9/16/2018

15.401 Industry Overview 15.401 Industry Overview 15.401

Fixed-income securities are financial claims with promised cashflows of

known fixed amount paid at fixed dates.
Classification of Fixed-Income Securities:
 Treasury Securities
– U.S. Treasury securities (bills, notes, bonds)
Finance Theory – Bunds, JGBs, U.K. Gilts
– ….
 Federal Agency Securities
– Securities issued by federal agencies (FHLB, FNMA $\ldots$)
 Corporate Securities
– Commercial paper
– Medium-term notes (MTNs)
– Corporate bonds
– ….
 Municipal Securities
Lectures 4–6: Fixed-Income Securities  Mortgage-Backed Securities Courtesy of SIFMA. Used with permission. The Securities Industry and Financial Markets Association (SIFMA) prepared
this material for informational purposes only. SIFMA obtained this information from multiple sources believed to be reliable
 Derivatives (CDO’s, CDS’s, etc.) as of the date of publication; SIFMA, however, makes no representations as to the accuracy or completeness of such third
party information. SIFMA has no obligation to update, modify or amend this information or to otherwise notify a reader
thereof in the event that any such information becomes outdated, inaccurate, or incomplete.

Slide 3 Slide 5

Critical Concepts 15.401 Industry Overview 15.401 Industry Overview 15.401

 Industry Overview
 Valuation
 Valuation of Discount Bonds U.S. Bond Market Debt 2006 ($Billions) U.S. Bond Market Issuance 2006 ($Billions)
 Valuation of Coupon Bonds
 Measures of Interest-Rate Risk Asset-Backed,
2,016.70, 8%
Municipal,
2,337.50, 9% Municipal,
 Corporate Bonds and Default Risk Money Treasury,
Asset-Backed,
674.6, 16%
265.3, 6%
Markets, 4,283.80, 16% Treasury,
 The Sub-Prime Crisis 3,818.90, 14% Federal 599.8, 14%
Agency, 546.9,
Federal 13%
Agency,
Readings 2,665.20, 10%
Corporate, Mortgage-
Mortgage- 748.7, 17% Related,
Corporate, Related, 1,475.30, 34%
5,209.70, 19% 6,400.40, 24%

Slide 2 Slide 4 Slide 6

1
 9/16/2018

Industry Overview 15.401 Valuation 15.401 Valuation of Discount Bonds 15.401

Components of Valuation If r Varies Over Time

Fixed-Income Market Participants
 Time value of principal and coupons  Denote by Rt the one-year spot rate of interest in year t
 Risks
– Inflation
Investors:
Issuers: Intermediaries:  Governments
– Credit
 Governments
 Corporations
 Primary Dealers  Pension Funds – Timing (callability)  But we don’t observe the entire sequence of future spot rates today!
 Other Dealers  Insurance
 Commercial Banks
 Investment Banks 
Companies
Commercial Banks
– Liquidity
 States
 Credit-rating Agencies – Currency
 Municipalities  Hedge
 Credit Enhancers MutualFunds
Funds
 SPVs
 Liquidity Enhancers
 Foreign Institutions  Individuals
Foreign Institutions For Now, Consider Riskless Debt Only
 U.S. government debt (is it truly riskless?)
 Consider risky debt later  Today’s T-year spot rate is an “average” of one-year future spot rates
 (P0,F,r0,T) is over-determined

Slide 9 Slide 11 Slide 13

Valuation 15.401 Valuation of Discount Bonds 15.401 Valuation of Discount Bonds 15.401

Cashflow: Pure Discount Bond Example:

 Maturity  No coupons, single payment of principal at maturity On 20010801, STRIPS are traded at the following prices:
Example. A 3-year bond with principal of $1,000
 Coupon and annual coupon payment of 5% has the  Bond trades at a “discount” to face value
 Principal following cashflow:  Also known as zero-coupon bonds or STRIPS*
 Valuation is straightforward application of NPV

For the 5-year STRIPS, we have

 Note: (P0, r, F) is “over-determined”; given two, the third is determined

Now What If r Varies Over Time?

 Different interest rates from one year to the next
 Denote by rt the spot rate of interest in year t

*Separate Trading of Registered Interest and Principal Securities

Slide 10 Slide 12 Slide 14

2
 9/16/2018

Valuation of Discount Bonds 15.401 Valuation of Discount Bonds 15.401 Valuation of Discount Bonds 15.401

Suppose We Observe Several Discount Bond Prices Today Term Structure Contain Information About Future Interest Rates More Generally:
 Forward interest rates are today’s rates for transactions between two
future dates, for instance, t1 and t2.
 For a forward transaction to borrow money in the future:
– Terms of transaction is agreed on today, t = 0
– Loan is received on a future date t1
– Repayment of the loan occurs on date t2
 Note: future spot rates can be (and usually are) different from current
corresponding forward rates
 Implicit in current bond prices are forecasts of future spot rates!
 These current forecasts are called one-year forward rates
 To distinguish them from spot rates, we use new notation:

Term Structure of Interest Rates

Slide 15 Slide 17 Slide 19

Valuation of Discount Bonds 15.401 Valuation of Discount Bonds 15.401 Valuation of Discount Bonds 15.401

Example:
Term Structure Contain Information About Future Interest Rates Term Structure Contain Information About Future Interest Rates
r0,t As the CFO of a U.S. multinational, you expect to repatriate $10MM from
a foreign subsidiary in one year, which will be used to pay dividends
one year afterwards. Not knowing the interest rates in one year, you
would like to lock into a lending rate one year from now for a period of
one year. What should you do? The current interest rates are:

Strategy:
 Borrow $9.524MM now for one year at 5%
Maturity  Invest the proceeds $9.524MM for two years at 7%

 What are the implications of each of the two term structures?

Slide 16 Slide 18 Slide 20

3
 9/16/2018

Valuation of Discount Bonds 15.401 Valuation of Discount Bonds 15.401 Valuation of Coupon Bonds 15.401

Example (cont): Example (cont):

Coupon Bonds
Outcome (in millions of dollars): Strategy:  Intermediate payments in addition to final principal payment
 Buy 20,000,000 of 3 year discount bonds, costing  Coupon bonds can trade at discounts or premiums to face value
 Valuation is straightforward application of NPV (how?)

Example:
 Finance this by (short)selling 4 year discount bonds of amount
 3-year bond of $1,000 par value with 5% coupon

 This creates a liability in year 4 in the amount $21,701,403

 The locked-in 1-year lending rate one year from now is 9.04%, which  Aside: A shortsales is a particular financial transaction in which an
is the one-year forward rate for Year 2 individual can sell a security that s/he does not own by borrowing the
security from another party, selling it and receiving the proceeds, and
then buying back the security and returning it to the orginal owner at a
later date, possibly with a capital gain or loss.

Slide 21 Slide 23 Slide 25

Valuation of Discount Bonds 15.401 Valuation of Discount Bonds 15.401 Valuation of Coupon Bonds 15.401

Example: Example (cont): Valuation of Coupon Bonds

Suppose that discount bond prices are as follows:  Cashflows from this strategy (in million dollars):

 Since future spot rates are unobservable, summarize them with y

A customer would like to have a forward contract to borrow $20MM three

years from now for one year. Can you (a bank) quote a rate for this  y is called the yield-to-maturity of a bond
forward loan?  It is a complex average of all future spot rates
 There is usually no closed-form solution for y; numerical methods
All you need is the forward rate f4 which should be your quote for the must be used to compute it (Tth-degree polynomial)
forward loan  The yield for this strategy or “synthetic bond return” is given by:
 (P0, y, C) is over-determined; any two determines the third
 For pure discount bonds, the YTM’s are the current spot rates
 Graph of coupon-bond y against maturities is called the yield curve

Slide 22 Slide 24 Slide 26

4
 9/16/2018

Valuation of Coupon Bonds 15.401 Valuation of Coupon Bonds 15.401 Valuation of Coupon Bonds 15.401

Liquidity Preference Model

U.S. Treasury Yield Curves
 Investors prefer liquidity
 Long-term borrowing requires a premium
 Expected future spot < current forward

E[Rk] < fk
E[Rk] = fk − Liquidity Premium

Source: Bloomberg

Slide 27 Slide 29 Slide 31

Valuation of Coupon Bonds 15.401 Valuation of Coupon Bonds 15.401 Valuation of Coupon Bonds 15.401

Time Series of U.S. Treasury Security Yields Models of the Term Structure
Another Valuation Method for Coupon Bonds
 Expectations Hypothesis
 Theorem: All coupon bonds are portfolios of pure discount bonds
 Liquidity Preference
 Preferred Habitat
 Valuation of discount bonds implies valuation of coupon bonds
 Market Segmentation  Proof?
 Continuous-Time Models
– Vasicek, Cox-Ingersoll-Ross, Heath-Jarrow-Morton Example:
 3-Year 5% bond
Expectations Hypothesis
 Sum of the following
 Expected Future Spot = Current Forward
discount bonds:
E0[Rk] = fk – 50 1-Year STRIPS
– 50 2-Year STRIPS
– 1050 3-Year STRIPS

Slide 28 Slide 30 Slide 32

5
 9/16/2018

Valuation of Coupon Bonds 15.401 Measures of Interest-Rate Risk 15.401 Measures of Interest-Rate Risk 15.401

Bonds Subject To Interest-Rate Risk Example:

Example (cont):
 Price of 3-Year coupon bond must equal the cost of this portfolio  As interest rates change, bond prices also change Consider a 4-year T-note with face value $100 and 7% coupon, selling at
 What if it does not?  Sensitivity of price to changes in yield measures risk $103.50, yielding 6%.
 For T-notes, coupons are paid semi-annually. Using 6-month intervals,
In General: the coupon rate is 3.5% and the yield is 3%.
P = C P0,1 + C P0,2 + · · · + (C + F )PO,T

 If this relation is violated, arbitrage opportunities exist

 For example, suppose that
P > C P0,1 + C P0,2 + · · · + (C + F )PO,T
 Short the coupon bond, buy C discount bonds of all maturities up to T
and F discount bonds of maturity T
 No risk, positive profits ⇒ arbitrage

Slide 33 Slide 35 Slide 37

Valuation of Coupon Bonds 15.401 Measures of Interest-Rate Risk 15.401 Measures of Interest-Rate Risk 15.401

Macaulay Duration Example (cont):

What About Multiple Coupon Bonds?  Duration (in 1/2 year units) is
 Weighted average term to maturity
XT q
X
Dm = k · ωk ωk = 1
k=1 k=1
Ck /(1 + y)k PV(Ck )  Modified duration (volatility) is
ωk = =
P P

 Suppose n is much bigger than T (more bonds than maturity dates)  Sensitivity of bond prices to yield changes
 This system is over-determined: T unknowns, n linear equations T
X Ck
P =  Price risk at y=0.03 is
 What happens if a solution does not exist? k=1
(1 + y)k
 This is the basis for fixed-income arbitrage strategies ∂P −1 T
X Ck
= k·
∂y 1 + y k=1 (1 + y)k
1 ∂P Dm
= −  Note: If the yield moves up by 0.1%,
P ∂y 1+ y
= −Dm∗ the bond price decreases by 0.6860%
Modified Duration

Slide 34 Slide 36 Slide 38

6
 9/16/2018

Measures of Interest-Rate Risk 15.401 Measures of Interest-Rate Risk 15.401 Measures of Interest-Rate Risk 15.401

Macaulay Duration
 Relation between duration and convexity: 1 8
XkCk
 Duration decreases with coupon rate ∗
Dm = = 3.509846
 Duration decreases with YTM ∂P ∂ 2P (y0 − y)2 1 + 0.06 k=1 2P (1 +
0.06 )k
P (y0) ≈ P (y) + (y) · (y0 − y) + (y) · 2 2
 Duration usually increases with maturity ∂y ∂y2 2 1 8
X k(k + 1)Ck
· ¸ Vm = = 14.805972
– For bonds selling at par or at a premium, duration always increases = P (y) · ∗ (y0 − y) + V1 (y0
1 − Dm − y) 2 (1 + 0.06 )2 4P (1 + 0.06 )k
m 2³ k=1 2
with maturity 2
P (y0) ≈ P (0.06) 1 − 3.509846(y0 − 0.06)+
– For deep discount bonds, duration can decrease with maturity !
 Second-order approximation to bond-price function (y0 − 0.06)2
– Empirically, duration usually increases with maturity  Portfolio versions: 14.805972
X
2
P = Pj
j
P (0.08) ≈ P (0.06)(1 − 0.0701969+0.0029611)
1 ∂P X Pj
Dm∗ (P) ≡ − = D ∗m,j ≈ 93.276427
P ∂y j
P
1 ∂ 2P X Pj ∗
Vm∗ (P) ≡ − = Vm,j P (0.08) = 93.267255
P ∂y2 j
P

Slide 39 Slide 41 Slide 43

Measures of Interest-Rate Risk 15.401 Measures of Interest-Rate Risk 15.401 Corporate Bonds and Default Risk 15.401

Macaulay Duration Non-Government Bonds Carry Default Risk

 For intra-year coupons and annual yield y  A default is when a debt issuer fails to make a promised payment
T
X (interest or principal)
Annual Dm = k · ωk/q  Credit ratings by rating agencies (e.g., Moody's and S&P) provide
k=1 indications of the likelihood of default by each issuer.
∗ y
Annual Dm = Annual Dm/(1 + )
q Credit Risk Moody's S&P Fitch

Investment Grade
Convexity Highest Quality Aaa AAA AAA
High Quality (Very Strong) Aa AA AA
 Sensitivity of duration to yield changes Upper Medium Grade (Strong) A A A
Medium Grade Baa BBB BBB

∂ 2P T
X
1 Ck Not Investment Grade
= k · (k + 1) · Somewhat Speculative Ba BB BB
∂y2 (1 + y)2 k=1
(1 + y)k Speculative B B B
Highly Speculative Caa CCC CCC
1 ∂ 2P Most Speculative Ca CC CC
Imminent Default C C C
= Vm
P ∂y2 Default C D D

Slide 40 Slide 42 Slide 44

7
 9/16/2018

Corporate Bonds and Default Risk 15.401 Corporate Bonds and Default Risk 15.401 Corporate Bonds and Default Risk 15.401

Decomposition of Corporate Bond Yields

Moody’s Baa 10-Year Treasury Yield Decomposition of Corporate Bond Yields
 Promised YTM is the yield if default does not occur
 Expected YTM is the probability-weighted average of all possible
yields
 Default premium is the difference between promised yield and
expected yield
 Risk premium (of a bond) is the difference between the expected
yield on a risky bond and the yield on a risk-free bond of similar
maturity and coupon rate

Example: Suppose all bonds have par value $1,000 and

 10-year Treasury STRIPS is selling at $463.19, yielding 8%
 10-year zero issued by XYZ Inc. is selling at $321.97
 Expected payoff from XYZ's 10-year zero is $762.22

Source: Fung and Hsieh (2007)

Slide 45 Slide 47 Slide 49

Corporate Bonds and Default Risk 15.401 Corporate Bonds and Default Risk 15.401 The Sub-Prime Crisis 15.401

What’s In The Premium?  For the 10-year zero issued by XYZ:

Why Securitize Loans?
 Expected default loss, tax premium, systematic risk premium (Elton, et
Repack risks to yield more homogeneity within categories
al 2001)
More efficient allocation of risk
– 17.8% contribution from default on 10-year A-rated industrials
Creates more risk-bearing capacity
 Default, recovery, tax, jumps, liquidity, and market factors (Delianedis
and Geske, 2001) Provides greater transparency
– 5-22% contribution from default Supports economic growth
 Credit risk, illiquidity, call and conversion features, asymmetric tax Benefits of sub-prime market
treatments of corporates and Treasuries (Huang and Huang 2002)
– 20-30% contribution from credit risk But Successful Securitization Requires:
 Liquidity premium, carrying costs, taxes, or simply pricing errors Diversification
(Saunders and Allen 2002) Accurate risk measurement
“Normal” market conditions
Reasonably sophisticated investors

Slide 46 Slide 48 Slide 50

8
 9/16/2018

The Sub-Prime Crisis 15.401 An Illustrative Example 15.401 An Illustrative Example 15.401

“Confessions of a Risk Manager” in The Economist, August 7, 2008: Consider Simple Securitization Example:
Assuming
 Two identical one-period loans, face value $1,000 $1,000 Independent Defaults
Like most banks we owned a portfolio of different tranches of
collateralised-debt obligations (CDOs), which are packages of
 Loans are risky; they can default with prob. 10% I.O.U. Portfolio
asset-backed securities. Our business and risk strategy was to buy  Consider packing them into a portfolio Value Prob.
pools of assets, mainly bonds; warehouse them on our own
 Issue two new claims on this portfolio, S and J $2,000 81%
balance-sheet and structure them into CDOs; and finally distribute
them to end investors. We were most eager to sell the non-  Let S have different (higher) priority than J Portfolio $1,000 18%
investment-grade tranches, and our risk approvals were  What are the properties of S and J?
conditional on reducing these to zero. We would allow positions $0 1%
of the top-rated AAA and super-senior (even better than AAA)  What have we accomplished with this “innovation”? $1,000
tranches to be held on our own balance-sheet as the default risk I.O.U.
Let’s Look At The Numbers!
was deemed to be well protected by all the lower tranches, which
would have to absorb any prior losses.
© The Economist. All rights reserved. This content is excluded from our Creative Commons license.
For more information, see http://ocw.mit.edu/fairuse .

Slide 51 Slide 53 Slide 55

The Sub-Prime Crisis 15.401 An Illustrative Example 15.401 An Illustrative Example 15.401

“Confessions of a Risk Manager” in The Economist, August 7, 2008:

In May 2005 we held AAA tranches, expecting them to rise in
90%
value, and sold non-investment-grade tranches, expecting them to $1,000 $1,000 $1,000
Price = 90% × $1,000 + 10% × $0
go down. From a risk-management point of view, this was perfect: I.O.U. I.O.U. C.D.O.
have a long position in the low-risk asset, and a short one in the = $900
higher-risk one. But the reverse happened of what we had 10%
expected: AAA tranches went down in price and non- $0 (Default)
investment-grade tranches went up, resulting in losses as we Senior Tranche
marked the positions to market. Portfolio
This was entirely counter-intuitive. Explanations of why this had
happened were confusing and focused on complicated cross- 90%
correlations between tranches. In essence it turned out that there $1,000 $1,000 $1,000
Price = 90% × $1,000 + 10% × $0
had been a short squeeze in non-investment-grade tranches, I.O.U. = $900 I.O.U. C.D.O.
driving their prices up, and a general selling of all more senior
structured tranches, even the very best AAA ones. 10% $0 (Default)
© The Economist. All rights reserved. This content is excluded from our Creative Commons license.
For more information, see http://ocw.mit.edu/fairuse . Junior Tranche

Slide 52 Slide 54 Slide 56

9
 9/16/2018

An Illustrative Example 15.401 An Illustrative Example 15.401 An Illustrative Example 15.401

Assuming Independent Defaults Assuming Independent Defaults Assuming Perfectly Correlated Defaults

$1,000 Portfolio $1,000 Portfolio $1,000 Portfolio

Senior Junior Senior Junior Senior Junior
Value Prob. C.D.O. Value Prob. Value Prob.
C.D.O. Tranche Tranche Tranche Tranche C.D.O. Tranche Tranche

$2,000 81% $1,000 $1,000 $2,000 81% $1,000 $1,000 $2,000 90% $1,000 $1,000

Senior Tranche $1,000 18% $1,000 $0 Senior Tranche $1,000 18% $1,000 $0 Senior Tranche $0 10% $0 $0

$0 1% $0 $0 $0 1% $0 $0
Bad State For
Senior Tranche
$1,000 $1,000 Price for Senior Tranche = 99% × $1,000 + 1% × $0 $1,000
Bad State For (10%) Bad State For
C.D.O. Senior C.D.O. = $990 C.D.O.
Bad State Junior Tranche
Tranche (1%) For Junior Price for Junior Tranche = 81% × $1,000 + 19% × $0 (10%)
Tranche (19%) = $810
Junior Tranche Junior Tranche Junior Tranche

Slide 57 Slide 59 Slide 61

An Illustrative Example 15.401 An Illustrative Example 15.401 An Illustrative Example 15.401

Assuming Independent Defaults Assuming Perfectly Correlated Defaults

$1,000 Portfolio $1,000 Portfolio

Non-Investment Grade Senior Junior Senior Junior
“Similar” to Value Prob. Value Prob.
C.D.O. Tranche Tranche C.D.O. Tranche Tranche
Junior
Tranche? $2,000 81% $1,000 $1,000 $2,000 90% $1,000 $1,000
Senior Tranche $1,000 18% $1,000 $0 Senior Tranche $0 10% $0 $0

“Similar” to $0 1% $0 $0
Senior Price for Senior Tranche = 90% × $1,000 + 10% × $0
Tranche? = $900 (was $990)
$1,000 $1,000
C.D.O. But What If Defaults Become Highly Correlated? C.D.O. Price for Junior Tranche = 90% × $1,000 + 10% × $0
= $900 (was $810)

Source: Moody’s Junior Tranche Junior Tranche

Slide 58 Slide 60 Slide 62

10
 9/16/2018

Implications 15.401

To This Basic Story, Add:

Very low default rates (new securities)
Very low correlation of defaults (initially)
Aaa for senior tranche (almost riskless)
Demand for senior tranche (pension funds)
Demand for junior tranche (hedge funds) $1,000
Fees for origination, rating, leverage, etc. C.D.O.
Insurance (monoline, CDS, etc.)
Equity bear market, FANNIE, FREDDIE

Then, National Real-Estate Market Declines

Default correlation rises
Senior tranche declines
Junior tranche increases
Ratings decline
Unwind ⇒ Losses ⇒ Unwind ⇒ …

Slide 63

11