May 2023
Jatin Misra, Ph.D., CFA
Co-Head of Structured Credit
Prepayment strategies:
Essentials, analysis, and investing
Executive summary
Prepayment risk is one of the four major risks present in
bond markets, along with term structure (or duration) risk,
credit risk, and liquidity risk. We believe prepayment risk,
largely concentrated in U.S. markets, is one of the least
understood and most underutilized risks in a well-diversified
portfolio. The risk is significant enough to impact the
domestic U.S. volatility surface and rates market, as well as
potentially impact global developed market rates.
The intent of this paper is to help institutional investors
better understand prepayment opportunities and how
best to incorporate prepayment securities into a broader
fixed income allocation. The paper outlines the history
and drivers behind the prepayment markets and delves
into distinct security types and their attributes, potential
opportunities, and how prepayment strategies may be
incorporated into a broader portfolio. Following is a
summary of four key topics we cover.
Refinancing drivers. The starting point for uncovering
value in prepayment markets is understanding how and
why borrowers prepay their mortgages. The majority
of U.S. homeowners have the option to prepay their
mortgage at any time without penalty to take advantage
of macroeconomic and mortgage industry developments
for their financial benefit. The primary drivers of
homeowner prepayments include a reduction in interest
rate, home price appreciation, and government housing
policies that impact the aordability and availability of
mortgage financing.
Tranching and prepayment speeds. The securities used
to pursue prepayment strategies are created based on
the cash flows of a pool of mortgage loans. Mortgage
principal and interest payments can be restructured into
a variety of collateralized mortgage obligation (CMO)
tranches. Redistributing prepayment cash flows may be
categorized as time tranching and may further be divided
into contraction and extension risks.
Jo Anne Ferullo, CFA
Portfolio Manager
For use with institutional investors and investment professionals only.
2
May 2023 | Prepayment strategies: Essentials, analysis, and investing
For use with institutional investors and investment professionals only.
Agency IOs. Interest-only (IO) securities are the purest
expression of prepayment risk. They tend to increase in
value in rising interest-rate environments and generally
decrease in value when interest rates are falling or
remain very low. While IOs may exhibit a high degree
of volatility, they are demonstrable risk mitigants
within a broader portfolio when interest-rate risk is
hedged. Duration hedging also allows for the capture of
prepayment risk premium and an attractive risk-adjusted
return. Additionally, we believe agency IOs to be a good
diversifier within a broader portfolio given their low
correlation to other asset classes, including other types
of structured credit.
Identifying opportunities. To compare valuations and
characteristics of securities in the prepayment markets,
it is critical to have a thorough grasp of the key influential
factors for estimating their path-dependent cash flows.
The primary factor is how in or out of the money the
underlying loans of a mortgage pool are when compared
with interest rates. The market generally allows a risk
premium for securities that are most susceptible to
errors in commonly used prepayment valuation models.
We find the pricing of this risk premium to be inconsistent,
which creates opportunities for sophisticated investors
to buy and sell securities with ineicient pricing.
Successful investing in prepayment strategies requires
a thorough understanding of the mortgage market,
underlying collateral characteristics and behavior,
and cash flow structuring by an experienced team
adept at identifying and extracting value based on
borrower behavior.
Prepayment essentials
Prepayment risk is one of the four major risks present in
bond markets, along with term structure (or duration)
risk, credit risk, and liquidity risk. Of these four return
drivers, prepayment risk is one of the least understood
and, consequently, one of the most underutilized
risks in a well-diversified portfolio.
At the highest level, prepayment securities and strategies
are based on the timing of cash flows — the receipt by
the investor of a payment either earlier or later than
anticipated. At its heart, this risk is similar to call risk in
the government or corporate credit markets. It manifests
itself in a more developed form in agency mortgage-
backed securities (MBS). Being guaranteed by Fannie
Mae, Freddie Mac, or Ginnie Mae (collectively, the
government-sponsored enterprises, or GSEs), agency
MBS are insulated against exposure to the credit of the
underlying borrowers but are fully exposed to the U.S.
borrower’s unique option to pay down or refinance their
mortgage at any time, without penalty. As a result,
prepayment strategies are constructed around the ability
of the investor to predict the future timing of cash flows.
In the context of providing a framework for considering
prepayment strategies, we focus on agency IO bonds.
Following we provide an overview of the securitization
process that creates agency MBS pools, the structuring of
these pools to create IOs, and the performance dynamics
of IO securities, as well as a conceptual framework for
analyzing and investing in these securities with an eye
toward their use within a broader portfolio framework.
The U.S. mortgage market
Prepayment securities are created based on the cash
flows of the underlying mortgage loans. If the investor can
reliably estimate the cash flows of the underlying mortgage
loans for the valuation and analysis of a mortgage pool
security, the cash flow projection and the same valuation
and analysis framework may be applied to mortgage
structures that direct those cash flows in a variety of ways.
The fundamentals of estimating cash flows and option-
adjusted analysis are key to analyzing prepayment
strategies and are applicable to investing in agency IOs.
We provide an overview of these essential concepts as a
foundation for exploring interest-only securities.
3
May 2023 | Prepayment strategies: Essentials, analysis, and investing
For use with institutional investors and investment professionals only.
History of the Federal Housing Administration
The FHA was established in 1934 and became
part of the U.S. Department of Housing and Urban
Development’s (HUD) Oice of Housing in 1965.
1
Part
of the FHA’s mission is to provide aordable mortgage
insurance on loans made by FHA-approved lenders.
The insurance protects the lender against default
by the borrower, and the FHA strictly dictates the
required characteristics of the loans it insures. One of
the requirements instituted early in the life of the FHA
was that loans must have constant, or fixed, payments
and be directly amortizing, meaning that each monthly
principal payment must be applied directly to the
amortization of the loan. This was not particularly
common leading up to the Great Depression. Because
loans at that time were not directly amortizing, payo
was random and not tied to a particular maturity
date per se, but most were paid o in roughly 11 to 12
years. Commercial banks were also not encouraged
to extend long-term financing.
2
This changed when
the 30-year loan term was approved for the FHA by
Congress in 1948
3
for loans on newly constructed
homes and in 1954 for those on existing homes.
4
The
FHA works alongside the GSEs in providing guarantees
on loans, supporting the ongoing extension of credit to
homebuyers (for more information, see pages 6–7).
1 https://www.hud.gov/program_oices/housing/fhahistory (accessed 12/21/2021).
2
https://www.atlantafed.org/blogs/real-estate-research/2012/04/05/debunking-a-popular-myth-about-mortgage-lending (accessed 12/20/2021).
3 Congressional Research Service, A Chronology of Housing Legislation, 1892–2003, 2004 (accessed 12/21/2021).
4
https://www.aei.org/economics/housing-finance/housing-finance-fact-or-fiction-fha-pioneered-the-30-year-fixed-rate-mortgage-during-the-great-
depression/ (accessed 12/20/2021).
Creating a mortgage pool
A mortgage loan is established when a borrower receives
credit from a lending institution, most oen from a bank
or via an intermediary mortgage broker who works with
multiple lending institutions, in order to purchase a home.
Mortgage loans may have fixed or floating interest rates,
dierent maturities, or other diering characteristics,
but the most common is the 30-year loan with a fixed
monthly payment. The Federal Housing Administration
(FHA) played an important role in encouraging the use of
the fixed-payment, 30-year loan.
The originator of the loan underwrites, or assesses, the
borrower’s ability to repay the loan based on a variety
of characteristics, including the borrower’s income and
credit history as well as the amount of the loan versus
the value of the home, among others. These risk
characteristics determine the interest rate the borrower
pays, typically the combination of a spread over the
10-year Treasury yield plus a risk charge based on the
possibility the borrower may default. Additional fees are
assessed, including fees for Fannie Mae, Freddie Mac, or
Ginnie Mae; and fees for providing a guarantee as well as
for servicing. The last of these, servicing, entails keeping
track of and recording principal and interest payments,
ensuring payments are passed through to investors,
managing delinquencies and defaults, and so forth.
While the payments on a fixed-rate mortgage remain
the same in total each month, the mix of principal and
interest changes with more interest paid earlier in the life
of the loan and declining over time, while the principal
payment makes up the dierence and increases over time
(see the illustration for calculating mortgage payments).
4
May 2023 | Prepayment strategies: Essentials, analysis, and investing
For use with institutional investors and investment professionals only.
Calculating monthly mortgage payments and abbreviated amortization schedule
Amortization schedule: Interest payments decline and principal payments increase over time
Months
Beginning
loan balance
Monthly
payment
Interest
payment
Guarantee
and servicing fees
Principal
payment
Ending loan
balance
1 $1,000,000.00 $4,774.15 $2,916.67 $416.67 $1,440.82 $998,559.18
2 $998,559.18 $4,774.15 $2,912.46 $416.07 $1,445.62 $997,113.56
3 $997,113.56 $4,774.15 $2,908.25 $415.46 $1,450.44 $995,663.13
4 $995,663.13 $4,774.15 $2,904.02 $414.86 $1,455.27 $994,207.85
65 $897,403.14 $4,774.15 $2,617.43 $373.92 $1,782.81 $895,620.33
66 $895,620.33 $4,774.15 $2,612.23 $373.18 $1,788.75 $893,831.58
67 $893,831.58 $4,774.15 $2,607.01 $372.43 $1,794.71 $892,036.87
68 $892,036.87 $4,774.15 $2,601.77 $371.68 $1,800.69 $890,236.18
Note: Typically, the guarantee and servicing fees are incorporated in the total interest rate charged on the loan but are shown separately in the
example for the sake of clarity.
Source: Putnam.
5 Frank J. Fabozzi, The Handbook of Mortgage Securities, 7th edition, p. 88.
Payment calculation formula
Monthly payment
5
= MP = MB
0
* [ i (1 + i)
n
]
[ (1 + i )
n
– 1 ]
Where: i = loan interest rate / 12
n = # of months in the term of the loan = 12 months * 30 years = 360 months
MB
0
= Mortgage balance = $1,000,000
5
May 2023 | Prepayment strategies: Essentials, analysis, and investing
For use with institutional investors and investment professionals only.
FIGURE 1
Example A, part 1: Mortgage pool
GNMA
FNMA
FHLMC
Mortgage pool
Par: $100M
Coupon: 3.0%
Price: $106.82
OAD: 5.73 years
DV01: 6.12
Investor objective
Benefit from
monthly cash flow,
higher yields
Principal and
interest payments
Par: $100M
Coupon: 3.0%
Price: $106.82
OAD: 5.73 years
DV01: 6.12
Residential mo
rtgage loans
Aer a loan is closed (when the terms are settled and
contracts are signed), the originator may sell the loan
to one of the GSEs, sell the loan to an aggregator that
buys mortgage loans, or pool similar loans together on
its balance sheet and sell the pool to broker/dealers who
then sell these pools on to end investors (Figure 1).
The three GSE agencies generate profit by borrowing
at lower rates (because federal backing is assumed)
and then investing in loans and mortgage securities
purchased from lenders/originators. The loans and
mortgage securities are then packaged into pools of loans
and sold to investors at a profit. The agencies also receive
guarantee fees to assume the credit risk of the borrowers,
which are paid by the investor when the pooled loans
are sold into the market. The investor is willing to pay
the fee to avoid the credit risk and rely on the agency,
rather than the borrower, to make the timely payment of
principal and interest. To mitigate the credit exposure to
the borrower, the agencies have strict requirements each
borrower and loan must meet.
Source: Putnam.
For illustrative purposes only. The above example is not intended to be relied upon as a forecast, research, or investment advice, and is not
a recommendation, oer, or solicitation to buy or sell any securities or to adopt any investment strategy. As with any investment, there is a
potential for profit as well as the possibility of loss.
6
May 2023 | Prepayment strategies: Essentials, analysis, and investing
For use with institutional investors and investment professionals only.
The roles of U.S. housing agencies
Government-sponsored enterprises (GSEs)
The Federal National Mortgage Association (Fannie Mae)
is the oldest of the agencies and was established in 1938
by an amendment to the National Housing Act of 1934 in
order to help solve some of the housing finance issues that
arose during the Great Depression. Fannie Mae’s original
mandate allowed it to buy FHA-insured loans from private
lenders and later, in 1948, loans from the U.S. Department
of Veterans Aairs’ (VA) Home Loan Program for Veterans.
As part of the Housing and Urban Development Act of 1968,
Congress reorganized Fannie Mae, splitting it in two.
Fannie Mae was reorganized as a private, for-profit
company focused on conventional (non-U.S. government-
insured or government-guaranteed) mortgages that
conform to specific underwriting standards. At the same
time, the Government National Mortgage Association
(Ginnie Mae, or GNMA) was also created. Ginnie Mae
remains a direct government agency within HUD, earning
a fee for guaranteeing privately issued mortgage-backed
securities collateralized by government-insured or
guaranteed loans. Freddie Mac was established in 1970
as a government-chartered corporation, owned by the
11 Federal Home Loan Banks, the system’s Oice of
Finance, and the federally insured savings institutions,
which owned stock in the Federal Home Loan Banks.
As part of the Financial Institutions Reform, Recovery,
and Enforcement Act (FIRREA) in 1989, Freddie Mac was
rechartered as a private corporation in the manner of
Fannie Mae. Currently, both Fannie Mae and Freddie Mac
are in conservatorship with the federal government as
a consequence of the bailout required during the global
financial crisis (GFC).
6
GSEs’ role in the U.S. residential housing market
Ginnie Mae is solely focused on guaranteeing the
principal and interest of mortgage-backed securities
backed by pools of loans that meet the standards of
several government lending programs. The loans are
backed by the FHA, HUD, the VA Home Loan Program, the
U.S. Department of Agriculture’s (USDA) Rural Development
Housing and Community Facilities Programs and Rural
Development Guaranteed Rural Rental Housing Program,
and the Oice of Public and Indian Housing programs.
Putting the full faith and credit of the U.S. government
behind the government-sponsored programs supports the
available supply of mortgage financing and lowers the cost.
7
The role of Fannie Mae and Freddie Mac is primarily to
provide liquidity to the American mortgage-lending
system, focused on low- to middle-income earners.
Specifically, they are federally mandated to maintain
stability in the secondary market for residential mortgages,
increase the liquidity of mortgage investments, and make
more money available for residential mortgage financing.
By purchasing loans from originators, assuming the loans
meet the agencies’ strict requirements for size, credit, and
underwriting standards, they free up the originators’ capital
to make additional loans. In addition, the two agencies
guarantee the timely payment of principal and interest on
the loans to outside investors.
The lender, or originator, is incented to oer the long-term,
fixed-payment mortgages because they know Fannie Mae
and Freddie Mac will most likely purchase them, freeing
the lender to create new loans. Investors are also incented
to purchase the pooled loans because they do not have to
worry about the credit risk of the borrower. Fannie Mae and
Freddie Mac guarantee the principal and interest payments.
8
6 https://www.fhfaoig.gov/Content/Files/History%20of%20the%20Government%20Sponsored%20Enterprises.pdf (accessed 12/21/2021).
7 https://www.ginniemae.gov/about_us/who_we_are/Pages/our_history.aspx (accessed 12/21/2021).
8 Center for American Progress, “7 Things You Need to Know About Fannie Mae and Freddie Mac,” September 6, 2012,
https://americanprogress.org/article/7-things-you-need-to-know-about-fannie-mae-and-freddie-mac/ (accessed 12/1/2021).
7
May 2023 | Prepayment strategies: Essentials, analysis, and investing
For use with institutional investors and investment professionals only.
To be clear, Fannie Mae and Freddie Mac are the guarantors,
not the U.S. government. Because the two agencies are
federally mandated to provide liquidity and stability to
the residential mortgage market, there is an implicit, not
explicit, guarantee by the U.S. government. The secretary
of the Treasury is authorized to buy up to $2.25 billion of
securities from each company to support its liquidity, and
both agencies are generally exempt from state and local
taxes. The Federal Housing Finance Agency (FHFA) was
established by the Housing and Economic Recovery Act of
2008 (HERA) and is responsible for the eective supervision,
regulation, and housing mission oversight of Fannie Mae,
Freddie Mac, and the Federal Home Loan Bank system.
Since 2008, FHFA has also served as conservator of Fannie
Mae and Freddie Mac. It is the responsibility of FHFA to
regulate, enforce, and monitor Fannie Mae and Freddie
Mac’s capital standards and limit the size of their mortgage
investment portfolios. HUD is responsible for monitoring
Fannie Mae and Freddie Mac’s general housing missions.
Estimating cash flows and prepayments
As with most fixed income instruments, the value of
a prepayment-sensitive security is determined by the
present value of its expected future cash flows. The notable
dierence is that the variability of cash flows is not driven by
the likelihood of default, but rather by the rate at which the
underlying borrowers prepay their mortgages. If a security’s
realized prepayment speeds exceed expectations, then the
security loses value, and vice versa. Understanding how
and why borrowers prepay their mortgages is imperative
to uncovering value in prepayment markets.
Prepayments represent “unscheduled” principal payments,
or those principal payments received faster than expected
in the amortization schedule. The most common metric
used to measure prepayment speeds is the conditional
prepayment rate, which is the percentage of principal
expected to be received ahead of schedule per year.
However, borrowers make mortgage payments on a
monthly basis and, as such, prepayment rates are typically
reported each month. These prepayments are reported
as the single monthly mortality rate (SMM), which is the
amount of principal paid early during the month divided
by the expected end of month principal balance.
9
The SMM
is then used to calculate the conditional prepayment rate,
which provides investors with a forward-looking estimate
of a mortgage pool’s prepayment rate, and can be applied
to project the pool’s future cash flows (see illustration for
calculation formula and example).
9 dv01.freshdesk.com/support/solutions/articles/42000052036-cpr-calculation (accessed 9/24/2021).
8
May 2023 | Prepayment strategies: Essentials, analysis, and investing
For use with institutional investors and investment professionals only.
Calculating conditional prepayment rate (CPR) with single monthly mortality rate (SMM)
SMM = Unscheduled principal received
Beginning loan balance – Scheduled principal payment
The SMM is annualized to calculate the CPR:
CPR = 1 – (1 – SMM)
12
Where, MB
0
= Mortgage pool balance = $100,000,000
SMM = Given
i = loan interest rate / 12 = 3.0% / 12
n = # of months in the term of the loan = 12 months * 30 years = 360 months
Monthly payment
5
= MB
0
* [ i (1 + i)
n
]
[ (1 + i )
n
– 1 ]
However, CPR has its limitations. Most notably, it assumes
a constant rate of prepayments based upon the most
recent month’s data. In actuality, the prepayment
rate will change over time as factors such as the
macroeconomic environment and government policy
evolve. Understanding how these external factors impact
prepayment speeds is crucial.
Due to the borrowers’ ability to prepay their loan in part
or in full at any time without penalty, prepayments occur
for a number of reasons (see “Analyzing refinancing
behavior” on p. 20). The primary driver of prepayments is
refinancing the loan for a lower interest rate — a reaction
to changes in the interest-rate environment. This behavior
means that the monthly payments from the pool are not
fixed or determinable. The pool cash flows change based
on the borrowers’ reaction to interest rates; as such,
probabilistic methods such as Monte Carlo simulations
are used to estimate the monthly cash flows and the
borrowers’ propensity to prepay depending on the loan
and borrower characteristics, as well as the historical
experience of the underlying loans in dierent interest-
rate environments.
Sample mortgage pool cash flows, including CPR
Months
Beginning
loan balance SMM
Monthly
payment
Interest
payment
Scheduled
principal Prepayment Total principal Total cash flow
1 $100,000,000.00 0.00095 $1,520,055.93 $250,000.00 $1,270,055.93 $93,793.45 $1,363,849.38 $89,636,150.62
2 $98,636,150.62 0.00123 $1,520,055.93 $246,590.38 $1,273,465.55 $119,756.10 $1,393,221.66 $97,242,928.97
3 $97,242,928.97 0.00155 $1,520,055.93 $243,107.32 $1,276,948.61 $148,747.27 $1,425,695.88 $95,817,233.09
4 $95,817,233.09 0.00182 $1,520,055.93 $239,543.08 $1,280,512.85 $172,056.83 $1,452,569.68 $94,364,663.41
5 $94,364,663.41 0.00211 $1,520,055.93 $235,911.66 $1,284,144.27 $196,399.90 $1,480,544.17 $92,884,119.25
6 $92,884,119.25 0.00247 $1,520,055.93 $232,210.30 $1,287,845.63 $226,242.80 $1,514,088.43 $91,370,030.82
9
May 2023 | Prepayment strategies: Essentials, analysis, and investing
For use with institutional investors and investment professionals only.
Valuation and adjusting for the eects of interest
rates and prepayment behavior
Option-adjusted spread (OAS) analysis is used to value
debt securities with diering cash flow patterns and has
two distinct dierences from traditional bond valuation.
The first is that it does not employ a constant discount
rate to determine the present value of the security’s cash
flows but uses discount rates based on the term structure
of interest rates. The second is that interest rates are
not assumed to remain constant over time, but multiple
interest-rate paths are used to evaluate the bond over
its life.
A probabilistic model (e.g., Monte Carlo simulation) is
used to generate a large number of interest-rate paths
based on the current term structure of interest rates
as well as historical interest-rate behavior. For each
individual interest-rate path, cash flows are estimated and
discounted. The discount rates are based on the swap
rates associated with that particular path plus a spread.
The spread that makes the average price across all paths
equal the market price is the OAS. Most importantly, the
OAS generated by the model can be used to assess the
relative prepayment risk of dierent mortgage securities.
Consider OAS a representation of the incremental yield
over swaps/Treasuries aer adjusting for variations in
interest rates and prepayment behavior on cash flows.
Duration and convexity for interest rate-contingent
cash flows
Neither Macaulay nor modified duration are accurate
measures of price sensitivity to changes in yield for
prepayment securities. Macaulay duration measures
how long to hold a bond so that the present value of the
cash flows received is equal to the current market price;
it is eectively the time horizon of receiving cash flows.
Modified duration measures a bond’s price sensitivity to
changes in yields but does not account for the fact that
the cash flows themselves can change due to changes in
interest rates. To incorporate the optionality associated
with the cash flows, option-adjusted duration (OAD, also
eective duration) is a better alternative. The OAD is also
useful in calculating the DV01, or dollar duration, of the
mortgage pool security.
The process to calculate OAD is relatively simple; the
OAS is assumed to remain constant, the initial yield of
the short-term Treasury is shied up and down, and new
estimated prices for the security are calculated. The new
prices oer a means of estimating convexity (option-
adjusted, or eective, convexity). Convexity measures
the rate of change in duration, or the trend of acceleration
or deceleration in a bond’s price sensitivity to changes
in yield. The linear approximation of duration breaks
down for large changes in yield, making convexity an
important measure in evaluating a mortgage pool.
When interest rates rise, bond prices fall. In this case, a
bond with positive convexity does not lose as much value
as an option-free bond. However, a bond with negative
convexity does not gain as much in value when interest
rates fall.
10
May 2023 | Prepayment strategies: Essentials, analysis, and investing
For use with institutional investors and investment professionals only.
Calculating option-adjusted duration (OAD)
Estimating convexity
10 TreasuryDirect, U.S. Department of the Treasury, Bureau of Fiscal Service, “Timeline of Separate Trading of Registered Interest and Principal
Securities (STRIPS),” March 10, 2011, https://www.treasurydirect.gov/indiv/research/history/histtime/histtime_strips.htm (accessed 9/24/2021).
11 Elizabeth Balboa, “This Day In Market History: Fannie Mae Rolls Out First Stripped Mortgage-Backed Securities,” Benzinga, July 11, 2019,
https://www.benzinga.com/general/education/18/07/12004681/this-day-in-market-history-fannie-mae-rolls-out-first-stripped-mort
(accessed 9/24/2021).
12 Bloomberg U.S. Aggregate Bond Index, as of 12/31/22.
13 Putnam Investments.
OAD = V
–
– V
+
2V
o
∆r
Where, V
o
= Current price
V
–
= Price aer a shi down in Treasury yield
V
+
= Price aer a shi up in Treasury yield
∆r = Change in Treasury yield
Option-adjusted convexity approximation = V
+
– 2V
o
+ V
–
2V
o
∆r
2
Typically, mortgage securities demonstrate negative
convexity. The eect is that upside price potential is lower
than that of an option-free bond. This is a cost of investing
in mortgage securities.
Having a thorough grasp of the key influencing factors
for estimating the path-dependent cash flows is critical
for analyzing and comparing security valuation and
characteristics. The primary factor is how in or out of the
money the underlying loans of the pool are compared
with interest rates. The results of the cash flow estimation
impact all the subsequent calculations. But once the
investor is comfortable with the cash flows, the investor
can use them to analyze the various structures derived
from the mortgage pool.
Interest-only securities
The U.S. Treasury conducted the first modern issuance
of IO and principal-only (PO) Treasury STRIPS (Separate
Trading of Registered Interest and Principal of Securities)
in 1985.
10
This paved the way for creating other forms
of stripped securities issued by the GSEs, who followed
closely behind beginning in 1986.
11
Since then, the agency
IO market has grown to approximately $100–$200 billion,
a relatively small — but powerful — slice of the U.S.
residential agency mortgage market, which weighs in at
$6.8 trillion.
12
To put the agency IO market into context
with other opportunities, the private, or non-agency,
residential mortgage market is approximately $600–$700
billion, while the commercial mortgage market is slightly
smaller at $500–$600 billion.
13
11
May 2023 | Prepayment strategies: Essentials, analysis, and investing
For use with institutional investors and investment professionals only.
Allocating prepayment risk
The holder of an agency mortgage pool receives their
pro rata share of both principal and interest. These
principal and interest payments can be restructured into
a variety of collateralized mortgage obligation tranches.
CMO structures are intended to match the objectives
of investors with varying tolerance for prepayment risk.
Prepayment risk is not eliminated by redirecting it; it
simply means that some tranches carry more prepayment
risk and others, less. It is merely divided up; for those
investors willing to assume less prepayment risk, other
investors need to be willing to take on more.
The redistribution of the principal and interest cash flows
may be broadly categorized as time tranching or credit
tranching. Time tranching creates structures with dierent
maturity profiles and divides up prepayment risk. Interest
and principal cash flows are directed to dierent classes
of the CMO in order to satisfy dierent investor desires
regarding time to maturity or the degree of prepayment
risk assumed. In the examples of creating an IO or an
inverse IO, the interest cash flow is directed to the IO and
the principal cash flow to the PO or other CMO tranche.
Credit tranching directs credit losses (default risk) to
specified tranches in a similar manner; however, because
agency CMO securities carry either an explicit or implicit
government guarantee, there is no credit tranching.
Prepayment risk, part of time tranching, may be divided
into contraction and extension risk. Contraction risk
occurs when prepayments happen more quickly than
anticipated due to a decline in interest rates. The adverse
results of contraction are twofold: the investor will be
reinvesting cash flows at lower interest rates and the
security’s inherent negative convexity diminishes the
associated upside price potential. Extension risk is the
opposite; interest rates rise and prepayments slow more
than anticipated. The investor has less cash flow to
reinvest at higher rates, and the downside price potential
is exacerbated by negative convexity — by prepayments
coming in slower than anticipated, there is more principal
outstanding invested at a lower-than-market interest rate.
In constructing CMO tranches with dierent prepayment
risk profiles, contraction risk is allocated to certain
tranches and extension risk to others. Scheduled
principal and prepayments are directed first to the
shortest maturity, then the next shortest, and so on.
The shortest tranche also has the lowest interest rate —
the principal balance of the tranche is outstanding for the
shortest amount of time so it generates less interest. The
tranche outstanding for longer will have a higher interest
rate. The CMO structures follow a strictly defined payment
order. The allocation of risk allows greater participation
in the mortgage-backed securities market by satisfying
varying investor risk tolerance. We focus on the CMO
structure most exposed to prepayment risk, namely
the forms of IO securities. An IO receives no principal
payments, only interest payments. This structure, with
IOs receiving only interest cash flow, concentrates the
eects of both forms of prepayment risk.
Creation of an IO
An IO is created when the mortgage pool is structured to
direct the interest component of the monthly payments
to an IO security and the principal component to a
principal-only security. PO investors, such as insurance
companies or banks, tend to appreciate more cash flow
stability with lower volatility versus the IO. Conversely, an
IO is subject to more cash flow uncertainty because the
cash flows consist solely of interest payments that are a
function of outstanding principal and are highly sensitive
to prepayments.
As shown in Figure 2, the structuring process
concentrates prepayment risk into the IO. The price
of the pool equals the price of the PO plus the price of
the IO ($106.82 = $85.32 + $21.50). The basis for the price
split between the PO and the IO is the pool cash flow
generated by the probabilistic model process described
earlier. The scheduled principal and prepayment cash
flows are directed to the PO, and the price is based on the
present value of the cash flows. A PO is essentially a
zero-coupon bond with principal repayment based on the
scheduled and prepaid principal. The price is higher
because there is more certainty associated with the
scheduled principal. The interest cash flow is directed to
the IO, and the price is based on the present value of the
flows. That price is much lower with greater uncertainty of
cash flow.
12
May 2023 | Prepayment strategies: Essentials, analysis, and investing
For use with institutional investors and investment professionals only.
FIGURE 2
Example A, part 2: Mortgage pool and IO/PO construction
GNMA
FNMA
FHLMC
Mortgage pool
Par: $100M
Coupon: 3.0%
Price: $106.82
OAD: 5.73 years
DV01: 6.12
Principal only (PO)
Par: $100M
Coupon: 0%
Price: $85.32
OAD: 11.02 years
DV01: 9.40
Investor objective
Benefit from
faster prepayments
(refinancing)
Benefit from
slower prepayments
(refinancing)
Interest only (IO)
Par: $100M
Coupon: 3.0%
Price: $21.50
OAD: -15.26 years
DV01: -3.28
Residential mo
rtgage loans
Source: Putnam.
For illustrative purposes only. The above example is not intended to be relied upon as a forecast, research, or investment advice, and is not
a recommendation, oer, or solicitation to buy or sell any securities or to adopt any investment strategy. As with any investment, there is a
potential for profit as well as the possibility of loss.
The mortgage pool is priced at $106.82 in Figure 2.
Principal payments are received at $100, par value.
For purposes of illustration, we can assume the pool is
composed of one loan. If prepayments come in faster
than anticipated, the pool owner may lose $6.82 in
premium. An IO, which receives no principal payments
whatsoever, is dependent on the outstanding principal
to generate interest. If prepayments come in faster
than anticipated and the principal balance is paid
down to zero, there is no interest for the IO, and the
full investment of $21.50 may be lost. Alternatively,
the same scenario is beneficial for the PO investor.
If they pay $85.32 for the PO security and prepayments
come in faster than anticipated, the PO investor may gain
$14.68 ($100 – $85.32 = $14.68). Another way to frame the
scenario: The sooner prepayments are received, the
greater the present value of the PO; while less interest is
generated, the price of the IO declines.
The combination of the IO/PO cash flows exactly replicates
the cash flow of the pool. The par amount of the PO equals
the par amount of the pool and is the principal received
over time. The IO interest is calculated based on the total
par amount of the pool, but the holder of the IO does not
receive any principal payments. As it relates to the
calculation of interest on the IO, the pool par amount may
be thought of as a notional reference value.
13
May 2023 | Prepayment strategies: Essentials, analysis, and investing
For use with institutional investors and investment professionals only.
Calculating the coupon of a mortgage pool
Pool coupon = IO coupon + PO coupon
3.00% * $100M = (3.00% * $100M) + (0% * $100M)
$3M = $3M + $0
OAD is also divided between the IO and PO securities. The
allocation of interest-rate risk is driven by the interplay of
prepayment cash flows and interest-rate movements. As
interest rates rise, that risk is borne nearly entirely by the
PO. Prepayments of principal slow dramatically, the PO
extends, and the average time until repayment of principal
is much longer. Coupled with a higher discount factor in the
discounted cash flow price calculation due to increasing
interest rates, the value of expected future principal
payments is also lower. These two factors result in a
substantial decrease in the present value of those
payments. This gives the PO security a duration
significantly longer than that of the original pool. The IO is
aected in the opposite manner. As interest rates fall, the
lower discount rate increases the value of the scheduled
interest payments; however, because the interest
payments decline/terminate when principal prepayments
increase, it lowers the value of the IO. Typically, the
termination of interest payments tends to dominate the
eect of lower discount rates and results in a negative
duration for the IO.
14
The dollar duration (DV01) of the IO/PO
pairing replicates the DV01 of the pool under the
assumption that all three (the pool, the IO, and the PO) are
run at the same OAS.
The point to note is the negative OAD and DV01 of the IO.
With this high degree of negative OAD and DV01, it is a
relatively volatile asset.
Calculating dollar duration (DV01) of a mortgage pool
Pool DV01 = IO DV01 + PO DV01
(Pool price * Pool OAD) = IO price * IO OAD + PO price * PO OAD
($1.0682 * 5.73) = ($0.2150 * –15.26) + ($0.8532 * 11.02)
6.12 = –3.28 + 9.40
Note the negative OAD and DV01 of the IO, –15.26 and –3.28, respectively.
14 Alan J. Marcus and Arnold Kling, NBER Working Paper Series #2340: Interest-Only/Principal-Only Mortgage-Backed Strips: A Valuation and Risk Analysis,
August 1987, https://www.nber.org/system/files/working_papers/w2340/w2340.pdf (accessed 11/9/2021).
14
May 2023 | Prepayment strategies: Essentials, analysis, and investing
For use with institutional investors and investment professionals only.
IO cash flows and performance
At the highest level, the coupon interest earned by an
IO drives the price of the security and is a function of
how long the associated principal balance remains
outstanding. The longer principal exists to generate
interest, which may be earned by the investor holding the
IO. Most interest-only strategies are negatively correlated to
interest rates, demonstrating negative duration as shown in
the earlier IO/PO example, and tend to outperform when
interest rates are rising. As interest rates rise and mortgage
rates move higher, there is less financial incentive for
the borrower to refinance or prepay their mortgage.
Prepayment speeds slow, principal remains outstanding
longer, more interest income is generated by the IO
security, and the price of the IO increases. This is in direct
contrast to other fixed income securities, whose prices
most oen fall as interest rates rise.
The price of an IO will generally fall in the opposite
scenario. When interest rates are falling or remain very
low, it creates an incentive for the borrower to refinance
at a lower mortgage rate. As more borrowers refinance,
prepayments increase, principal is reduced, the amount
of interest earned decreases, and the price of the security
falls. The IO is highly sensitive to prepayment speeds
because the investor only receives interest payments;
there are no principal payments. The investor loses the
future interest payments when the original loan is paid o
and the principal balance used to calculate the interest
payment goes to zero.
Beyond the characteristic of negative duration, IO
securities also display negative convexity. There is no
obligation for the borrower to refinance, creating an
asymmetry. If the economic incentive in a falling interest-
rate environment is great enough, it prompts the
borrower to act and refinance at a lower mortgage rate,
resulting in principal erosion (contraction) and the IO
investor losing future interest cash flows. However, there
is no financial incentive for the borrower to refinance in
a rising interest-rate environment. Principal remains
outstanding for longer (extension), and the IO generates
interest payments for a longer time. The degree of
negative convexity, or asymmetry, is influenced by the
backdrop of the existing mortgage market relative to
the current environment. From a price perspective, if the
security is priced at a discount, which typically occurs
when expectations are for reduced interest, such as in a
declining interest-rate environment, the asymmetry works
against the investor. The security cannot extend further
but may contract significantly if interest rates decline,
borrowers refinance, and principal erosion occurs.
Premium is the opposite scenario; as a premium-priced
security extends, it may add a fair degree of balance,
reducing asymmetry. The degree of convexity depends
on the coupon of the IO security relative to those that are
at the money, or in line with the current market. This is a
more nuanced eect than experienced in the corporate
bond market.
The above describes the general unhedged performance
characteristics of an IO. In a rising interest-rate
environment, the fixed-rate IO is capped at the existing
fixed coupon rate (Figure 3). The pace of receiving
unscheduled principal, or the prepayment speed,
determines the steepness or flatness in our valuation
example. The valuation charts (see Figures 3 and 6) are
simplified representations of the typical unhedged
experience.
15
May 2023 | Prepayment strategies: Essentials, analysis, and investing
For use with institutional investors and investment professionals only.
FIGURE 3
Valuation of fixed-rate IO shows negative convexity as price declines
with falling interest rates but remains flat above the IO coupon rate
Source: Putnam.
For illustrative purposes only. The above example is not intended to be relied upon as a forecast, research, or investment advice, and is not
a recommendation, oer, or solicitation to buy or sell any securities or to adopt any investment strategy. As with any investment, there is a
potential for profit as well as the possibility of loss.
IO coupon %
Principal erosion
and declining interest
Falling interest rates
Interest-rate change
Coupon
Rising interest rates
Fixed-rate
interest payments
6%
5
4
3
1
2
0
Creation of an IO: The case of the inverse IO
Another common example of IO construction is the inverse
IO (Figure 4). The inverse IO (IIO) is created by stripping a
floater from a fixed-rate pool. The coupon of the floater is
based on a reference rate plus a spread. The floater also
includes a cap on the coupon rate, a natural consequence
of the underlying pool having a fixed-rate coupon. Once the
floater is stripped from the pool, the resulting inverse IO
also has a coupon that floats, but it is an inverse floater and
moves in the opposite direction. The inverse IO also carries
an eective floor. The spread, cap, and floater on the two
tranches are coordinated so that the total interest cash flow
matches the interest cash flow of the underlying mortgage
pool. The scheduled principal and prepayments are
directed to the floater tranche, while the inverse IO only
receives its designated interest.
16
May 2023 | Prepayment strategies: Essentials, analysis, and investing
For use with institutional investors and investment professionals only.
FIGURE 4
Example B: Mortgage pool and inverse IO/floater construction
Source: Putnam.
For illustrative purposes only. The above example is not intended to be relied upon as a forecast, research, or investment advice, and is not
a recommendation, oer, or solicitation to buy or sell any securities or to adopt any investment strategy. As with any investment, there is a
potential for profit as well as the possibility of loss.
Residential mo
rtgage loans
GNMA
FNMA
FHLMC
Mortgage pool
Par: $100M
Coupon: 6.50%
Price: $119.32
OAD: 3.79 years
DV01: 4.52
Floater
Par: $100M
Coupon: SOFR + 20 bps
Cap: 6.50%
Price: $96.61
OAD: 1.00 years
DV01: 0.97
Investor objective
Floating-rate asset with
relative price stability
Amplified
return potential
Inverse IO
Par: $100M
Coupon: (6.30% – SOFR)
Price: $22.71
OAD: 15.63 years
DV01: 3.55
The structuring process also concentrates prepayment risk into
the inverse IO. The price of the pool equals the price of the floater
plus the price of the inverse IO ($119.32 = $96.61 + $22.71). The basis
for the price split between the floater and the inverse IO is again
the pool cash flow generated by the probabilistic model process
described earlier. The scheduled principal and prepayment cash
flows are directed to the floater along with its designated interest,
and the floater price is based on the present value of the cash
flows. The floater, because its coupon resets based on either a
1-month or 3-month reference rate, will price close to par. Similar
to the IO construction in Example A, the inverse interest cash flow
is directed to the inverse IO, and the price is based on the present
value of the flows. Note that the inverse IO price is much lower than
the floater and aligns with the greater uncertainty of cash flow, as
also illustrated with the IO in Example A.
In Example B, the mortgage pool is priced at $119.32. Principal
payments are received at $100, par value. If prepayments come
in faster than anticipated, the pool owner may lose $19.32 in
premium. The inverse IO, in the same vein as the IO, receives no
principal payments and is dependent on the outstanding principal
to generate interest. If prepayments come in faster than
anticipated and the principal balance is paid down to zero, there
is no interest for the inverse IO and the full investment of $22.71
may be lost. Alternatively, the same scenario is beneficial for the
floater investor. If they pay $96.61 for the floater security and
prepayments come in faster than anticipated, the floater investor
may gain $3.39 ($100 – $96.61 = $3.39). Another way to frame the
scenario: The sooner prepayments are received, the greater the
present value of the floater, while less interest is generated and the
price of the inverse IO declines.
17
May 2023 | Prepayment strategies: Essentials, analysis, and investing
For use with institutional investors and investment professionals only.
Calculating a pool coupon with inverse IOs and floater
Pool coupon = IIO coupon + Floater coupon
Pool coupon 6.50% = (6.30% – SOFR) * (Floater par / IIO par) + (SOFR + 0.20%, Max 6.50%)
When SOFR = 2.00%:
6.50% = (6.30% – 2.00%) * ($100M / $100M) + (2.00% + 0.20%, Max 6.50%)
6.50% = 4.30% + 2.20%
Calculating pool dollar duration with inverse IOs and floater
Pool DV01 = IIO DV01 + Floater DV01
(Pool price * Pool OAD) = IIO price * IIO OAD + Floater price * Floater OAD
($1.1932 * 3.79) = ($0.2271 * 15.63) + ($0.9661 * 1.00)
4.52 = 3.55 + 0.97
The floating coupons of the inverse IO and the floater
must move in opposite directions and perfectly oset
each other. In this example, the floater cap is activated
when SOFR reaches 6.30%.
The pool coupon (6.50%) and the cap set on the floater
(6.50%) define the par amount associated with each
component. In this example, the par amounts of the three
instruments are equal to one another because we have set
the pool coupon and the floater cap both equal to 6.50%.
The inverse IO/floater pairing also preserves the
replication of the pool’s price, OAD, and DV01.
Note that the OAD of the inverse IO is amplified, similar to
the negative OAD of the previous IO example; however, it
demonstrates amplified positive duration because it
reacts in the opposite manner of the IO. The inverse IO
does continue to demonstrate negative convexity, but to
a greater degree. The floater has a very low OAD, and its
price will typically fluctuate very little. This implies that
the price fluctuations of the pool are carried by the
inverse IO and are also amplified in the same manner as
the duration.
18
May 2023 | Prepayment strategies: Essentials, analysis, and investing
For use with institutional investors and investment professionals only.
Falling rates Rising rates
SOFR
-1.00% -0.50% 0.00% 1.00% 3.00% 5.00% 6.00% 6.50%
IIO coupon
6.50% 6.50% 6.50% 5.50% 3.50% 1.50% 0.50% 0.00%
Max coupon Min coupon
Inverse IO cash flows and performance
An inverse IO demonstrates dierent characteristics than
a fixed-rate IO. An inverse IO is an IO with an inverse
adjustable (floating) rate coupon. Instead of the coupon
rate being calculated as a reference rate (typically, LIBOR
or SOFR) plus a spread with a specified cap on the coupon
rate, the inverse IO coupon is calculated as a fixed interest
rate minus a reference rate with a coupon range based on
a specified level of the reference rate.
Continuing the example, the inverse IO uses SOFR as the
reference rate. In this example, the security has both an
eective cap (6.50%) and a floor (0%) determined based
on the level of SOFR.
The minimum coupon rate is 0% for this bond when
SOFR rises above 6.30%; the security does not pay any
interest. However, because this occurs in a rising-rate
environment, the associated principal is more protected
and less prone to prepayments.
If SOFR rises above 6.30% and the bond coupon resets to 0%, this
eectively turns the inverse IO into an option — a derivative that
may be compared in some respects to an interest-rate floor.
There is a subtle dierence between the two.
1 The coupon formula for a rate floor, the minimum
guaranteed coupon if interest rates fall, may look like:
MAX (SOFR + 0.20%, 6.50%)
In this instance, if SOFR falls below 6.30%, the coupon on
the security will be 6.50% and is a rate floor. Corporate
floating-rate bank loans can have a specified floor when
the reference rate falls below a specified threshold.
2 The interest-rate derivative market version of an
interest-rate floor has the coupon formula:
MAX (6.30% – SOFR, 0%)
If SOFR is above 6.30%, the coupon floor is 0%. But if
SOFR is below 6.30%, it has a coupon greater than zero.
In the familiar instance of a rate floor, a threshold is
established for the floor when interest rates are falling,
while in the instance of the inverse IO, the threshold
for the minimum coupon is established when interest
rates are rising due to the inverse nature of the coupon
calculation (see Figure 5).
FIGURE 5
Calculation of inverse IO coupons using SOFR rates
19
May 2023 | Prepayment strategies: Essentials, analysis, and investing
For use with institutional investors and investment professionals only.
As a result, in a rising-interest-rate environment, interest
payments will be lower as rates rise until they reach the
minimum coupon rate of 0%. The inverse IO tends to
underperform in a rising interest-rate environment,
demonstrating positive duration as demonstrated earlier.
Its floating-rate nature also creates exposure to the short
end of the curve. In a falling interest-rate environment,
interest cash flows will rise but only until principal erodes
as borrowers take advantage of the lower rates and
refinance (Figure 6). However, the value of the IO asset can
grow: The potential to earn cash flow exists as rates fall
and the price increases until principal degrades.
The security may not generate interest when rates move
higher, but it does maintain value due to the potential
for interest to increase if rates fall, at least until principal
goes away.
Interestingly, many times investors shy away from inverse
IOs when they have no cash flows and will likely sell them
extremely cheaply (e.g., 1000 OAS). Understanding how
to monetize the optionality when there are no cash flows
can be an advantageous strategy. The valuation of the
nonpaying inverse IO as essentially a floor may result in an
opportunity: An investor can purchase it at a discount to
the equivalent interest-rate floor.
FIGURE 6
Valuation of inverse IO — Payments increase
as rates fall only until refinance threshold
Source: Putnam.
For illustrative purposes only. The above example is not intended to be relied upon as a forecast, research, or investment advice, and is not
a recommendation, oer, or solicitation to buy or sell any securities or to adopt any investment strategy. As with any investment, there is a
potential for profit as well as the possibility of loss.
Rising interest ratesFalling interest rates
IO coupon %
Principal erosion and
declining interest
9%
8
7
6
5
4
3
2
1
0
Adjustable (floating) rate
interest payments up to maximum rate
until principal erosion occurs
Adjustable (floating) rate
interest payments down to a
minimum rate of 0%, residual
value of “floor” option
Interest-rate change
Coupon
20
May 2023 | Prepayment strategies: Essentials, analysis, and investing
For use with institutional investors and investment professionals only.
Investing in prepayment
securities
Analyzing refinancing behavior
Homeowners prepay their mortgage for various reasons,
some of which are structural in nature — such as when a
homeowner moves to a new home due to job relocation,
marriage/divorce, or to upgrade living standards. In these
events, dubbed “turnover,” the homeowner typically sells
the property and uses the proceeds to prepay/pay o
the remaining principal balance due on their mortgage.
These decisions are oen driven by life changes rather
than a financial incentive. While not as variable in nature
as refinancing based on financial incentive, even an
increase of 1 CPR in turnover can have a significant
impact on the valuation of prepayment securities.
Prepayments driven by financial incentive can vary
significantly over time due to changes in several factors,
such as interest rates, home price appreciation, and
government policy. The U.S. homeowners’ option to
prepay their mortgage at any time without penalty
aords them the opportunity to take advantage of
macroeconomic and mortgage industry developments
for their own financial benefit. In such an event, the
borrower refinances their mortgage by entering into a
new loan agreement for the same property and uses
all or a portion of the proceeds to prepay/pay o the
principal balance on their existing loan. This optionality
creates the potential for significant variability in
prepayment speeds and is the primary driver of
performance in prepayment markets.
Macro drivers
Understanding and analyzing the dierent macro
environment variables and their influence, either alone
or in combination, on prepayment speeds is crucial to
successfully implementing a prepayment strategy.
Interest rates
Interest rates have historically been the most influential
driver of prepayment speeds for traditional mortgage
borrowers, and thus the price of prepayment-sensitive
assets. When a borrower has the option to refinance their
existing mortgage at a lower interest rate, they may be
able to reduce their monthly payments and save money
over the long run if the decrease in their interest payments
exceeds the additional costs borne to originate the new
mortgage loan. Since mortgage rates oered on new
loans are driven by long-term interest rates (e.g., 10-year
U.S. Treasury note), expectations for the long end of the
U.S. yield curve play a key role in forecasting prepayment
speeds.
It is critically important to understand the context of
the existing mortgage within the current interest-rate
environment, such as the mortgage rates paid by
existing borrowers relative to the prevailing mortgage
rates oered on new loans. If most current borrowers
have mortgage rates of 4% and the prevailing mortgage
rate falls 100 bps from 4.5% to 3.5%, then a financial
incentive arises for those borrowers to refinance.
However, such an incentive does not result from a
100 bps drop if prevailing rates are at 5.5%.
15
The borrower’s response function to interest-rate
movements is best illustrated by the “S curve” shown
in Figure 7. The x-axis illustrates the interest-rate change
a borrower may achieve if they refinanced (refinancing
incentive), with the right side of the graph representing
a decrease in their interest rate and, hence, a potential
financial incentive to refinance. The le side represents
an increase in interest rate for the borrower, and a
potential financial disincentive to refinance. The y-axis
displays the rate at which borrowers prepay their
mortgages, represented by CPR. As seen on the le side
of the graph, there is a relatively low percentage of
borrowers that refinance even when they have a
disincentive, a result of structural prepayments that are
not driven by financial incentive (described previously
as “turnover”). To the right side of the x-axis, the curve
slopes upward as the refinancing incentive increases and
borrowers capitalize on it, until it gradually levels o.
The borrower response function to interest rates
(the slope of the S curve) does change over time and
is highly dependent on the existing mortgage market
at the time. The evolution of government policy and
technological advancements also impact borrower
behavior and are addressed in subsequent sections.
15 Siddhartha Jha, Interest Rate Markets: A Practical Approach to Fixed Income, 1st Edition, Wiley, 2011.
21
May 2023 | Prepayment strategies: Essentials, analysis, and investing
For use with institutional investors and investment professionals only.
FIGURE 7
Borrower refinancing behavior: The S curve
Sources: Bloomberg, Putnam, as of April 2021. For illustrative purposes only.
-3.00-2.00 -1.000.001.002.003.00
3M CPR rate
Refinancing incentive (%)
0
5
10
15
20
25
30
35
40
45
50
Interest rates may also fall to very low levels and remain
there for a considerable period of time. In this scenario,
prepayments typically spike in the early months as
borrowers react to the newly created financial incentive.
Over time, prepayment speeds are less likely to be as
sensitive to further interest-rate movements because the
bulk of borrowers who are interested and able to do so
have already refinanced. This is termed “burn out.”
Home price appreciation (HPA)
Another significant macro driver influencing the
prepayment environment is home price appreciation.
Faster HPA generally increases prepayment speeds
through a variety of mechanisms such as higher turnover,
easier and faster loan underwriting, and higher “cash
out” refinancings. When a home increases in value, the
borrower’s equity in the home also increases, and default
rates and loss severities decline. However, this gain is
unrealized until the borrower either sells the home or
unlocks the equity through a cash-out refi. This occurs
when a borrower enters into a new loan agreement with
a higher principal balance, which is made possible by the
increased value of the property. A portion of the proceeds
from the new loan are used to prepay the existing loan
principal, and the remainder is made available to the
homeowner. This type of refinancing activity becomes
more popular when home prices rise substantially.
For example, consider a borrower who purchased a home
for $250,000, currently owes $200,000 in principal, and has
homeowner’s equity of $50,000. If the home appreciates
in value and is now worth $350,000, the homeowner’s
equity increases to $150,000. The borrower can access
this equity by refinancing into a larger loan of $300,000,
providing them with an extra $100,000 in cash for other
expenses, e.g., building an addition to the home.
22
May 2023 | Prepayment strategies: Essentials, analysis, and investing
For use with institutional investors and investment professionals only.
Due to the potential for faster prepayments, faster HPA
can negatively impact the value of an IO, which in general
tend to benefit from slower HPA and likely slower speeds.
This is in contrast to bonds in the residential credit space,
where faster HPA improves the borrower’s equity in their
home and makes it less likely for them to default on
their mortgage.
Collateral factors
Within the context of the macroeconomic environment,
it is important to remember that all borrowers do not
necessarily behave in a similar manner. An individual
borrower’s response function can vary for a multitude of
reasons, such as the terms of their current loan, their credit
profile, and geography. As such, the collateral composition
of a mortgage pool plays an important role in forecasting
the pool’s prepayment speed and the resulting impact on
its cash flows.
Loan type and terms
Beyond the interest rate, there are several loan factors that
aect a borrower’s refinance incentive. The cost/benefit
analysis must consider metrics such as principal balance,
loan age, and loan-to-value ratio, among others. Consider
the two borrowers outlined in Figure 8; both currently pay
5% interest on their mortgages at a time when prevailing
mortgage rates are 4.5%. Both borrowers have a similar
credit profile and can refinance at a rate of 4.5%; however,
the first borrower has a $500,000 principal balance
outstanding on their current mortgage, while the second
borrower has $100,000 outstanding. In this scenario, the
first borrower is much more likely to refinance and realize
the cost savings than the second borrower, given the
available information.
Similar calculations and comparisons can be made among
the various loan maturities and interest-rate options
available to borrowers. While the 30-year fixed-rate
mortgage has historically been the most popular product,
lenders also oer mortgages with shorter terms, including
20 years, 15 years, and 10 years. The length of the loan
has significant implications for the borrower’s amortized
payment schedule and will impact the financial incentive
accordingly. Also, borrowers may choose a fixed-rate
mortgage that maintains a set interest rate throughout the
life of the loan or an adjustable-rate mortgage that resets
at specified periods based on a reference rate. As such,
a borrower’s response function to a change in market
interest rates will dier depending on the interest-rate
structure of their current loan.
There are also mortgage products that serve a dierent
purpose than traditional mortgages, such as reverse
mortgages. Reverse mortgages extend a line of credit
to borrowers age 55 and older,
16
which allows retirees to
access their home equity. Unlike a conventional mortgage,
which amortizes over time, a reverse mortgage loan does
not get repaid until one of a few possible events, including
mortality, morbidity, or refinancing. The refinancing
incentive to the borrower is quite narrowly beneficial, and
unlike prepayment incentives on typical mortgages, the
level of interest rates plays a weak role in the refinancing
activity of reverse mortgages. Rather, policy changes have
historically been the primary driver of refinancing events.
16 Borrowers must be 55 years of age in all states with the following exceptions: Louisiana, New Jersey, and Washington require 60 years of age,
while North Carolina, Texas, and Utah require 62 years.
23
May 2023 | Prepayment strategies: Essentials, analysis, and investing
For use with institutional investors and investment professionals only.
Borrower credit and demographics
A borrower’s credit profile impacts not only their ability to
access credit, but also the mortgage rate they are oered
and the lending platform under which they may qualify.
These factors can influence the borrower’s response
function to the refinance incentive and are important to
consider within the context of the macro environment
when determining the rate at which a given pool of loans
will prepay. Compare two borrowers currently paying 5%
interest on their mortgages at a time when the prevailing
mortgage rate is 4.5%. Ostensibly, both borrowers have
an incentive to refinance at the lower rate. However, if
one of these borrowers has a credit score (FICO) of 800
and the other a FICO of 690, lenders will likely oer them
dierent mortgage rates. The borrower with the higher
FICO score may receive a mortgage rate of 4.5%, in line
with the prevailing mortgage rate, which maintains the
refinance incentive. If the second, weaker borrower
receives a mortgage rate of 4.8%, the potential interest
savings from refinancing may not exceed the additional
costs associated with the refinancing process, eliminating
the incentive. The adjustment to the mortgage rate for
borrower quality characteristics is outlined in more detail
within the “Security selection: Assessing underlying
collateral section” (see page 26).
What is a FICO score?
17
The FICO score is a widely used metric by mortgage
lenders to help assess the likelihood a loan applicant
will repay their loan. The score was created by Fair, Isaac
and Company (now known as FICO). A borrower’s FICO
score impacts their ability to receive credit and the
interest rate charged on the loan. Scores range from
300 to 850 and incorporate five key data points from an
individual’s credit report: payment history, amounts owed
currently, length of credit history, new credit, and credit
mix (types of credit).
History has shown that the geography of the borrower
and property also influence the refinancing incentive
and/or the borrower’s behavior. The reasons for this
vary. Some states have more rigorous local regulations
that aect the ease and speed with which a refinance
can be processed. Others, such as New York, impose
additional taxes that result in higher up-front costs for
the borrower, thereby decreasing the refinance incentive
for its residents. Cultural factors may also play a role.
For instance, borrowers in Puerto Rico have historically
shown a greater reluctance to refinance their mortgages
despite a clear financial incentive, and as a result,
collateral in Puerto Rico tends to prepay slower than in
other regions.
17 https://www.myfico.com/credit-education/whats-in-your-credit-score (accessed 9/24/2021).
Borrower 1 Borrower 2
Current Refi Refi incentive Current Refi Refi incentive
Home value $750,000 $750,000 — $750,000 $750,000 —
Principal balance $500,000 $500,000 — $100,000 $100,000 —
Term (months) 360 360 — 360 360 —
Interest rate 5.0% 4.5% 0.5% 5.0% 4.5% 0.5%
Monthly payment $2,684 $2,533 $151 $537 $507 $30
Total interest due $466,479 $412,034 $54,445 $93,256 $82,407 $10,849
Source: Putnam.
FIGURE 8
Comparison of refinancing incentives
24
May 2023 | Prepayment strategies: Essentials, analysis, and investing
For use with institutional investors and investment professionals only.
Non-model considerations
Government policy
The U.S. government plays a key role in the functioning of
the mortgage finance system and the agency MBS market.
Housing policies, such as those implemented through the
FHFA and GSEs, can impact the aordability and availability
of mortgage financing and are subject to ever-changing
political, economic, and financial considerations. As such,
prepayment speeds can be influenced by new or changing
government policies and are an important consideration
of the valuation process.
For example, the FHFA implemented the Home Aordable
Refinance Program (HARP) in the wake of the 2008 housing
crisis when many borrowers found themselves “under
water”; the principal they owed on their mortgage exceeded
their property value. HARP encouraged borrowers that
remained current on their mortgage payments to refinance
despite the decline in their home’s value. This led to
meaningful dierences in prepayment speeds in mortgage
pools with otherwise similar collateral characteristics.
The process to comprehensively evaluate policy risk and its
impact on an individual security’s cash flows is important
in determining the risk/reward trade-o in prepayment-
sensitive sectors. Policy risk is ever evolving and does not
fit within the constraints of a quantitative model, thus
requiring careful consideration of policy changes on model
inputs and outputs.
All models try to fit historical data, but there is no such
thing as a perfect fit in empirical data modeling. As a
result, irrespective of sophistication, all models have
inherent biases and constraints. Being aware of these
biases allows for improving and deepening the security
selection process.
Technology’s impact on the mortgage industry
In the wake of 2008, lenders such as Quicken Loans,
Guaranteed Rate, and LoanDepot emerged as a growing
source of mortgage credit. Through technological
innovation, they oer borrowers a more seamless and
streamlined application process that is now conducted
almost entirely online. The growth of fintech accelerated
during the pandemic as companies were ]forced to adopt
technology to continue operations remotely at a time
when the mortgage industry faced a surge in demand.
The result oers benefits to both borrowers and lenders.
Improved experience: Applicants can receive
preapproval within minutes, connect to their financial
institutions electronically for streamlined asset
verification, sign documents from a mobile device, and
access tools (e.g., real-time payment calculators) that
may improve their decision-making.
Faster originations: On average, fintech lenders
process mortgage applications 20% faster than
non-fintech lenders (approximately 10 days).
18
The increased eiciency is even more pronounced for
refinance applications.
Supply elasticity: Increased automation of
underwriting processes and electronic data collection
reduces the eect of staing constraints, allowing
fintech lenders to respond rapidly to surges in demand.
In an industry where profitability is driven by volume,
technology that allows originators to process loans
more quickly and nimbly accommodate fluctuating
demand is a disruptor. Traditional mortgage lenders
have taken notice and are increasingly implementing
similar technology on their own lending platforms.
The result is a more eicient mortgage industry and, in
turn, an increased borrower response rate to refinance
incentives as evidenced by a historically steep S curve
during the most recent refinancing wave.
18 FRBNY, “The Role of Technology in Mortgage Lending,” February 2018 (accessed 11/5/2021).
25
May 2023 | Prepayment strategies: Essentials, analysis, and investing
For use with institutional investors and investment professionals only.
Security selection
Assessing underlying collateral
An agency mortgage pool comprises thousands of
underlying mortgages with similar maturities (e.g., 30
years) and interest rates. At first glance, an investor may
view two recently issued mortgage pool securities with
the same maturity and weighted average coupon (WAC)
as identical, and thus be ambivalent about which to
purchase. However, the collateral underlying these
securities can consist of a very dierent mix of
characteristics that potentially impact the pool’s
prepayment speed (see descriptions below).
Key metrics for evaluating prepayment securities
Spread at origination (SATO)
SATO measures the credit premium borrowers pay over
the prevailing market rate to account for the higher
credit risk associated with the borrower. This metric
amalgamates a number of factors related to a loan’s
credit profile, including observable metrics such as the
borrower’s FICO score and the loan-to-value (LTV) ratio,
as well as unobservable metrics, such as the borrower’s
previous payment history. On average, SATO is a good
measure of the pool’s credit profile and can help ascertain
whether the underlying borrowers may face additional
frictions that make it harder for them to refinance.
Loan age
Borrowers tend to be more responsive to a refinance
incentive in the first 1 to 2 years aer their loan is
originated. This is largely a behavioral phenomenon,
partly because a recent borrower will face fewer
frictions during the underwriting process, as well as
the complacency and comfort felt by more seasoned
borrowers that have been paying their mortgage for
several years.
Loan purpose
Mortgages can be originated to purchase a home or to
refinance an existing loan. Additionally, the mortgage may
be used to purchase the borrower’s primary residence or
to finance an investment property used for rental income.
This data is provided to investors and can be helpful in
understanding the underlying borrower behavior.
Originator type
With the regulations placed on banks in the wake of
the 2008 financial crisis, an increasing number of non-
bank lenders have entered the industry and increased
their market share. These lenders have developed and
adopted technology to help build their platforms, which
has led to notably faster prepayment speeds on the loans
they originate.
Geographical distribution
As discussed previously, borrower behavior can be
influenced by local regulations and/or customs. The
geographic distribution of the underlying collateral can
have a material impact on the pool’s prepayment rate.
26
May 2023 | Prepayment strategies: Essentials, analysis, and investing
For use with institutional investors and investment professionals only.
Identifying value
Most prepayment market participants use one of the
predominant vendor models (e.g., Yield Book, Bloomberg)
to evaluate prepayment securities. These are sophisticated
and complex models that consider various aspects of
the collateral, such as weighted average coupon, loan
size, loan age, SATO, loan-to-value, purpose of loan,
geographical distribution of borrowers, type of originator,
and the specific servicer, to project future probability
of prepayments and the resulting future cash flows.
Aer averaging spread expectations across hundreds of
simulations of forward interest rates, the resulting output
is an option-adjusted spread that can be used to assess a
security’s worth and determine its value relative to other
securities in the sector. Many investors take a black-box
model approach, making investment decisions based upon
the o-the-shelf model output.
The market generally allows a risk premium for securities
that are most susceptible to errors in these commonly
used prepayment models. We find the pricing of this
risk premium to be inconsistent, which creates potential
opportunities to buy and sell securities with ineicient
prepayment pricing. In particular, we believe the agency
IO sector provides a relatively liquid market for isolating
and capitalizing on these ineiciencies. We use the third-
party vendor models and then adjust various inputs and
assumptions to tailor the models to our specifications,
inherently creating a custom model. This allows us to
compare the outputs from our custom model with the
o-the-shelf model(s) to identify pockets of value that
may be overlooked, or conversely, pinpoint overvalued
securities. Given the complexity of this model, significant
eort is expended monitoring its outputs. As an example,
our monitoring system compares the mortgage rate
being used by the model with what is truly available
to borrowers in the primary market. We use periods of
dislocation between the actual primary rate and model
to capture value.
We also closely track prepayment speed errors in the third-
party models for various collateral types. We are proactive
in reviewing sectors that are either paying much slower
or faster than model projections, because these have the
potential to be significant return opportunities and are,
in fact, the first indicator of changes in the prepayment
environment. Each month when prepayment speeds are
released, we conduct a rigorous analysis of sectors where
the model seems to be missing on speeds and use that as
feedback to determine what changes, if any, need to be
made to the model parameters. Any dierences between
model speeds and actual speeds provide opportunities to
pursue alpha.
Value is ultimately determined using a combination of an
OAS framework and manager insights into non-modellable
aspects of the investment process, such as policy risks
and regime shis. An OAS framework works well when the
future is anticipated to behave like the past: The model
is, aer all, a calibration of prior empirical data. This
“static-ness” can be disrupted by policy risks or market
factors that change the S curve and the prepayment
environment. Additionally, technical factors, such as
supply/demand, the performance of related sectors such
as MBS pools, TBAs, and demand for the PO or floater play
a role in determining value for various types of bonds in the
agency IO market.
Prepayment strategies as
part of a broader portfolio
As concerns surrounding rising interest rates and inflation
continue to percolate throughout the economy, many
investors are seeking strategies to diversify their fixed
income exposure away from more interest-rate and
corporate credit-sensitive sectors. Among the four principal
risks inherent to fixed income investments (credit, interest
rate, liquidity, and prepayment), we believe prepayment
risk may be underrepresented as a risk premium in
many global investors’ fixed income portfolios. Agency
IOs tend to receive less attention from investors due
to the specialization required to extract value and the
sector’s relatively small size. However, we believe this
segment of the market oers exposure to a risk premium
that can complement portfolios and provide significant
diversification benefits.
27
May 2023 | Prepayment strategies: Essentials, analysis, and investing
For use with institutional investors and investment professionals only.
Diversification and correlation
Agency IO securities have historically performed better and
tend to increase in value in rising interest-rate environments
as U.S. homeowners lose the incentive to refinance.
Consequently, they are oen used to hedge portfolios
against interest-rate risk. For example, in a Federal Reserve
tapering environment, the Fed gradually slows the pace of
large-scale asset purchases initially designed to support
the economy and subsequently reduces its balance sheet
by not replacing maturing securities. A prominent example
of tapering as well as the diversification benefit of agency
IOs was experienced in 2013. Fears that the Fed was
beginning to unwind its accommodative policy following
the GFC led to a spike in market interest rates. During this
time, interest-rate and credit-sensitive strategies generally
underperformed, while agency IO securities experienced
positive excess returns versus Treasuries.
Additionally, we believe agency IOs to be a good diversifier
as a part of a broader portfolio given their historically low
correlation versus other asset classes, including other types
of structured credit. The low correlations stem from the fact
that agency IOs are exposed to a dierent risk premium,
prepayment risk, and investing in agency IOs is one of
the purest ways of expressing a view on the prepayment
environment. The factors that drive prepayment behavior
are independent of factors that drive the balance sheets of
other asset classes.
The diversification benefit of agency IO securities can be
demonstrated through empirical correlation analysis.
Running correlation analysis on various financial sectors is
useful because it helps measure and quantify the degree
of association between the sectors based on historical
returns. Eective portfolio diversification is enhanced when
assets are either uncorrelated or negatively correlated.
Typically, a correlation between two variables of less than
0.60 is considered lower/weaker, and a correlation of less
than 0.30 generally indicates there is no linear correlation
or relationship. The correlation table below (Figure 9),
constructed using duration-hedged monthly returns to
remove the eect of duration, illustrates that agency IO
securities have had very low correlations and, in some
cases, nearly no linear relationship to other sectors.
Given these low correlation values, agency IO securities
prove to be very eective in diversifying portfolio exposure
and helping to mitigate risk.
28
May 2023 | Prepayment strategies: Essentials, analysis, and investing
For use with institutional investors and investment professionals only.
FIGURE 9
Correlation of duration-hedged monthly excess returns,
10 years rolling
Historical returns versus other fixed income sectors
An important consideration when analyzing a diversified
portfolio is the excess return generated by the various
sector allocations within the portfolio. Additionally,
investors generally focus on maximizing returns in the
context of managing risk. The Sharpe ratio may be used
to measure the risk-adjusted performance of individual
sectors, as well as the portfolio itself, by comparing the
associated total returns in excess of the risk-free rate per
unit of volatility (standard deviation of total return). In
general, a higher Sharpe ratio is preferred as it means that
the sector or portfolio has achieved a better return relative
to its risk. If two sectors oer similar returns, the one with
the higher standard deviation will have a lower Sharpe
ratio, and in order to compensate for the higher standard
deviation, the sector must deliver correspondingly higher
returns to maintain a higher Sharpe ratio. The Sharpe
ratio is a good measure of how much additional return an
investor earns by taking on additional risk.
Corporate debt Equities Securitized
Global U.S. Euro U.K. Global U.S. Euro U.K.
NA
RMBS
Agency
IO CMBS
Agency
MBS
Corporate
debt
Global
—
U.S.
0.98 —
Euro
0.91 0.83 —
U.K.
0.82 0.73 0.81 —
Equities
Global
0.80 0.79 0.71 0.62 —
U.S.
0.72 0.71 0.62 0.48 0.94 —
Euro
0.79 0.75 0.74 0.70 0.92 0.78 —
U.K.
0.77 0.75 0.68 0.67 0.88 0.74 0.94 —
Securitized
NA RMBS
0.60 0.59 0.53 0.51 0.47 0.39 0.47 0.45 —
Agency IO
0.22 0.20 0.20 0.15 0.10 0.09 0.10 0.08 0.15 —
CMBS
0.50 0.47 0.45 0.58 0.49 0.42 0.49 0.55 0.44 0.12 —
Agency MBS
0.54 0.52 0.45 0.37 0.46 0.37 0.49 0.55 0.16 0.34 0.30 —
Sources: Bloomberg, Putnam as of 12/31/22. For illustrative purposes only. Indices used in the above calculations include the Bloomberg Global
Aggregate Corporates Index (USD hedged), Bloomberg U.S. Corporate Index, Bloomberg Euro-Aggregate Corporates Index (USD hedged),
Bloomberg Sterling Aggregate Corporates Index (USD hedged), MSCI World Index (USD hedged), S&P 500 Index, Stoxx Europe 600 USD, and the FTSE
100 USD. Where there is no available representative index, data is based on a universe of securities selected by Putnam that are representative of
various fixed income sectors and subsectors within the mortgage market. Past performance is not a guarantee of future results. Diversification does
not assure a profit or protect against loss. It is possible to lose money in a diversified portfolio.
29
May 2023 | Prepayment strategies: Essentials, analysis, and investing
For use with institutional investors and investment professionals only.
FIGURE 10
Annual excess returns over risk-free rate and Sharpe ratios
In Figure 10, most sectors have demonstrated Sharpe
ratios of either above 1 or close to 1. A Sharpe ratio above 1
indicates that the sector is achieving higher returns above
the risk-free rate relative to its risk. However, the agency IO
sector has achieved a Sharpe ratio of 0.12 during the same
time period. This is due to the fact that the total return used
in the numerator for the agency IO calculation includes
embedded duration. Agency IO securities exhibit negative
duration, meaning that as interest rates rise, the prices of
agency IOs also rise, and vice versa. During the historical
time period reflected in the Sharpe ratios calculated below,
interest rates have declined overall, and since agency
IOs exhibit negative duration, this has weighed more
heavily on returns compared with other sectors and has
contributed to heightened volatility. When investing in
agency IO securities, using the total unhedged return in the
numerator results in a skewed Sharpe ratio for agency IOs
compared with other asset classes.
The information ratio (IR) is also a measure to gauge risk-
adjusted returns; however, it compares excess returns with
the returns of a benchmark, while also accounting for the
volatility of those returns. In other words, the IR represents
risk-adjusted alpha. In general, the higher the IR the better,
and if the IR is less than zero, it signifies the manager did
not achieve the objective of outperforming the benchmark.
With regard to agency IOs, unlike the Sharpe ratio
formula in which the unhedged total return is used in the
numerator, the information ratio uses duration-hedged
excess returns, eliminating the duration bias. The duration
of our agency IO securities is managed at the portfolio
level. This is accomplished by aggregating the term
structure exposure of each bond via four risk factors: level,
slope, bend, and wave at the portfolio level. These factor
exposures are matched by building a replicating basket of
swaps/swaptions for the term structure exposures. The
resulting duration-hedged excess return equals the total
return of agency IOs minus the total return of the basket of
hedging securities.
As evidenced in the table below, the agency IO sector has
achieved an information ratio of 0.69 since 2009, showing
that the sector has outperformed on a risk-adjusted basis
(Figure 11, on following page).
Since 2009
INV GR
Corporates
HY
Corporates LOANS EM USD S&P 500
MSCI
World NA RMBS Agency IO CMBS
Agency
MBS
Annual excess return 3.94% 8.20% 6.63% 5.30% 12.53% 9.21% 10.45% 2.26% 5.23% 0.94%
Annual volatility 5.95 8.42 6.21 7.40 15.39 15.73 9.51 18.40 6.13 3.26
Sharpe ratio 0.66 0.97 1.07 0.72 0.81 0.59 1.10 0.12 0.85 0.29
Sources: Bloomberg, Putnam as of 12/31/22. Data is provided for informational use only. Past performance is no guarantee of future results.
IG corporates are represented by the Bloomberg U.S. Corporate Index. High yield is represented by the Bloomberg U.S. High Yield Index.
Loans are represented by the Morningstar® LSTA® US Leveraged Loan Index. Emerging market debt is represented by the Bloomberg EM USD
Aggregate Index. S&P 500 is represented by the S&P 500® Index. MSCI World is represented by the MSCI World Index (USD hedged). Where there
is no available representative index, data is based on a universe of securities selected by Putnam that are representative of various fixed income
sectors and subsectors within the mortgage market. Diversification does not assure a profit or protect against loss. It is possible to lose money
in a diversified portfolio.
30
May 2023 | Prepayment strategies: Essentials, analysis, and investing
For use with institutional investors and investment professionals only.
FIGURE 11
Annual excess returns over duration-matched swaps
and information ratios
Since 2009
INV GR
Corporates
HY
Corporates LOANS EM USD S&P 500
MSCI
World NA RMBS Agency IO CMBS
Agency
MBS
Annual excess return 2.79% 7.24% 7.30% 4.10% 13.23% 9.88% 8.27% 8.13% 7.05% 0.22%
Annual volatility 5.23 8.98 6.23 7.78 15.44 15.78 7.82 11.85 6.71 1.98
Information ratio 0.53 0.81 1.17 0.53 0.86 0.63 1.06 0.69 1.05 0.11
Sources: Bloomberg, Putnam as of 12/31/22. Data is provided for informational use only. Past performance is no guarantee of future results.
IG corporates are represented by the Bloomberg U.S. Corporate Index. High yield is represented by the Bloomberg U.S. High Yield Index.
Loans are represented by the Morningstar LSTA US Leveraged Loan Index. Emerging market debt is represented by the Bloomberg EM USD
Aggregate Index. S&P 500 is represented by the S&P 500 Index. MSCI World is represented by the MSCI World Index (USD hedged). Where there is
no available representative index, data is based on a universe of securities selected by Putnam that are representative of various fixed income
sectors and subsectors within the mortgage market. Diversification does not assure a profit or protect against loss. It is possible to lose money
in a diversified portfolio.
It was noted earlier that other investment managers
typically use agency IOs, given their negative duration
characteristic, to hedge their broader portfolio duration.
Rather than use agency IOs as a duration hedge, we
invest in IOs as a prepayment strategy and hedge the
strategy’s duration, seeking to capture the prepayment
risk premium rather than be subject to the interest-rate
risk of the securities. This approach is intended to provide
our investors with the attractive duration-hedged,
risk-adjusted return potential of the strategy whether
in the context of a multisector fixed income portfolio or
a stand-alone prepayment portfolio.
Consider investing in
prepayment risk
Of the four primary bond risks, prepayment risk is typically
underrepresented within investor portfolios because of
the limited degree of market understanding. The heart
of the challenge lies in analyzing the unique ability of the
U.S. homeowner to prepay their mortgage loan at any
time without a penalty. We believe successful investing in
prepayment strategies requires a thorough understanding
of the mortgage market, cash flow structuring, and the
means to analyze borrower behavior incentives in order to
identify and extract value.
Agency IO securities are the purest expression of
prepayment risk. While IOs may exhibit a high degree of
volatility based on their sensitivity to borrower refinancing,
they are demonstrable risk mitigants within a broader
portfolio when duration is hedged. Hedging the interest-
rate risk of the strategy also allows for the capture of
prepayment risk premium and an attractive risk-adjusted
return potential.
31
May 2023 | Prepayment strategies: Essentials, analysis, and investing
For use with institutional investors and investment professionals only.
This material is prepared for use by institutional investors and investment
professionals and is provided for limited purposes. This material is a general
communication being provided for informational and educational purposes
only. It is not designed to be investment advice or a recommendation of any
specific investment product, strategy, or decision, and is not intended to
suggest taking or refraining from any course of action. The opinions
expressed in this material represent the current, good-faith views of the
author(s) at the time of publication. The views are provided for informational
purposes only and are subject to change. This material does not take into
account any investor’s particular investment objectives, strategies, tax
status, or investment horizon. Investors should consult a financial advisor
for advice suited to their individual financial needs. Putnam Investments
cannot guarantee the accuracy or completeness of any statements or data
contained in the material. Predictions, opinions, and other information
contained in this material are subject to change. Any forward-looking
statements speak only as of the date they are made, and Putnam assumes
no duty to update them. Forward-looking statements are subject to
numerous assumptions, risks, and uncertainties. Actual results could dier
materially from those anticipated. Past performance is not a guarantee of
future results. As with any investment, there is a potential for profit as well
as the possibility of loss.
Use of models and analytical, quantitative, and risk management tools and
techniques is no guarantee of investment success or positive performance.
These models are dynamic, and as a result, the variables or factors or
components may change over time. Models are based on historical data,
which is not indicative of future results. Other tools may produce
substantially dierent success and failure outcomes. Asset-backed
securities, including mortgage-backed securities, may have structures that
make their reaction to interest rates and other factors diicult to predict,
making their prices volatile, and they are subject to liquidity risk. Mortgage-
and asset-backed securities are subject to prepayment risk, which means
that they may increase in value less than other bonds when interest rates
decline and decline in value more than other bonds when interest rates rise.
Traditional debt investments typically pay a fixed rate of interest until
maturity, when the entire principal amount is due. By contrast, payments
on securitized debt instruments, including mortgage backed and asset-
backed investments, typically include both interest and partial payment of
principal. Principal may also be prepaid voluntarily, or as a result of
refinancing or foreclosure. Compared to debt that cannot be prepaid,
mortgage-backed investments are less likely to increase in value during
periods of declining interest rates and have a higher risk of decline in value
during periods of rising interest rates. They may increase the volatility of the
portfolio. Some mortgage-backed investments receive only the interest
portion or the principal portion of payments on the underlying mortgages.
The yields and values of these investments are extremely sensitive to
changes in interest rates and in the rate of principal payments on the
underlying mortgages. The market for these investments may be volatile
and limited, which may make them diicult to buy or sell. Asset-backed
securities are structured like mortgage-backed securities, but instead of
mortgage loans or interests in mortgage loans, the underlying assets may
include such items as motor vehicle installment sales or installment loan
contracts, leases of various types of real and personal property, and
receivables from credit card agreements. Asset-backed securities are
subject to risks similar to those of mortgage-backed securities. Exposure to
mortgage-backed securities may make an investment strategy more
susceptible to economic, market, political, and other developments
aecting the housing or real estate markets and the servicing of mortgage
loans secured by real estate properties.
This material or any portion hereof may not be reprinted, sold, or
redistributed in whole or in part without the express written consent of
Putnam Investments. The information provided relates to Putnam
Investments and its ailiates, which include The Putnam Advisory Company,
LLC and Putnam Investments Limited®.
Issued in the United Kingdom by Putnam Investments Limited®. Putnam
Investments Limited is authorized and regulated by the Financial Conduct
Authority (FCA). For the activities carried out in Germany, the German
branch of Putnam Investments Limited holds a permit as a financial
investment broker in accordance with Sec. 34f para. 1 sentence no. 1 of the
German Trade Ordinance and is registered under the registration number
D-F-16077N8-19 in the publicly available broker register. Putnam
Investments Limited is also permitted to provide cross-border investment
services to certain EEA member states. In Europe, this material is directed
exclusively at professional clients and eligible counterparties (as defined
under the FCA Rules, or the German Securities Trading Act
(Wertpapierhandelsgesetz) or other applicable law) who are knowledgeable
and experienced in investment matters. Any investments to which this
material relates are available only to or will be engaged in only with such
persons, and any other persons (including retail clients) should not act or
rely on this material. Furthermore, this material is only intended for the
recipient receiving it directly from Putnam Investments Limited and should
not be forwarded to, or relied upon by, the recipient’s underlying clients.
This material is prepared by Putnam Investments for use in Japan by
Putnam Investments Japan Co., Ltd. (“Putnam Japan”). Putnam Japan is
registered with Kanto Local Finance Bureau in Japan as a financial
instruments business operator conducting the investment advisory and
agency business, and is a member of Japan Investment Advisers
Association. This material is prepared for informational purposes only, and
is not intended as a solicitation to invest in any securities or other financial
product in Japan, or to enter into with Putnam Japan an investment
advisory contract or a discretionary investment management contract.
Prepared for use in Canada by Putnam Investments Canada ULC (o/a
Putnam Management in Manitoba). Where permitted, advisory services
are provided in Canada by Putnam Investments Canada ULC (o/a
Putnam Management in Manitoba) and its ailiate, The Putnam Advisory
Company, LLC.
Prepared for use with wholesale investors in Australia by Putnam
Investments Australia Pty Limited, ABN 50 105 178 916, AFSL No. 247032.
This material has been prepared without taking account of an investor’s
objectives, financial situation and needs. Before deciding to invest,
investors should consider whether the investment is appropriate for them.
BLOOMBERG® is a trademark and service mark of Bloomberg Finance L.P.
and its ailiates (collectively “Bloomberg”). Bloomberg or Bloomberg’s
licensors own all proprietary rights in the Bloomberg Indices. Neither
Bloomberg nor Bloomberg’s licensors approve or endorse this material, or
guarantee the accuracy or completeness of any information herein, or make
any warranty, express or implied, as to the results to be obtained therefrom,
and to the maximum extent allowed by law, neither shall have any liability
or responsibility for injury or damages arising in connection therewith.
