity premium theory.

Over the years, economists have proposed and discarded numerous theories to explain the term strncture of interest rates. We can benefit from their hard work and ignore all the ones they found wanting. The first one we will focus on, called the ~"' 1,,.~0''"

is straightforward and intuitive. If we think about yields as the sum of a risk-free interest rate and a risk premium, the expectations hypothesis focuses on the first of those elements. It begins with the observation that the risk-free interest rate can be computed, assuming there is no uncertainty about the

The Term Structure of Interest Rates Chapter 7 I 173

18 16 14

'"012 v 10 :;;::: 8

6 4 2

The Term Structure of Treasury Interest Rates

OL---~---L--~--~L_ __ L_ __ ~ __ _L __ ~ ____ L_ __ ~---L--~--~~~

1971 1974 1977 1980 1983 1986 1989 1992 1995 1998 2001 2004 2007 2010 2013

- 3-Month Treasury Bills ~ 10-Year Treasury Bonds

SOURCE: Board of Governors oftl!e Federal Reserve System. (FRED data codes: TB3MS and GSJO).

future. That is, we know not just the yield on bonds available today but the yields that will be available on bonds next year, the year after that, and so on.

To understand the implications of this statement, think about an investor who wishes to purchase a bond and hold it for two years. Because there is no uncertainty, the inves-tor knows the yield today on a bond with two years to maturity, as well as the yields on a one-year bond purchased today and on a second one-year bond purchased one year from now. Being sure about all of these, the investor will be indifferent between holding the two-year bond and holding a series of two one-year bonds. Certainty means that bonds of different maturities are pe~fect substitutes for each otha This is the essence of the expectations hypothesis.

To see how this works, assume that the current one-year interest rate is 5 percent. The expectations hypoth-esis implies that the curr-ent two-year interest rate should equal the average of 5 percent and the one-year interest rate one year in the future. If that future interest rate is 7 percent, then the curr-ent two-year interest rate will be (5 + 7)/2 = 6%.

According to the expectations hypothesis, then, when interest rates are expected to rise in the future, long-term interest rates will be higher than short-term interest rates. This means that the which plots the yield to maturity on the vertical axis and the time to maturity on the horizontal axis, will slope up. (The Wall Street Joumal's Credit Market column often includes a plot of a yield curve for U.S. Treasury issues like the one shown in Figure 7.4.) Analogously, the expectations hypothesis implies that if interest rates are expected to fall, the yield

3.5

3.0

' 2.5 B 2 2.0 8 1.5

::s ~ 1.0

0.5

The U.S. Treasury Yield Curve

0 1 3 months years +---- maturity ------+

The figure plots the yields on Treasury bills and bonds for January 31, 2013. SOURCE: Board of Governors oftl!e Federal Reserve System.

17 4 I Chapter 7 The Risk and Term Structure of Interest Rates

The Expectations Hypothesis and Expectations of Future Short-term Interest Rates

Time to Maturity

Interest rates are expected to rise.

Time to Maturity

Interest rates are expected to remain unchanged.

Titne to Maturity

Interest rates are expected to fall.

curve will slope down. And if interest rates are expected to remain unchanged, the yield curve will be fiat. (See Figure 7.5.)

If bonds of different matmities are pelfect substitutes for each other, then we can construct investment strategies that must have the same yields. Let's look at the inves-tor with a two-year hmizon. Two possible strategies are available to this investor:

A. Invest in a two-year bond and hold it to maturity. We will call the interest rate as-sociated with this investment i21 ("i" stands for the interest rate, "2" for two years, and "t" for the time period, which is today). Investing one dollar in this bond will yield (1 + i2)(1 + i2) two years later.

B. Invest in two one-year bonds, one today and a second when the first one matures. The one-year bond purchased today has an interest rate of i 11 ("1" stands for one year). The one-year bond purchased one year from now has an interest rate of i~t+l' where the "t+ 1" stands for one time period past period t, or next year. The "e," which stands for expected, indicates that this is the one-year interest rate investors expect to obtain one year ahead. Because we are assuming that the future is known, this expectation is certain to be correct. A dollar invested using this strategy will return (1 + ilt)(1 + i~ 1+ 1 ) in two years.

The expectations hypothesis tells us that investors will be indifferent between these two strategies. (Remember, the bonds are petfect substitutes for each other.) Indiffer-ence between strategies A and B means that they must have the same return, so

(3) Expanding equation (3) and taking an approximation that is very accurate, we can write the two-year interest rate as the average of the current and future expected one-year interest rates: 13

(4) 13Expanding (3) gives us 2i2, + i~, = i1, + i;,+I + (i11)(I~ 1+ 1 ). The squared term on the left-hand side and the product term on the right-hand side of this equation are small, and their difference is even smaller. Using the example of 5 percent and 7 percent for the one-year interest rates, we can see that ignoring the two product terms means ignoring ((.06)2 - (.05'.07))/2 = (0.0036 - 0.0035)/2 = 0.00005, an en-or of 0.005 percentage points.

The Expectations Hypothesis of the Term Structure:

Today Year 0 One-year interest rate

today

~ill_____..

One-year interest rate, one year ahead

e .---- 111+1~

The Term Structure oflnterest Rates Chapter 7 I 175

Year 2

Year One-year interest rate, 3

two years ahead e .---- 111+2~

Three-year interest rate = Average of three one-year rates

If the one-year interest rate today is i11 = 5%, one-year interest rate, one year ahead is i~1+1 = 6%, and the one-year interest rate two years ahead, i~t+z = 7%, then the expectation hypothesis tells us that the three-year interest rate will be i31 = (5% + 6% + 7%)/3 = 6%.

For a comparison between a three-year bond and three one-year bonds, we get

(5)

where the notation i31 stands for a three-year interest rate and i~t+2 for the expected one-year interest rate two years from now.

The general statement of the expectations hypothesis is that the interest rate on a bond with n years to maturity is the average of n expected future one-year interest rates:

(6)

What are the implications of this mathematical expression? Does the expectations hy-pothesis of the term structure of interest rates explain the three observations we started with? Let's look at each one.

1. The expectations hypothesis tells us that long-term bond yields are aU averages of expected future short-term yields-the same set of short-term interest rates-so interest rates of different maturities will move togetl1e1: From equation (6) we see that if the current one-year interest rate, i11 , changes, all the yields at higher maturities will change with it.

2. The expectations hypothesis implies that yields on slwrt-tenn bonds will be more volatile than yields on long-term bonds. Because long-term interest rates are av-erages of a sequence of expected future short-term rates, if the current 3-month interest rate moves, it will have only a small impact on the 1 0-year interest rate. Again, look at equation (6). 14

14Take a simple example in which the one-year and two-year interest rates, i 1, and i2,, are both 5 percent. If the one-year interest rate increases to 7 percent, then the two-year interest rate will rise to 6 percent. The two move together, and the short-term rate is more volatile than the long-term rate.


A picture can be worth a thousand words, but only if you know what it represents. To decode charts and graphs, use these strategies:

1. Read the title of the chart. This point may seem trivial, but titles are often very descriptive and can give you a good start in understanding a chart.

2. Read the label on the horizontal axis. Does the chart show the movements in a stock price or in the inter-est rate over minutes, hours, days, weeks, months, or years? Are the numbers spaced evenly?

Look at Figure 7.6, a sample of the Treasury yield curve that appears In The Wall Street Journal every day. The hori-zontal axis extends from three months to 30 years, but the in-crements are not evenly spaced. In fact, a distance that starts out as three months on the left-hand corner becomes over 10 years at the far right. The axis is drawn in this way for two reasons. First, it focuses the reader's eye on the shorter end of the yield curve. Second, the telescoped axis narrows so that it takes up less space.

Interestingly, this particular yield curve shows a slight downward slope from three months to one year, followed by a steep upward slope. This pattern suggests that investors ex-pected interest rates to decline sharply for the next year and then rise after that, which is exactly what happened.

3. Read the label on the vertical axis. What is the range of the data? This is a crucial piece of information because most charts are made to fill the space available. As a result, small movements can appear to be very large. Compare Panel A and Panel 8 of Figure 7.7 on page 177 on inflation. The first shows the percentage change in the consumer price index from 1960 to 2012; the

second shows the same data starting 25 years later, in 1985.

In Panel A, the vertical axis ranges from -4 percent to 16 percent; in Panel 8, it covers only -3 to 7 percent. To fill the second panel visually, the artist changed the vertical scale.

Treasury Yield Curve

April 23, 2001

6.00%

5.50

5.00

4.50

4.00

3.50

3.00 3 6 mos. yr.

2 5 10 30 maturities

SOURCE: The 1Val/ Street Jouma/, April 23, 2001. Used with pennission of Dow Jones & Company, Inc. via Copyright Clearance Center.

3. The expectations hypothesis cannot explain why long-term yields are nonnally higher than slwrt-tenn yields because it implies that the yield curve slopes up-ward only when interest rates are expected to rise. To explain why the yield curve normally slopes upward, the expectations hypothesis would suggest that interest rates are normally expected to rise. But as the data in Figure 7.3 show, interest rates have been trending downward for nearly 30 years, so anyone constantly forecasting interest-rate increases would have been sorely disappointed.

The expectations hypothesis has gotten us two-thirds of the way toward understand-ing the term structure of interest rates. By ignoring risk and assuming that investors view short- and long-term bonds as petfect substitutes, we have explained why yields at different maturities move together and why short-term interest rates are more vola-tile than long-term rates. But we have failed to explain why the yield curve normally slopes upward. To understand this, we need to extend the expectations hypothesis to include risk. After all, we all know that long-term bonds are riskier than short-term

Taking only a superficial look at these two charts can be misleading. Without noticing the difference in their vertical scales, we could conclude that the decline of inflation from 2008 to 2009 was as dramatic as the decline of inflation from 1980 to 1983. But, on closer inspection, Inflation fell from

u.s. Consumer Price Inflation

A. 12-Month Percentage Change 16 14 12

The Term Structure of Interest Rates Chapter 7 I 177

5.4 percent to -2.0 percent in the recent episode, while in the 1980-1983 period, it dropped from 14.6 percent to 2.5 percent. A proper reading of the charts leads to the correct conclusion that the earlier decline was nearly twice as large as the recent one.

B. 12-Month Percentage Change 7 6 5

10 4 8 3 6 2 4 1 2 0 0 -1

-2 -2 -4~~~~~~~~~~~~~~ -3~~~~~~~~LLLLLL~LL~

1960 1970 1980 1990 2000 2010 1985 1990 1995 2000 2005 2010

SOURCE: Bureau of Labor Statistics (FRED data code: CPJAUCSL).

bonds. Integrating this observation into our analysis will give us the liquidity premium theoJ)' of the term stmcture of interest rates.

Premium Theory Throughout our discussion of bonds, we emphasized that even default-free bonds are risky because of uncertainty about inflation and future interest rates. What are the impli-cations of these risks for our understanding of the term stmcture of interest rates? The answer is that risk is the key to understanding the usual upwards lope of the yield curve. Long-tenn interest rates are typically higher than short-term interest rates because long-term bonds are riskier than short-term bonds. Bondholders face both inflation and inter-est-rate risk. The longer the term of the bond, the greater both types of risk.

The reason for the increase in inflation risk over time is clear-cut. Remember that bondholders care about the purchasing power of the return-the real return-they


receive from a bond, not just the nominal dollar value of the coupon payments. Com-puting the real return from the nominal return requires a forecast of future inflation, or expected future inflation. For a three-month bond, an investor need only be concerned with inflation over the next three months. For a 10-year bond, however, computation of the real return requires a forecast of inflation over the next decade.

In summary, uncertainty about inflation creates uncertainty about a bond's real re-turn, making the bond a risky investment. The further we look into the future, the greater the uncertainty about inflation. We are more uncertain about the level of infla-tion several years from now than about the level of inflation a few months from now, which implies that a bond's inflation risk increases with its time to maturity.

What about interest-rate risk? Interest-rate risk alises from a mismatch between the investor's investment horizon and a bond's time to maturity. Remember that if a bondholder plans to sell a bond prior to maturity, changes in the interest rate (which cause bond prices to move) generate capital gains or losses. The longer the term of the bond, the greater the price changes for a given change in interest rates and the larger the potential for capital losses.

Because some holders of long-term bonds will want to sell their bonds before they mature, interest-rate risk concerns them. These investors require compensation for the risk they take in buying long-term bonds. As in the case of inflation, the risk increases with the term to maturity, so the compensation must increase with it.

What are the implications of including risk in our model of the term structure of inter-est rates? To answer this question, we can think about a bond yield as having two parts, one that is risk free and another that is a risk premium. The expectations hypothesis ex-plains the risk-free part, and inflation and interest-rate risk explain the risk premium. To-gether they form the the of interest rates. Adding the 1isk premium to equation (6), we can express this theory mathematically as

. +'e +'e + +e llt llt+l llt+2 , , , llt+n-J im = rp" + n (7)

where J]J11

is the risk premium associated with an n-year bond. The larger the risk, the higher the risk premium, l]J

11, is. Because risk rises with maturity, I]J

11 increases with

n, the yield on a long-term bond includes a larger risk premium than the yield on a short-term bond.

To get some idea of the size of the lisk premium I]J11

, we can look at the average slope of the term structure over a long period. From 1985 to mid-2013, the difference between the interest rate on a 10-year Treasury bond and that on a 3-month Treasury bill averaged nearly two percentage points. It is important to keep in mind that this risk premium will vary over time. For example, if inflation is very stable or the variability of the real interest rate were to fall, then the 10-year bond risk premium could easily fall below one percentage point.

Can the liquidity premium theory explain all three of our conclusions about the term structure of interest rates? The answer is yes. Like the expectations hypothesis, the liquidity premium theory predicts that interest rates of different maturities will move together and that yields on short-term bonds will be more volatile than yields on long-term bonds. And by adding a risk premium that grows with time to maturity, it explains why long-term yields are higher than short-term yields. Because the risk premium in-creases with time to maturity, the liquidity premium theory tells us that the yield curve will normally slope upward; only rarely will it lie flat or slope downward. (A flat yield curve means that interest rates are expected to fall; a downward-sloping yield curve suggests that the financial markets are expecting a significant decline in interest rates.)

The Information Content of Interest Rates Chapter 7 I 179

Standing in the middle of an open field during a thunder-storm is a good way to get hurt, so few people do it. Instead, they take shelter. Investors do exactly the same thing dur-ing financial storms; they look for a safe place to put their investments until the storm blows over. In practical terms, that means selling risky investments and buying the safest instruments they can: U.S. Treasury bills, notes, and bonds. An increase in the demand for government bonds coupled with a decrease in the demand for virtually everything else is called a When it happens, there is a dra-matic increase in the difference between the yields on safe and risky bonds-the risk spread rises.

n

When the government of Russia defaulted on its bonds in August 1998, the shock set off an almost unprecedented flight to quality. Yields on U.S. Treasuries plummeted, while those on corporate bonds rose. Risk spreads widened quickly; the difference between U.S. Treasury bills and commercial paper rates more than doubled, from its normal level of half a per-centage point to over one percentage point. The debt of coun-tries with emerging markets was particularly hard hit.

This flight to quality was what William McDonough called "the most serious financial crisis since World War II" (see the opening of this chapter). Because people wanted to hold only U.S. Treasury securities, the financial markets had ceased to function properly. Mr. McDonough worried that the problems in the financial markets would spread to the econ-omy as a whole. They didn't in the 1998 episode, but they did in the much larger financial crisis of 2007-2009.

The risk and term stmcture of interest rates contain useful information about overall economic conditions. These indicators are helpful in evaluating both the present health of the economy and its likely future course. Risk spreads provide one type of infor-mation, the term structure another. In the following sections we will apply what we have just learned about interest rates to recent U.S. economic history and show how forecasters use these tools.

When the overall growth rate of the economy slows or turns negative, it strains private businesses, increasing the risk that corporations will be unable to meet their financial obligations. The immediate impact of an impending recession, then, is to raise the risk premium on privately issued bonds. Importantly, though, an economic slowdown or recession does not affect the risk of holding government bonds.

The increased risk of default is not the same for all firms. The impact of a reces-sion on companies with high bond ratings is usually small, so the spread between U.S. Treasuries and Aaa-rated bonds of the same maturity is not likely to move by much. But for issuers whose finances were precarious prior to the downturn, the effect is quite different. Those borrowers who were least likely to meet their payment obligations when times were good are even less likely to meet them when times turn bad. There is a real chance that they will fail to make interest payments. Of course, firms for whom even the slightest negative development might mean disaster are the ones that issue low-grade bonds. The lower the initial grade of the bond, the more the default-risk premium rises as general economic conditions deteriorate. The spread between U.S. Treasury bonds and junk bonds widens the most.

Panel A of Figure 7.8 shows annual GDP growth over four decades superimposed on shading that shows the dates of recessions. (We'lllearn more about recession dat-ing in Chapter 22.) Notice that during the shaded peliods, growth is usually negative. In Panel B of Figure 7.8, GDP growth is drawn as the red line and the green line is the spread between yields on Baa-rated bonds and U.S. Treasury bonds. Note that the two lines move in opposite directions. (The correlation between the two series is -0.56.) That is, when the risk spread rises, output falls. The risk spread provides a


7.8 The Risk Spread and GDP Growth

1975 1980 1985 1990 1995 2000 2005 2010

1- GDP Growth I B. GDP Growth with Risk Spread

~ 10 6 0 8 5 Q) 6 b1)

-------~ 4 4 ?f2-A u 2 3 '"d cO

Q) 0 Q) b1) 2 H cO 0. ~ -2 Cf) 13 -4 1

~ -6 0 P-. 1970 1975 1980 1985 1990 1995 2000 2005 2010

1- GDP Growth (left scale) -Risk Spread (right scale) I

SOURCE: Bureau of Economic Analysis and Board of Governors of the Federal Rese1w System. GDP growth is the percentage c/umge.fromthe same quarter of the previous yem; while the yield spread is the dijference between/he average yield on Baa and 10-year U.S. Treaslll)' bonds during the quarte1: Shaded periods denote recessions. (FRED data codes: GDPCJ, BAA, GSJO, and USREC.)

useful indicator of general economic activity, and because financial markets operate every day, this information is available well before GDP data, which is published only once every three months. During the financial crisis of 2007-2009, the spread reached its widest level (5.6 percentage points) since the Great Depression of the 1930s.

of Like information on the risk structure of interest rates, information on the term structure-particularly the slope of the yield curve-helps us to forecast general eco-nomic conditions. Recall that according to the expectations hypothesis, long-term in-terest rates contain information about expected future short-term interest rates. And according to the liquidity premium theory, the yield curve usually slopes upward. The key term in this statement is usually. On rare occasions, short-term interest rates ex-ceed long-term yields. When they do, the term structure is said to be inverted, and the yield curve slopes downward.


The Term Spread and GDP Growth

A. Cunent Term Spread and GDP Growth ,..-..._ 10 cf< 8


By Caroline Baum January 7, 2010

If you were to stop the average person on the street and ask him about the Treasury yield curve, he'd probably look at you with a blank stare.

"What's a yield curve and why should I care about it?'' our passer-by might say.

I'm sure lots of people are wondeling the same thing, now that the yield curve is garnering headlines.

The yield curve, or the "term structure of interest rates" in economists' parlance, is the pictorial representation of two points connected by a line.

Not any two points, mind you. One point is a short-term interest rate that is either pegged by the central bank (the federal funds rate) or influenced by it (the yield on the three-month Treasury bill). The other point is some long-term interest rate, the yield on the 10- or 30-year Treasury, whose price is determined by the marketplace.

What's so important about these two particular points? Sin1ple. These two points, in relationship to one another,

succinctly describe the stance of monetary policy .... When short-term rates are below long-term rates, banks

have an incentive to go out and lend. "Borrow short, lend long" is the first rule of banking. That's why a steep yield curve is expansionary. It's a means to an end, the end being

money and credit creation. The central bank provides the raw material in the form of bank reserves. Depository institutions do-or don't do, in the present situation-the rest.

When the yield curve is inverted, with short rates higher than long rates, bank lending screeches to a halt.

The current ultra-steep yield curve is reminiscent of the early 1990s. Back then, banks were saddled with-what else?-bad real estate loans. History repeats itself with in-creased frequency.

Courtesy of what was at the time an unheard-oflow level for the funds rate (3 percent) and a vertical yield curve, banks bought boatloads of U.S. goverlllllent securities, which carry a zero risk-based capital weighting. The profit went right to the bottom line. Banks healed themselves, then got back to the business of allocating capital to the private sector.

People who are up in arms about the banks getting free money from the Fed and buying Treasuries should relax. That's how it always works. Banks lend to Uncle Sam so they can lend to you. During recessions, there's never much private-sector credit demand, so the Treasury's appetite satisfies the banks' need for high-quality assets.

As a devout believer in the predictive powers of the spread, I have always challenged those who try to decon-struct it, to mold its meaning to fit the current fashion. History shows the "what" matters more than the "why."

An inverted yield curve is one of the most reliable har-bingers of recession. Yet when the curve inverted in 2006 and stayed that way through 2007, the protests went out: This time is different! An influx of foreign capital-the

difference between the 10-year and 3-month yields-what is called a term Panel A of Figure 7.9 shows GDP growth (as in Figure 7.8) together with the contem-poraneous term spread (the growth and the term spread at the same time). Notice that when the term spread falls, GDP growth tends to fall somewhat later. In fact, when the yield curve becomes inverted, the economy tends to go into a recession roughly a year later. Panel B of Figure 7.9 makes this clear. At each point, GDP growth in the current year (e.g., 1990) is plotted against the slope of the yield curve one year earlier (e.g., 1989). The two lines clearly move together; their correlation is +0.42. What the data show is that when the term spread falls, GDP growth tends to fall one year later. The yield curve is a valuable forecasting tool.

An example illustrates the usefulness of this information. In the left-hand panel of Figure 7 .10, we can see that on January 23, 2001, the yield curve sloped downward from


"savings glut"-is depressing long-term rates. The inverted yield doesn't mean what it usually means.

It wasn't different. It just took longer. So what is the spread telling us now? All systems are

go. With the funds rate close to zero and 10-year Treasur-ies yielding about 3.8 percent, there's a lot of carry in the curve.

Yet money and credit aren't growing. The M2 money supply, which includes savings and time deposits and money market mutual funds in addition to Ml 's cmTency and demand deposits, rose an annualized 1.2 percent from May to November, according to the Fed. The three-month annualized increase of 4.2 percent is encouraging if it continues.

On the other side of the balance sheet, bank credit fell an annualized 5 percent in the same six-month period. Loans and leases plunged 9.2 percent at an annual rate.

"Banks aren't lending," says Paul Kasriel, chief econo-mist at the Northern Tmst Corp. in Chicago.

At least not to non-government entities. Bank purchases of U.S. Treasmy and agency secmities rose an annualized 21 percent from May to November. The next step is a return to private lending.

For all its allure, the yield curve isn't acting as a trans-mission mechanism just yet. That doesn't mean that it won't. When the financial system is impaired, it takes a steeper curve for a longer period of time to produce an expansion in money and credit-just as it did in the early 90s. The invita-tion to party stands. Banks have yet to RSVP.

With time, the spread will translate into more borrow-ing and lending. It's the Fed's job to prevent a trickle from becoming a cascade, fanning what would be the third bubble in two decades.

SOURCE: B/oombeJg.com, January 7, 2010. Used with permission of Bloomberg L.P. Copyright 2010. All rights reserved.

lESSONS Of THE ARTIClE The slope of the yield curve can help predict the direction and speed of economic growth. At the beginning of 2010, the yield curve was unusually steep, pointing to a strong economic expansion (see Figure 7.98, where the steep slope in 2010 is shown as a high value for the term spread). However, in the aftermath of the financial crisis of 2007-2009, lenders were especially cautious about extending credit to risky borrowers, even though risk spreads had narrowed sharply from crisis peaks (see Figure 7.88). The author argues that it is only a matter of time until the steep yield curve encourages lenders to start lending again. However, the willingness to lend depends on how quickly intermediaries gain confidence that borrowers will repay. Even in 2013, many poten-tial borrowers were still repairing their balance sheets, which had been damaged by the record plunge of U.S. housing prices.

3 months to 5 years, then upward for maturities to 30 years. This pattem indicated that interest rates were expected to fall over the next few years. Eight months later, after mon-etaty policy had eased and the U.S. economy had slowed substantially, the Treasury yield curve sloped upward again (see the right-hand panel of Figure 7.1 0). At that point, growth was at a virtual standstill, and policymakers were doing eve1ything they could to get the economy moving again. They had reduced interest rates by more than 3 percentage points over a period of less than nine months. Thus, investors expected little in the way of short-term interest-rate reductions. This prediction turned out to be wrong, however; interest rates kept falling after the terrorist attacks of September 11, 2001. They continued to fa11 through the remainder of 2001 as the economy went into a mild recession.

The yield curve did not predict the depth or duration of the recession of 2007-2009. One- and two-year market rates did not anticipate the persistent plunge of overnight


7JO The U. S. Treasury Yield Curve in 2001

January 23, 2001 6.50%

-Monday 1 week ago

6.00 4 k - wee sago

5.50 tttl+ 5.00

August 17, 2001 6.00% .

-Fnday 5.50 1 week ago 5.00 -4 weeks ago

4.50 ~~ I 4.00 /~++!

3.50 450 _f6--1~:z-.s 10 30 300 3. 6--1 2 5 10 30

mos. yr. maturities mos. yr. maturities

SOURCE: Board of Governors of the Federal Resene System.

rates. As financial institutions weakened in the crisis, the widening risk spread signaled a severe economic downturn, providing a more useful predictor in this episode.

We started this chapter by asking why different types of bonds have different yields and what it is we can learn from those differences. After a bit of work, we can now see that differences in both risk and time to maturity affect bond yields. The less likely the issuer is to repay or the longer the time to maturity, the riskier a bond and the higher its yield. Even more importantly, both increases in the risk spread and an inverted yield curve suggest troubled economic times ahead.

Terms benchmark bond, 168 commercial paper, 167 expectations hypothesis of the term

stmcture, 172 fallen angel, 166 flight to quality, 179 interest-rate spread, 162 inverted yield curve, 181 investment-grade bond, 164 junk bond, 166 liquidity premium theory of the term

stmcture, 178 municipal bonds, 171

prime-grade commercial paper, 167 rating, 163 ratings downgrade, 167 ratings upgrade, 167 risk spread, 168 tisk stmcture of interest rates, 168 spread over Treasuries, 168 taxable bond, 171 tax-exempt bond, 171 term spread, 182 term stmcture of interest rates, 172 yield curve, 173

Chapter Lessons Chapter 7 I 185

Using FRED: Codes for Data in This Chapter

Data Series

Moody's Aaa corporate bond yield Moody's Baa corporate bond yield BofA Merrill Lynch US corporate BBB effective yield BofA Merrill Lynch US high-yield BB effective yield BofA Merrill Lynch US high-yield B effective yield BofA Merrill Lynch US high-yield CCC or below effective yield 1 0-year Treasury constant maturity rate 3-month Treasury bill rate 3-month AA nonfinancial commercial paper rate Consumer price index Real GOP Brazil Treasury bill rate

Chapter Lessons

FRED Data Code

AAA BAA BAMLCOA4CBBBEY BAMLHOA1HYBBEY BAMLHOA2HYBEY BAMLHOA3HYCEY GS10 TB3MS CPN3M CPIAUCSL GDPC1 INTGSTBRM193N

1. Bond ratings summarize the likelihood that a bond issuer will meet its payment obligations. a. Highly rated investment-grade bonds are those with the lowest risk of default. b. If a firm encounters financial difficulties, its bond rating may be downgraded. c. Commercial paper is the short-tenn version of a privately issued bond. d. Junk bonds are high-risk bonds with very low ratings. Firms that have a high

probability of default issue these bonds. e. Investors demand compensation for default risk in the form of a risk premium.

The higher the risk of default, the lower a bond's rating, the higher its risk pre-mium, and the higher its yield.

2. Municipal bonds are usually exempt from income taxes. Because investors care about the after-tax returns on their investments, these bonds have lower yields than bonds whose interest payments are taxable.

3. The term stmcture of interest rates is the relationship between yield to maturity and time to maturity. A graph with the yield to maturity on the vertical axis and the time to maturity on the horizontal axis is called the yield curve. a. Any theory of the term stmcture of interest rates must explain three facts:

i. Interest rates of different maturities move together. ii. The yields on short-term bonds are more volatile than the yields on long-term

bonds. iii. Long-term yields are usually higher than short-term yields.

b. The expectations hypothesis of the term stmcture of interest rates states that long-term interest rates are the average of expected future short-term interest rates. This hypothesis explains only the first two facts about the term stmcture of interest rates.


c. The liquidity premium theory of the term structure of interest rates, which is based on the fact that long-term bonds are riskier than short-term bonds, explains all three facts in part a.

4. The tisk structure and the term structure of interest rates both signal financial mar-kets' expectations of future economic activity. Specifically, the likelihood of are-cession will be higher when a. The risk spread, or the range between low- and high-grade bond yields, is wide. b. The yield curve slopes downward, or is inverted, so that short-term interest rates

are higher than long-term interest rates.

Conceptual and Analytical Problems 1. Consider a firm that issued a large quantity of commercial paper in the period

leading to a financial crisis. ( LO I) a. How would you expect the credit rating of the commercial paper to evolve as

the crisis unfolds? b. Would you alter your prediction if, rather than commercial paper, the firm was

instead issuing asset-backed commercial paper? 2. Suppose that a major foreign government defaults on its debt. What, if anything,

will happen to the position and slope of the U.S. Treasury yield curve? (L02) 3. What was the connection between house price movements, the growth in

subprime mortgages, and securities backed by these mortgages-on the one hand-and-on the other hand-the difficulties encountered by some financial institutions during the 2007-2009 financial crisis? (LOI)

4. Suppose that the interest rate on one-year bonds is 4 percent and is expected to be 5 percent in one year and 6 percent in two years. Using the expectations hypothesis, compute the yield curve for the next three years. (L02)

5. * According to the liquidity premium theory, if the yield on both one- and two-year bonds are the same, would you expect the one-year yield in one year's time to be higher, lower, or the same? Explain your answer. (L02)

6. You have $1,000 to invest over an investment horizon of three years. The bond market offers various options. You can buy (i) a sequence of three one-year bonds; (ii) a three-year bond; or (iii) a two-year bond followed by a one-year bond. The current yield curve tells you that the one-year, two-year, and three-year yields to maturity are 3.5 percent, 4.0 percent, and 4.5 percent, respectively. You expect that one-year interest rates will be 4 percent next year and 5 percent the year after that. Assuming annual compounding, compute the return on each of the three investments, and discuss which one you would choose. (L02)

7.* Suppose that the yield curve shows that the one-year bond yield is 3 percent, the two-year yield is 4 percent, and the three-year yield is 5 percent. Assume that the risk premium on the one-year bond is zero, the risk premium on the two-year bond is 1 percent, and the tisk premium on the three-year bond is 2 percent. (L02) a. What are the expected one-year interest rates next year and the following year? b. If the risk premiums were all zero, as in the expectations hypothesis, what

would the slope of the yield curve be?

*Indicates more difficult problems

Conceptual and Analytical Problems Chapter 7 I 187

8. * If inflation and interest rates become more volatile, what would you expect to see happen to the slope of the yield curve? (L02)

9. As economic conditions improve in countries with emerging markets, the cost of borrowing funds there tends to fall. Explain why. (L03)

10. Suppose your local government, threatened with bankruptcy, decided to tax the interest income on its own bonds as part of an effort to rectify serious budgetary woes. What would you expect to see happen to the yields on these bonds? (LOJ)

11. * If, before the change in tax status, the yields on the bonds described in Prob-lem 10 were below the Treasury yield of the same maturity, would you expect this spread to narrow, to disappear, or to change sign after the policy change? Explain your answer. (LOJ)

12. Suppose the risk premium on U.S. corporate bonds increases. How would the change affect your forecast of future economic activity, and why? (L03)

13. If regulations restricting institutional investors to investment-grade bonds were lifted, what do you think would happen to the spreads between yields on investment-grade and speculative-grade bonds? (LOJ)

14. Suppose a country with a struggling economy suddenly discovered vast quanti-ties of valuable minerals under government-owned land. How might the gov-ernment's bond rating be affected? Using the model of demand and supply for bonds, what would you expect to happen to the bond yields of that country's government bonds? (L03)

15. The misrating of mortgage-backed securities by rating agencies contributed to the financial c1isis of 2007-2009. List some recommendations you would make to avoid such mistakes in the future. ( LO 1)

16. How do you think the abolition of investor protection laws would affect the risk spread between corporate and government bonds? (LOJ)

17. You and a friend are reading The Wall Street Joumal and notice that the Treasury yield curve is slightly upward sloping. Your friend comments that all looks well for the economy but you are concerned that the economy is heading for trouble. Assuming you are both believers in the liquidity premium theory, what might account for your difference of opinion? (L03)

18. Do you think the term spread was an effective predictor of the recession that started in December 2007? Why or why not? (L03)

19. * Given the data in the accompanying table, would you say that this economy is heading for a boom or for a recession? Explain your choice. (L03)

3-month 10-year Baa corporate Treasury-bill Treasury bond 1 0-year bond

January 1.00% 3.0% 7.0% February 1.05 3.5 7.2 March 1.10 4.0 7.5 April 1.20 4.3 7.7 May 1.25 4.5 7.8


FRED ECOHOHIC PATA I ST. LOUIS FED

Scan here for quick access to the hints for these problems. Need a barcode reader? Try Scanlife, available in your app store.

Data Exploration For detailed instructions on using Federal Resen1e Economic Data (FRED) online to answer each of the follmFing problems, visit www.mhhe. com!moneyandbanking4e and click on Student Edition, then Data Exploration Hints.

1. Did the financial crisis of 2007-2009 affect financial and nonfinancial firms to the same extent? For the period beginning in 2006, plot the spread between the interest rates on three-month nonfinancial commercial paper (FRED code: CPN3M) and three-month Treasury bills (FRED code: TB3MS). Plot a similar spread using the interest rates on three-month financial commercial paper (FRED code: CPF3M) and Treasury bills (FRED code: TB3MS). Compare the evolution of these two spreads. (LOJ)

2. The Federal Reserve Bank of St. Louis publishes a weekly index of financial stress (FRED code: STLFSI) that summarizes strains in financial markets, including li-quidity problems. For the period beginning in 1994 plot this index and, as a second line, the difference between the Baa corporate bond yield (FRED code: WBAA) and the 10-year U.S. Treasury bond yield (FRED code: WGS 10YR). Does the index STLFSI provide an early warning of stress? (L03)

3. How did the Great Depression (1929-1933) and the Great Recession of 2007-2009 affect expectations of corporate default? To investigate, constmct for each of those periods a separate plot of the corporate bond yield spread. For the Depression period, plot from 1930 to 1933 the difference between the Baa corporate bond yield (FRED code: BAA) and the long-term government bond yield (FRED code: LTGOVTBD). For the Great Recession, plot from 2007 to 2009 the difference between the Baa yield (FRED code: BAA) and the 10-year Treasury bond yield (FRED code: GS10). Compare the plots. (LOJ)

4. How reliably does an inverted yield curve anticipate a recession? How far in ad-vance? Plot from 1970 (as in Figure 7.9A) the difference between the 10-year Trea-sury yield (FRED code: GS 1 0) and the three-month Treasury bill rate (FRED code: TB3MS). Discuss the vadability of the time between an inversion of the yield curve and the subsequent recession. (L03)

5. Download the data from the graph produced in Data Exploration Problem 4 and (a) find the most recent period for which the yield curve was (approximately) fiat and (b) the longest time period for which yield curve was inverted. (L02)

Stocks, Stock Markets, and Market Efficiency

Understand . . . LOt The characteristics of common stock L02 Measures of the level of the stock market L03 The valuation of stocks L04 Investing in stocks for the long run LOS The stock market's role in the economy

Stocks play a prominent role in our financial and economic lives. For individuals, they provide a key instrument for holding personal wealth as well as a way to diversify, spreading and reducing the risks that we face. Importantly, diversifiable risks are risks that are more likely to be taken. By giving individuals a way to transfer risk, stocks supply a type of insurance enhancing our ability to take risk. 1

For companies, they are one of several ways to obtain financing. Beyond that, though, stocks and stock markets are a central link between the financial world and the real economy. Stock prices are fundamental to the functioning of a market-based economy. They tell us the value of the companies that issued the stocks and, like all other prices, they allocate scarce investment resources. The firms deemed most valuable in the mar-ketplace for stocks are the ones that will be able to obtain financing for growth. When resources flow to their most valued uses, the economy operates more efficiently.

Mention of the stock market provokes an emotional reaction in many people. They see it as a place where fortunes are easily made or lost, and they recoil at its unfathomable booms and busts. During one infamous week in October 1929, the New York Stock Exchange lost more than 25 percent of its value-an event that marked the beginning of the Great Depression. In October 1987, prices fell nearly 30 percent in one week, including a record decline of 20 percent in a single day. Crashes of this magnitude have become part of the stock market's folklore, creating the popular im-pression that stocks are very risky.

In the 1990s, stock prices increased nearly fivefold and Americans forgot about the "black Octobers." By the end of the decade, many people had come to see stocks as 1This point was central to our discussions of risk in Chapter 5. Our ability to diversify risk either through the explicit purchase of insurance or through investment strategies means that we are able to do risky things that we otherwise would not do.

190 I Chapter 8 Stocks, Stock Markets, and Marl{et Efficiency

almost a sure thing; you could not afford not to own them. In 1998, nearly half of all U.S. households owned some stock, either directly or indirectly through mutual funds and managed retirement accounts.

When the market's inexorable rise finally ended, the ensuing decline seemed more like a slowly deflating balloon than a crash. From early 2000 to the week following the tenorist attacks of September 11, 2001, the stock prices of the United States' biggest companies, as measured by the Dow Jones Industrial Average, fell more than 30 per-cent. While many stocks recovered much of their loss fairly quickly, a large number did not. Dming the same period, the Nasdaq Composite index fell 70 percent, from 5,000 to 1,500; it has remained well below its March 2000 peak ever since. Because the Nas-daq tracks smaller, newer, more technologically oriented companies, many observers dubbed this episode the "Internet bubble."

The plunge of the U.S. stock market in the recent financial crisis was much broader and greater, roughly halving the market value by early 2009 from its 2007 peak of about $26 trillion. And, although the stock market surpassed this peak in the first quar-ter of 2013, it fared less well after accounting for inflation.

Yet, contrary to popular mythology, stock prices do tend to 1ise-even after adjust-ing for inflation-over the long term, collapsing only on those infrequent occasions when normal market mechanisms are out of alignment. For most people, the experi-ence of losing or gaining wealth suddenly is more memorable than the experience of making it gradually. By being preoccupied with the potential short-term losses associ-ated with crashes, we lose sight of the gains we could realize if we took a longer-term view. The goal of this chapter is to try to make sense of the stock market-to show what fluctuations in stock value mean for individuals and for the economy as a whole and look at a critical connection between the financial system and the real economy. We will also explain how it is that things sometimes go awry, resulting in bubbles and crashes. First, however, we need to define the basics: what stocks are, how they origi-nated, and how they are valued.

The Essential Characteristics of Common Stock Stocks, also known as common stock or are shares in a firm's ownership. A finn that issues stock sells part of itself, so that the buyer becomes a part owner. Stocks as we know them first appeared in the 16th century. They were created to raise funds for global exploration. Means had to be found to finance the dangerous voyages of explorers such as Sir Francis Drake, Henry Hudson, and Vasco de Gama. Aside from kings and queens, no one was wealthy enough to finance these risky ventures alone. The solution was to spread the risk through joint-stock companies, organizations that issued stock and used the proceeds to finance several expeditions at once. In exchange for investing, stockholders received a share of the company's profits.

These early stocks had two important characteristics that we take for granted today. First, the shares were issued in small denominations, allowing investors to buy as little or as much of the company as they wanted; and second, the shares were transferable, meaning that an owner could sell them to someone else. Today, the vast majority of large companies issue stock that investors buy and sell regularly. The shares normally are quite numerous, each one representing only a small fraction of a company's total value. The large number and small size of individual shares-prices are usually below $100 per share-make the purchase and sale of stocks relatively easy.

The Essential Characteristics of Common Stock Chapter 8 I 191

Examples of Stock Certificates

Until recently, all stockowners received a certificate from the issuing company. Figure 8.1, on the left, shows the first stock certificate issued by the Ford Motor Com-pany in 1903, to Henry Ford. The right-hand side of the figure shows a more recent stock certificate issued by the World Wrestling Federation (WWF), renamed World Wrestling Enterprises (WWE). The WWE is the media and entertainment company that produces the wrestling events involving characters like The Rock and Hulk Hogan. The former governor of Minnesota, Jesse Ventura, worked for the WWF before entering politics.

Today, most stockholders no longer receive certificates; the odds are that you will never see one. Instead, the infmmation they bear is computerized, and the shares are registered in the names of brokerage fitms that hold them on investors' behalf. This procedure is safer because computerized certificates can't be stolen. It also makes the process of selling the shares much easier.

The ownership of common stock conveys a number of rights. First and most im-portantly, a stockholder is entitled to participate in the profits of the enterprise. Impor-tantly, however, the stockholder is merely a residual claimant. If the company runs into financial trouble, only after all other creditors have been paid what they are owed will the stockholders receive what is left, if anything. Stockholders get the leftovers!

To understand what being the residual claimant means, let's look at the case of a software manufacturer. The company needs a number of things to make it run. The list might include rented office space, computers, programmers, and some cash balances for day-to-day operations. These are the inputs into the production of the company's software output. If we took a snapshot of the company's finances on any given day, we woul

chapter 7 to 9

Documents

high bond

aaa bond ratings

high rating

rating agencies

multitude of bonds

types of bonds

ratings agencies

wellknown bond rating