Active Bond Management

Active bond management strategies rely on expectations of

interest rate movements or changes in yield-spread relationships.
Active bond management contrasts with passive bond management
strategies which make no attempt to enhance performance by
capitalizing on these expectations. Rather, the objective of passive
strategies is to design a portfolio that will achieve the performance of
a predetermined benchmark. The benchmark might either be to
achieve the return on a specified index or it might be to accumulate
sufficient dollars to satisfy a future liability stream.

Active bond managers attempt to exploit the four general factors that
affect a fixed income portfolio’s return:

• changes in the level of interest rates

• changes in the shape of the yield curve
• changes in yield spreads across/between sectors
• changes in yield spreads for a particular instrument

Changes in the level of interest rates

For changes in the level of interest rates, interest rate anticipation

strategies can be used from an expected level change in interest
rates. A level change in interest rates is defined as a parallel shift in
the yield curve. Duration measures the inverse relationship between
a level change in interest rates and a fixed income portfolio’s value.
To enhance return, if rates are expected to decline a manager would
increase the duration of the portfolio; conversely, if rates are
expected to rise, the manager would shorten portfolio duration.

1
Changes in the shape of the yield curve

Because duration only measures the effect of a parallel shift in

interest rates, durational neutral strategies can be used to profit from
an expected change in the shape of the yield curve. The yield curve
can shift in various ways, but the two most common are 1) a
downward shift combined with a steepening and 2) an upward shift
combined with a flattening. Two portfolios can have similar durations,
but if their bond maturities are different they will react differently to
changes in the shape of the yield curve.
One method to measure the effects of a change in the shape of
the yield curve is to construct a bullet portfolio and a barbell portfolio,
each with an equivalent duration. A bullet portfolio is one where
maturities are centered at a single point on the yield curve. A barbell
portfolio is one where maturities are concentrated at two extreme
points on the yield curve, with one maturity shorter and the other
longer than the bullet portfolio’s maturity. In general, the bullet will
outperform if the yield curve steepens with long rates rising relative to
short rates, because of the capital loss on the longer term bonds in
the barbell portfolio. Conversely, if the yield curve flattens with long
rates falling relative to short rates, the barbell will almost certainly
outperform because of the positive effect of capital gains on long term
bonds. Even if the yield curve shifts in a parallel fashion, the relative
performance of bullets and barbells may be different, even if their
duration is the same. The reason is that the bullet often has a higher
yield, but the barbell has greater convexity. Thus, if the yield curve
rises by a small amount, the bullet may outperform because of its
yield advantage. However, for large yield curve increases, the barbell
may outperform because its convexity advantage will keep it from
falling as much in price as the bullet.
Partial duration can also be used to measure a portfolio’s
sensitivity to yield curve shape changes. Partial duration measures
the change in a bond’s value due to a shift in one point in the yield
curve while all other yield curve points remain the same. Using this

2
method, a manager can hold portfolio duration constant and select
bonds that provide superior performance for an expected change in
one point on the yield curve.

Changes in yield spreads across/between

sectors

Yield spread strategies can be used to profit from an expected

change in current bond sector spreads. The bond market can be
segregated into different sectors such as type (corporate, treasury, or
mortgage backed), quality (treasury, government agency, AAA, AA,
A, BBB, below investment grade), or call feature (callable or non-
callable).
Yield spread strategies are based on an assumption that current
yield spreads between sectors are not consistent with some “normal”
yield spread level. In these strategies, often called intermarket spread
swaps, a manager sells bonds in one sector and buys bonds in
another sector in the hopes of profiting as the yield spread moves
from its current level to its “normal” level. These strategies are
generally independent of interest rate anticipation strategies which
attempt to capitalize on expectations regarding the level of interest
rates. There are many instances when yield spread strategies can be
used, with two examples being given below.
During periods of economic expansion, corporate yield spreads
(the spread between treasury issues and non-treasury issues)
generally narrow, reflecting corporate bonds decreased credit risk.
Conversely, during recessions, corporate yield spreads generally
widen, reflecting the increased credit risk due to a weakening
economy. Thus, if an expansion is forecast, a manager would
purchase corporates and sell treasuries in anticipation of greater
price appreciation or less price erosion due to the spread narrowing.
On the other hand, if the economy is expected to weaken the

3
manager would sell corporates and buy treasuries to reduce the price
loss due to the spread widening.
Another example is the decision to purchase callable or
noncallable bonds. If interest rates are expected to decline, callable
yields generally widen because the call option is becoming more
valuable. Callable bonds are short a call option which increases in
value as interest rates drop, and buyers of callable bonds will
demand more compensation in yield spread to offset the call risk as
interest rates decline. If interest rates are expected to fall, a manager
would sell callable bonds and purchase non-callable bonds to avoid
the depreciation in price of callable bonds due to spread widening
(negative convexity at work). Conversely, if interest rates are
expected to rise and callable bond spreads to narrow, the manager
would sell non-callable bonds and purchase callable bonds. Interest
rate volatility also plays a role in the spread. As volatility increases,
the value of the embedded call option rises, causing callable bond
prices to fall and the yield spread to widen. Thus, if volatility is
expected to increase, the manager would sell callable bonds and
purchase non-callable bonds; if volatility is expected to decrease,
non-callable bonds would be sold and callable bonds would be
purchased.

Changes in yield spreads for a particular

instrument

Individual security yield spread strategies, often called substitution

swaps, can be used to select one of two bonds that are similar in all
aspects except that one has a higher yield. For example, in a given
credit quality sector, a bond might be selected if a manager feels that
its credit quality should be higher than other similarly rated bonds. If
this analysis is confirmed by the market via a rating upgrade, the
bond will go up in value.

4
 Another example is to compare two mortgage backed securities
(MBSs) of similar coupon, maturity, and type, where different
prepayment assumptions lead to different prices and yield spreads. If
a manager expects a different prepayment assumption than the
market, the manager can act on that assumption in the hopes that the
market will agree with his prepayment assumption in the future and
“correctly” value that particular MBS. For example, for a particular
discount MBS or a PO, if a manager expects faster prepayments than
the market, the manager could purchase that security to realize the
greater value of receiving principal repayments quicker than the
market anticipates. Similarly, for a particular premium MBS or an IO,
if the manager expects slower prepayments than the market, the
manager would purchase that security because its lower price is
overly compensating for the prepayment risk

Bond Portfolio Management

Strategies
Stock market investors will choose a particular risk level on the SML
and invest at this point, choosing only those securities that lie on the
SML (or above it). Stock investors have different levels of risk/return
requirements Bond investors will do the same thing. A young,
aggressive bond investor may choose a high risk bond & is willing to
risk his principal investment. A retiree may not be willing to take a
risky bond investment and may, instead invest in conservative bonds.

Individual investors choose to invest in bonds. Also, pension plans,

banks, insurance companies and other institutions invest in bonds. At
any rate, all investors are interested in a bond investment strategy.
There are three major types of strategies:

1. passive portfolio management

strategies

5
 2. active portfolio management
strategies
3. matched-funding strategies

In the 1950s the bond market was considered a safe, conservative

investment. At that time a buy-and-hold strategy was sufficient.
However, times changed, in the 1960s inflation increased, and interest
rates became more volatile. Thus, with more volatile interest rates,
there was a great amount of profit potential with bonds. Also, in the
1970s the Macauley duration measure was re-discovered.
Not all investors viewed the rise in interest rate volatility as a good
thing. The pension fund and insurance companies that invest in bond
found their job much more difficult. Thus, strategies based on duration
were developed to aid pension fund managers to match their liabilities
with properly constructed bond portfolios.
Passive Bond Portfolio Strategies
There are two major passive strategies:

• Buy-and-hold
• Indexing

Buy-and-hold Strategy
This strategy simply involves buying a bond and holding it until
maturity. Bond investors would examine such factors as quality
ratings, coupon levels, terms to maturity, call features and sinking
funds. These investors do not trade actively to earn returns, rather
they look for bonds with maturities or durations that match their
investment horizon.
There is also a modified buy-and-hold strategy in which investors buy
bonds with the intention of holding them until maturity, but they still
actively look for opportunities to trade into more desirable positions.
[However, if you modify this too much it turns into an active strategy.]
While the buy-and-hold strategy is a passive strategy, it still involves a
great deal of work. Agency issues typically provide high quality bonds
at a higher return than Treasury securities, callability affects the
attractiveness of an issue, etc. Plus, you may want to develop a
portfolio in which coupon payments are structured (and principal
repayments).
Techniques, Vehicles and Costs: Only default-free or very high
quality securities should be held. Also, those securities that are
callable by firm (allows the issuer to buy back the bond at a particular
price and time) or putable by holder (allows bondholder to sell the
bond to issuer at a specified price and time) will introduce alterations
in the firm's cash flows, and probably should not be included in the

6
buy-and-hold strategy. Also, those investors seeking to lock in a rate
of return may choose a zero-coupon bond--good strategy for college
tuition or retirement. The buy-and-hold strategy minimizes transaction
costs and, if implemented astutely, can be highly productive. For
example, if interest rates are currently high and are expected to
remain so for an extended period of time, the buy-and-hold strategy
will do well.
Indexing
Indexing involves attempting to build a portfolio that will match the
performance of a selected bond portfolio index, such as the Shearson
Lehman Hutton Government/Corporate Bond Index, Merrill Lynch
Index, etc. This portfolio manager is judged on his ability to track the
index.
Techniques, Vehicles and Costs: The fixed income market is
broader (in terms of security types) than the equity market. Also, even
though the Shearson Lehman Hutton Corporate Bond Index has over
4,000 securities, it only represents high quality corporate bond issues.
Thus, a compromise must be made when selected among different
indexes. Also, the strategy of buying every bond in a market index
according to its weight in the index is not a practical one. However, a
relevant subset is possible. We may choose to emulate a narrower
bond index.
Alternative Vehicles: We may choose to randomly select bonds from
the universe of bonds, or, we may choose the stratified approach
(segmenting the index into components from which individual
securities are chosen). When choosing the indexing option, bond
portfolio management cannot be considered entirely passive. Also,
there will be transaction costs associated with (1) purchasing the
issues used to construct the index; and (2) reinvesting cash payments
from coupon and principal repayments; and (3) rebalancing of portfolio
if the composition of your target index changes. Whereas full
replication of the target index would work best, this is impractical. If
you choose the stratified method, your performance will probably not
mirror your target index.
How many securities should you have in your portfolio if you use the
random sampling approach? McEnally and Boardman (1979) have
found that, once an index is selected, close replication is possible with
perhaps 40 bonds (for the long term).
Stratified Approach: Consists of analyzing the index to determine
various stratification levels (what portion of securities that make up
index are Treasury, Aaa Industrial, Baa Financial, of X years to
maturity, of X% coupon rate, etc.). The next step is to select the
securities for your portfolio. Typically, at selection and at the
rebalancing period (usually once a month) one security is chosen from

7
each category (there could be 40 categories). There's no requirement
as to which security is selected from each class.
Active Management Strategies
These strategies require major adjustments to portfolios, trading to
take advantage of interest rate fluctuations, etc. There are five major
active bond portfolio management strategies:

1. Interest rate anticipation

2. Valuation analysis
3. Credit analysis
4. Yield spread analysis
5. bond swaps

In each strategy, the manager hops to outperform the buy-and-hold

policy by using acumen, skill, etc.
Interest Rate Anticipation
This is the riskiest strategy because the investor must act on uncertain
forecasts of future interest rates. The strategy is designed to preserve
capital (lose as little as possible) when interest rates rise (and bond
prices drop) and to receive as much capital appreciation as possible
when interest rates drop (and bond prices rise).
These objectives can be obtained by altering the maturity or duration
of their portfolios. Longer maturity, or longer duration, portfolios will
benefit the most from an interest rate decrease and vice versa. Thus,
if a manager expects an increase in interest rates, they would
structure portfolio to have the lowest possible duration.
The problem faced with this type of strategy is the risk of mis-
estimating interest rate movements. It is difficult (EXTREMELY) to
predict (with accuracy) interest rate movements.
However, if this is your strategy, you should be concerned with:

• direction of the change in interest rates

• the magnitude of the change across maturities,
and
• the timing of the change.

How your bond will be affected by changes in interest rates can usually
be directly related to the security's duration. Thus, if you expect IR to
drop, you should shift to high duration securities. Also, the timing as to
when you expect the interest rate shift is important. You don't want to
shift too early, because you may compromise some return. Obviously,
you don't want to wait too late.
Scenario Analysis: Say, "what if" interest rates rise/fall by this much
over the next month/year/etc. Analyze the individual bonds within

8
your portfolio under each scenario and see how the returns are
affected under each scenario. [See p. 8-30] The scenario analysis
leads us to further analysis.
Relative Return Value Analysis: We can calculate the overall
expected return for each bond in our scenario (expected return under
each interest rate scenario weighted by the probability of that scenario
occurring) and the current duration of each bond in our portfolio and
graph the relationship. Those bonds falling above a regression line
(showing the general relationship) would be doing ok!
Strategic Frontier Analysis: We can graph the bonds in our portfolio
with the best case scenario (an interest rate decrease) on the vertical
axis and the worst case scenario on the horizontal axis, as shown
below:

Those securities which fall into Quadrant I represent aggressive

securities--if the best case happens, they will do well; however if the
worst case happens they will be the worst performers. Those securities
falling into Quadrant II are superior securities--they will perform well
regardless of which scenario occurs. Quadrant III represents defensive
securities--they will do well under the worst case scenario, but perform
poorly if the best case occurs. Quadrant IV securities are inferior as

9
they will perform poorly regardless of the scenario. You should sell
securities falling into Quadrant IV. Normally a few securities would fall
into Quadrants II and IV, with most falling into Quadrants I and III.
Valuation Analysis
The portfolio manager looks for undervalued bonds--those bonds that
have a computed value (according to the portfolio manager) higher
than the current market price). This also translates to those bonds
whose expected YTM is lower than the current YTM. This strategy
requires lots of analysis (continuous evaluations) and lots of trading
based on the analysis. Based on your confidence in your analysis, you
would buy undervalued bonds and sell overvalued bonds (or ignore
them if they are not in your portfolio).
Valuation Analysis: We can examine the term structure of pure
discount bonds (zero coupon) and thus determine the value of US
Treasuries, thus we can determine the default free characteristics of
any other type of bond. Then we can attempt to determine the other
factors that will affect bond yield by using multiple factor regression
analysis (looking at things such as: quality rating, coupon effect,
sector effect, call provision, sinking fund attributes, etc.) Using this
factor analysis, we can determine the expected yield for the security (if
the expected yield < current YTM then buy). However, there is some
subjectivity in factor analysis. For instance, if there is some "event
risk" (something affecting the financial stability of firm) missing from
the analysis, or if there is any anticipation of a market upgrade...
Credit Analysis
Credit analysis involves examining bond issuers to determine if any
changes in the firm's default risk can be identified. We try to
determine if the bond rating agencies are going to change the firm's
rating. Rating changes are prompted by internal changes within the
firm as well as external changes. Various factors examined include
financial ratios, GNP, inflation, etc.
Many more downgradings occur during economic contractions
[However from 1985-1990 downgradings increased substantially
despite an economic expansion.]. To be successful in utilizing bond
rating changes, you must accurately predict when the bond rating
change will occur and take action prior to the change. The market does
react to unexpected bond rating changes, however, it reacts quickly.
Credit Analysis of High-Yield Junk Bonds. Junk bonds have a wide
spread over bonds rated BBB and higher. Also, these yield spreads
widen over time (during poor economic times the spread widens).
Altman and Nammacher point out that the net return of junk bonds
(average gross return minus losses from bonds that defaulted) has
been superior to higher-rated debt [Of course, they're of higher risk.]
Other points to note: Even though the rating categories have not

10
changed, the quality of bonds today that fall into, let's say, the A
category, has lessened over time. "Specifically, the average values of
the financial ratios that determine whether bonds are included in the B
or CCC rating classes have declined over time."
Thus, bond portfolio managers will have to involved themselves in
detailed credit analysis to determine those bonds that will not default.
Credit Analysis: The assessment of default risk. Default risk has both
systematic and unsystematic elements. First, individual bond issuers
may experience difficulty in meeting their debt obligations. This could
be an isolated incident, and can be diversified away (or eliminated by
effective credit analysis). However, if default risk is precipitated by
adverse general business conditions, then this would require more
macro-oriented analysis. Many fixed-income investors complement the
bond ratings providing by bond agencies (Fitch's, Moody's, Duff &
Phelps, S&P's) with their own credit analysis, citing reasons such as:
more accurate, comprehensive, and timely analyses and
recommendations.
Yield Spread Analysis
A portfolio manager would monitor the yield relationships between
various types of bonds and look for abnormalities. If a spread were
thought to be abnormally high, you would trade to take advantage of a
return to a normal spread. Thus, you need to know what the "normal"
spread is, and you need the liquidity to make trades quickly to take
advantage of temporary spread abnormalities.
Spread Analysis: Involves anticipating changes in sectoral
relationships. For example, prices and yields on lower investment
grade bonds tend to move together (identifiable classes of securities
are referred to as sectors). Changes in relative yields (or the spread)
may occur due to:

• altered perceptions of the creditworthiness of a

sector of the market's sensitivity to default risk
• changes in the market's valuation of some
attribute or characteristic of the securities in the
sector (such as a zero coupon feature); or
• changes in supply/demand conditions.

The objective is to invest in the sector or sectors that will display the
strongest relative price movements. Brokerage firms maintain
historical records of yield spreads and are able to conduct specialized
analyses for clients, such as measurement of the historical average,
maximum, and minimum spread among sectors.
Potential drawbacks of this method include the need to make
numerous trades, the possibility of poor timing (how long will it take

11
for the market to realize the abnormal spread), and the danger that
overall changes in interest rates will dwarf these efforts.

12