Yield

Yield Curve
Curve Analysis
Analysis

www.dnatrainingconsulting.net © DNA Training & Consulting

 Bond market data

Coupon Face value Price

1-year note none 100 95

2-year note 8% paid 100 100

annually

3-year note 9.5%, paid 100 100

annually

www.dnatrainingconsulting.net © DNA Training & Consulting

 Deriving zero-coupon rates (1)

► Need to determine 2-year zero-coupon rate, e.g. for pricing a new

bond issue, or for valuing an investment proposal, or for pricing a
swap or other derivative

► Remember that for 2-year note can write

100 = (8 × DF1 ) + (108 × DF2 )

► But DF1 = 0.95, so can solve for DF2

► Rearranging above equation leads to

(100 − 8 × 0.95)
DF2 = = 0.8556
108
www.dnatrainingconsulting.net © DNA Training & Consulting
 Deriving zero-coupon rates (2)

► But we also know that

DF2 = (1 + z2 ) 2 = 0.8556

which leads to
z 2 = 1 / 0.8556 − 1 = 8.1125%
► From any discount factor you can derive the corresponding zero-
coupon rate, and from any zero-coupon rate you can derive the
corresponding discount factor

► In general 1
DFN =
(1+ zN )N
www.dnatrainingconsulting.net © DNA Training & Consulting
 Non-par cases

► Notes do not need to be trading at par

► If for example 2-year note is trading at 98% of par, simply write

98 = (8 × DF1 ) + (108 × DF2 )

and then proceed exactly as before

www.dnatrainingconsulting.net © DNA Training & Consulting

 Proceeding to 3-year zero rate

► Now we write

100 = (9.5 × DF1 ) + (9.5 × DF2 ) + (109.5 × DF3 )

and then solve for DF3, since we know already DF1 and DF2.

(100 − (9.5 × 0.95) − (9.5 × 0.8566)

► This leads to DF3 = = 0.7566
109.5

1
► So since DF3 = 0.7566 =
(1 + z3 )3

it follows that z 3 = 9 .7435 %

www.dnatrainingconsulting.net © DNA Training & Consulting
 Generalized approach

► Now assume we have already calculated DF1, DF2, etc all the way
to DFN-1, and now wish to calculate DFN from the note maturing in
N years and carrying an annual coupon of CN, we would write

Pr ice N = (C N × DF1 ) + (C N × DF2 ) + − − − + (C N × DFN −1 ) + ((100 + C N ) × DFN )

► Rearranging the above equation leads to

Pr ice N − (C N × DF1 ) + (C N × DF2 ) + − − − + (C N × DFN −1 ) = (100 + C N ) × DFN

So DF = Pr ice N − (C N × DF1 ) + (C N × DF2 ) + − − − + (C N × DFN −1 )

100 + C N
N

Pr ice N − C N × ( DF1 + DF2 + − − + DFN −1 )

=
100 + C N

www.dnatrainingconsulting.net © DNA Training & Consulting

 Bootstrapping
Benchmark Benchmark Discount Cumulative Zero-coupon
Tenor
Coupon Price Factor Sum of DFs Rates
This one entry is a discount
yield, not a coupon

1 5.26% 95% 0.9500 0.9500 5.26%

2 8.00% 100% 0.8556 1.8056 8.11%
3 9.50% 100% 0.7566 2.5622 9.74%
4 10.50% 100% 0.6615 3.2237 10.88%
5 11.00% 100% 0.5814 3.8051 11.45%
6 11.25% 100% 0.5141 4.3192 11.73%
7 11.38% 100% 0.4565 4.7757 11.85%
8 11.44% 100% 0.4071 5.1828 11.89%
9 11.48% 100% 0.3633 5.5461 11.91%
10 11.50% 100% 0.3248 5.8710 11.90%

www.dnatrainingconsulting.net © DNA Training & Consulting

 Flat

Benchmark Benchmark Discount Cumulative Zero-coupon

Tenor
Coupon Price Factor Sum of DFs Rates

This one entry is a discount

yield, not a coupon

1 5.26% 95% 0.9500 0.9500 5.26%

2 5.26% 100% 0.9025 1.8525 5.26%
3 5.26% 100% 0.8574 2.7099 5.26%
4 5.26% 100% 0.8145 3.5244 5.26%
5 5.26% 100% 0.7738 4.2982 5.26%
6 5.26% 100% 0.7351 5.0333 5.26%
7 5.26% 100% 0.6983 5.7316 5.26%
8 5.26% 100% 0.6634 6.3950 5.26%
9 5.26% 100% 0.6302 7.0253 5.26%
10 5.26% 100% 0.5987 7.6240 5.26%

www.dnatrainingconsulting.net © DNA Training & Consulting

 Inverted
Benchmark Benchmark Discount Cumulative Zero-coupon
Tenor
Coupon Price Factor Sum of DFs Rates

This one entry is a discount

yield, not a coupon

1 5.26% 95% 0.9500 0.9500 5.26%

2 5.11% 100% 0.9051 1.8551 5.11%
3 4.96% 100% 0.8650 2.7201 4.95%
4 4.81% 100% 0.8292 3.5493 4.79%
5 4.66% 100% 0.7973 4.3466 4.63%
6 4.51% 100% 0.7691 5.1157 4.47%
7 4.36% 100% 0.7443 5.8601 4.31%
8 4.21% 100% 0.7227 6.5827 4.14%
9 4.06% 100% 0.7039 7.2866 3.98%
10 3.91% 100% 0.6879 7.9746 3.81%

www.dnatrainingconsulting.net © DNA Training & Consulting

 More complex curves (1)

► What if shortest securities mature in less than one year, and/or coupons
on notes are paid more frequently than annually?

► Assume your bond market has the following instruments:

Issue date Maturity Coupon Face Price

date value
3-month June 30, Sep. 30, none 100 98
bill 2009 2009
6-month June 30, Dec. 30, none 100 96
bill 2009 2009
1-year Mar. 30, Mar. 30, 9% 100 100.8
note 2009 2010 quarterly
1-year June 30, June 30, 8% 100 100.1
note 2009 2010 quarterly

► Note that instrument in the third row has only 3 cash flows remaining
www.dnatrainingconsulting.net © DNA Training & Consulting
 More complex curves (2)

► For this note we can write

100 . 8 = ( 2 . 25 × DF 0 .25 ) + ( 2 . 25 × DF 0 . 5 ) + (102 . 25 × DF 0 .75 )

► But we know from the discount securities that

DF0.25 = 0.98, and DF0.5 = 0.96

100.8 − 2.25 × (0.98 + 0.96)

so DF0.75 = = 0.9431
102.25
► And for the note maturing on June 30, 2010 we can write
100 .1 = ( 2 × DF0.25 ) + ( 2 × DF0.5 ) + ( 2 × DF0.75 ) + (102 × DF1 )

100.1 − 2 × (0.98 + 0.96 + 0.9431)

leading us to DF1 = = 0.9250
102
www.dnatrainingconsulting.net © DNA Training & Consulting
 Deriving forward rates

► Definition: interest rate quoted by a financial institution which, like

any other interest rate, relates to a period with a defined tenor but
for a period starting in the future

► So could speak of 1-year forward 1-year rate, and denote this as

f1,2 which would mean the one-year zero-coupon rate for the time
period starting in one year from now (and therefore ending in two
years from now)

www.dnatrainingconsulting.net © DNA Training & Consulting

 Calculating f1,2 rate

► Done via no-arbitrage exercise

No-arbitrage pricing exercise

8.11%

0 1 YR 2 YRS

5.26% f1,2?

www.dnatrainingconsulting.net © DNA Training & Consulting

 Deriving forward rates

► Definition: interest rate quoted by a financial institution which relates to a period

with a defined tenor but for a period starting in the future

► So could speak of 1-year forward 1-year rate, and denote this as f1,2 and would
mean the 1-year zero-coupon rate for the time period starting one year from now.

► So we can write (1 + z 2 ) 2 = (1 + z1 ) × (1 + f1, 2 )

(1 + z 2 ) 2 (1 + 8 .11 %) 2
which gives us f 1, 2 = −1 = − 1 = 11 .04 %
(1 + z1 ) (1 + 5 .26 %)
► But a faster solution is simply to rearrange the equation above into
1 1
(1 + f1, 2 ) × = , so (1 + f1, 2 ) × DF 2 = DF1 ,
(1 + z 2 ) 2
(1 + z1 )
which leads to DF
f 1, 2 = 1
− 1 = 11 . 04 %
DF 2

www.dnatrainingconsulting.net © DNA Training & Consulting

 Generalized forward rates

► Can also speak of forward rate for a longer period: e.g. f3,7 refers to
the zero-coupon interest rate for the 4-year period starting 3 years
from today (and therefore ending 7 years from today)

► Most generally the zero-coupon rate for the period starting in a

years from today and finishing in b years from today is fa,b

www.dnatrainingconsulting.net © DNA Training & Consulting

 Calculating fa,b rate

► Done via no-arbitrage exercise

No-arbitrage pricing exercise

zb%

0 a YRS b YRS
fa,b?
Za%

www.dnatrainingconsulting.net © DNA Training & Consulting

 Generalized forward rates

► Can also speak of forward rate for a longer period: e.g. f3,7 refers to the
zero-coupon interest rate for the 4-year period starting 3 years from today
(and therefore ending 7 years from today)

► Most generally the zero-coupon rate for the period starting in a years from
today and finishing in b years from today is fa,b

(b−a )
So we can write (1 + z b ) = (1 + z a ) × (1 + f a ,b )
b a
►

(1 + f a ,b ) ( b − a ) 1
so =
(1 + z b ) b (1 + z a ) a

DF a 1 /( b − a )
which leads to f a ,b =( ) −1
DF b

www.dnatrainingconsulting.net © DNA Training & Consulting

 1-year Forwards
Zero- 1-year
Benchmark Benchmark Discount Cumulative
Tenor coupon Forward Notation
Coupon Price Factor Sum of DFs
Rates Rates

This one entry is a discount

yield, not a coupon

1 5.26% 95% 0.9500 0.9500 5.26% 5.26% f0,1

2 8.00% 100% 0.8556 1.8056 8.11% 11.04% f1,2
3 9.50% 100% 0.7566 2.5622 9.74% 13.08% f2,3
4 10.50% 100% 0.6615 3.2237 10.88% 14.37% f3,4
5 11.00% 100% 0.5814 3.8051 11.45% 13.77% f4,5
6 11.25% 100% 0.5141 4.3192 11.73% 13.10% f5,6
7 11.38% 100% 0.4565 4.7757 11.85% 12.61% f6,7
8 11.44% 100% 0.4071 5.1828 11.89% 12.14% f7,8
9 11.48% 100% 0.3633 5.5461 11.91% 12.05% f8,9
10 11.50% 100% 0.3248 5.8710 11.90% 11.84% f9,10

www.dnatrainingconsulting.net © DNA Training & Consulting

 1-year Forwards

16.00%

14.00% Par

12.00%
Zero-
coupon
10.00%

8.00% Forward

6.00%

4.00%

2.00%

0.00%
0 2 4 6 8 10

www.dnatrainingconsulting.net © DNA Training & Consulting

 2-year forwards
Zero- 1-year 2-year
Benchmark Benchmark Discount Cumulative
Tenor coupon Forward Forward Notation
Coupon Price Factor Sum of DFs
Rates Rates Rates

This one entry is a discount

yield, not a coupon

1 5.26% 95% 0.9500 0.9500 5.26% 5.26%

2 8.00% 100% 0.8556 1.8056 8.11% 11.04% 8.11% f0,2
3 9.50% 100% 0.7566 2.5622 9.74% 13.08% 12.05% f1,3
4 10.50% 100% 0.6615 3.2237 10.88% 14.37% 13.72% f2,4
5 11.00% 100% 0.5814 3.8051 11.45% 13.77% 14.07% f3,5
6 11.25% 100% 0.5141 4.3192 11.73% 13.10% 13.44% f4,6
7 11.38% 100% 0.4565 4.7757 11.85% 12.61% 12.85% f5,7
8 11.44% 100% 0.4071 5.1828 11.89% 12.14% 12.38% f6,8
9 11.48% 100% 0.3633 5.5461 11.91% 12.05% 12.10% f7,9
10 11.50% 100% 0.3248 5.8710 11.90% 11.84% 11.95% f8,10

www.dnatrainingconsulting.net © DNA Training & Consulting

 2-year Forwards

16.00%

14.00% Par

12.00%
Zero-
coupon
10.00%
1-yr
8.00% forward

6.00% 2-yrs
forward
4.00%

2.00%

0.00%
0 1 2 3 4 5 6 7 8 9 10

www.dnatrainingconsulting.net © DNA Training & Consulting

 Flat 2
Zero- 1-year
Benchmark Benchmark Discount Cumulative
coupon Forward
Coupon Price Factor Sum of DFs
Tenor Rates Rates

This one entry is a discount

yield, not a coupon

1 5.26% 95% 0.9500 0.9500 5.26% 5.26%

2 5.26% 100% 0.9025 1.8525 5.26% 5.26%
3 5.26% 100% 0.8574 2.7099 5.26% 5.26%
4 5.26% 100% 0.8145 3.5244 5.26% 5.26%
5 5.26% 100% 0.7738 4.2982 5.26% 5.26%
6 5.26% 100% 0.7351 5.0333 5.26% 5.26%
7 5.26% 100% 0.6983 5.7316 5.26% 5.26%
8 5.26% 100% 0.6634 6.3950 5.26% 5.26%
9 5.26% 100% 0.6302 7.0253 5.26% 5.26%
10 5.26% 100% 0.5987 7.6240 5.26% 5.26%

www.dnatrainingconsulting.net © DNA Training & Consulting

 6.00%

Par

Zero-
4.00% coupon

Forward

2.00%

0.00%
0 2 4 6 8 10

www.dnatrainingconsulting.net © DNA Training & Consulting

 Dangers of inverted curves

Tenor Coupon/Yield
1 9.0%
2 8.5%
3 8.0%
4 7.5%
5 7.0%
6 6.5%
7 6.0%
8 5.5%
9 5.0%
10 4.5%

www.dnatrainingconsulting.net © DNA Training & Consulting

 1-year forwards
Zero- 1-year
Benchmark Benchmark Discount Cumulative
Tenor coupon Forward Notation
Coupon Price Factor Sum of DFs
Rates Rates

This one entry is a discount

yield, not a coupon

1 9.00% 91.74% 0.9174 0.9174 9.00% 9.00% f0,1

2 8.50% 100% 0.8498 1.7672 8.48% 7.96% f1,2
3 8.00% 100% 0.7950 2.5622 7.95% 6.89% f2,3
4 7.50% 100% 0.7515 3.3137 7.40% 5.80% f3,4
5 7.00% 100% 0.7178 4.0315 6.86% 4.69% f4,5
6 6.50% 100% 0.6929 4.7244 6.30% 3.59% f5,6
7 6.00% 100% 0.6760 5.4004 5.75% 2.51% f6,7
8 5.50% 100% 0.6663 6.0667 5.21% 1.45% f7,8
9 5.00% 100% 0.6635 6.7302 4.66% 0.43% f8,9
10 4.50% 100% 0.6671 7.3973 4.13% -0.54% f9,10

www.dnatrainingconsulting.net © DNA Training & Consulting

 Dangers of inverted curves: conclusion

► Problem arose with choice of par curve in the first place, and specifically with the
last two entries at the bottom

► 9-year bond with an annual coupon of 5% and a 10-year bond with an annual
coupon of 4.5%, and both trading at par, are inconsistent with one another; you
would always earn a higher all-in return over 10 years with the 9-year bond given its
higher coupon, irrespective of what the 1-year interest rate in 9 years turns out to
be

► Additional 50 bps you have earned annually, again assuming in a worst case it is
reinvested at zero, would have aggregated by the end of year 9 a total of 4.5%, so
already your wealth under the 9-year note would be greater than what it would
become under the 10-year note in one year

► Conclusion: not all par curves are “possible”: given the coupon on any given note,
the coupon on the note one-year longer cannot be smaller by more than a certain
amount; if it is, the longer instrument becomes inferior to the shorter one under all
possible interest rate scenarios and would presumably find no buyers

www.dnatrainingconsulting.net © DNA Training & Consulting

 Alternative one-year investments

Consider the following choice between one-year investments:

1. Purchase the one-year note, yielding 5.26%, and hold it to

maturity; or

2. Purchase the 2-year zero-coupon note yielding 8.11%, and sell it

after one year

► Strategy 1 has a pre-determined return of 5.26%

► Strategy 2’s return over the 1-year holding period will depend
on what is the 1-year discounting rate in one year’s time, when
we come to selling the investment

www.dnatrainingconsulting.net © DNA Training & Consulting

 Alternative 1-year investments

Strategy 1 Strategy 2: Strategy 2:

Scenario 1 Scenario 2
Entry Price 95% 85.56% 85.56%

Exit Price 100% 100%/(1+10%) = 100%/(1+12%) =

90.91% 89.28%

All-in Return (100%/95%) – 1 (90.91%/85.56%) – 1 = (89.28%/85.56%) – 1 =

= 5.26% 6.26% 4.36%

Performance Outperform by 100 bps Underperform by 90 bps

versus
Strategy 1

www.dnatrainingconsulting.net © DNA Training & Consulting

 Alternative strategies and breakeven

Strategy 1 Strategy 2: Strategy 2: Strategy 2:

Scenario 1 Scenario 2 Scenario 3

Entry Price 95% 85.56% 85.56% 85.56%

Exit Price 100% 100%/(1+10%) = 100%/(1+12%) = 100%/

90.91% 89.28% (1+11.04%) =
90.06%
All-in Return (100%/95%) - 1 (90.91%/85.56%) (89.28%/85.56%) (90.06%/85.56%)
= 5.26% – 1 = 6.26% – 1 = 4.36% – 1 = 5.26%

Performance Outperform by Underperform by Matches exactly

versus 100 bps 90 bps Strategy 1
Strategy 1

www.dnatrainingconsulting.net © DNA Training & Consulting

 Forwards and realized returns

► More generally, if the yield changes implied by the forward rates are
subsequently realized, all zero-coupon bonds, regardless of maturity, earn
the same return over any holding period

► Specifically in our scenario 3 in which the 1-year forward 1-year rate,

11.04%, was actually realized, the 1-year zero coupon investment and the
2-year zero-coupon investment earned exactly the same holding period
return of 5.26%, despite their different maturities

► Corollary: if a bank funds itself and invests at different points of the same
zero-coupon curve, and if forward rates are subsequently realized, then the
bank will break even, irrespective of which two points of the curve it
chooses for its gapping strategy

www.dnatrainingconsulting.net © DNA Training & Consulting

 Are forwards good predictors?

► Few generalizations are possible given enormous variety of yield curves

around the world and difficulty of stripping away entirely all other risks and
leaving bank exposed exclusively to gapping risk

► For very short maturities in US risk-free curves, there appears to be a fairly

systematic tendency on the part of interest rate forwards to overshoot the
eventual outcome – i.e. it is significantly more common for the eventual
short-term rate to turn out to be lower than was implied by the forward at
some defined earlier time, than it is for the eventual short-term rate to turn
out to be higher than was implied by the forward at that earlier time

► So strategy known as running a negative gap, i.e. of lengthening the tenor

of assets versus the tenor of liabilities, will on average earn positive returns
for the bank, on a cumulative basis, over the long term

► But there may well be sustained periods of time in which the returns from
this strategy are in fact negative

www.dnatrainingconsulting.net © DNA Training & Consulting

 “Inevitability” of this truth

► This result is consistent with basic investment theory, especially when you
remember that 2-year zero has higher duration than one-year zero, and
hence higher price volatility

► If forwards are on average realized, then investor with one-year horizon

would always opt for 1-year zero over 2-year zero in our example, since
there would be no point in opting for the instrument with the higher price
volatility if the average expected return from holding it is the same as that of
the instrument with the lower price volatility

► 2-year instrument compensates for its higher duration and hence higher
price volatility by offering potential holders a higher expected return over a
1-year holding period than its shorter competitor with the lower price
volatility

www.dnatrainingconsulting.net © DNA Training & Consulting

 Yield curve shape

Three principal theories for explaining shape of yield curve:

► The expectations theory

► The liquidity or term premium theory

► The preferred habitat theory

► Can only summarize in the briefest of ways the principal arguments

for and against each theory

www.dnatrainingconsulting.net © DNA Training & Consulting

 Expectations theory

► States that curve shape reflects market expectations regarding future

interest rate

► Upward sloping curve reflects expectations of rising rates, flat curve reflect
neutral expectations, and inverted curve reflects expectations of falling
rates

► Certainly in cases of a steep upward sloping curve or one that is

significantly inverted, particularly at the short end, research confirms the
accuracy of this theory to some degree

► Flattish curves are also generally consistent with neutral expectations

regarding future short-term interest rates, but again no perfect correlation is
observed

www.dnatrainingconsulting.net © DNA Training & Consulting

 Liquidity or term premium theory (1)

► States that the longer the tenor of the instrument the greater should be its
expected return, all other things being equal

► Results from investor preference for short-tenor assets and specifically

ones with low price volatility and which can be turned into cash more easily
than others, either by selling them into secondary market at close to par or
by holding them until maturity

► Under this theory, even when interest rate expectations are neutral, we
should expect the curve to be upward sloping to some degree, otherwise no
incentive would exist for extending maturity and thus assuming greater
price risk and lower asset liquidity

www.dnatrainingconsulting.net © DNA Training & Consulting

 Liquidity or term premium theory (2)

► Over the last 50 years, say, US short-term interest rates have moved within
a range, so expectations have been cumulatively neutral, despite extended
periods in which rates were expected to be rising or falling steadily

► Yet curve has been upward sloping some 90% of the time during this half
century, suggesting some evidence of a term premium in bond yields – i.e.
a higher yield on account solely of maturity extension

► Note that the two factors we have analyzed so far can sometimes neutralize
one another: in a market in which short-term rates are generally expected to
fall, yet preference for low duration assets is high, the first factor would
cause the curve to invert, while the second factor would cause the curve to
slope upward

www.dnatrainingconsulting.net © DNA Training & Consulting

 Preferred habitat theory (1)

► Some investors, especially regulated institutional investors, are often

attracted to a point of the curve and repelled by another point for reasons
other than the quest for yield, including regulatory factors, tax factors,
accounting factors, and legal factors, among others

► For example pension funds may be severely penalized by the law if they
find themselves “underfunded”, which causes them to purchase very long-
dated bonds to match fund to the greatest degree possible their very long-
dated liabilities

► They do this even when the yields on long-dated bonds are substantially
lower than they are on medium-term bonds of comparable credit quality

www.dnatrainingconsulting.net © DNA Training & Consulting

 Preferred habitat theory (2)

► Example from opposite end of curve: money-market funds, central banks

and corporate treasurers craving for low duration

► These investors often tend to keep their cash invested in t-bills of very
shortest maturity, even when slight lengthening of maturity even when slight
lengthening of maturity can result in a yield pick-up of 25 or 50 bps annually
or more

► This causes shortest end of curve to slope upward significantly, even when
interest rate expectations are neutral or even are inclining towards lower
rates

www.dnatrainingconsulting.net © DNA Training & Consulting