Structured certificates of deposit:
Introduction and valuation

Geng Deng, Ph.D., CFA, FRMa, Tim Dulaney, Ph.D., FRMa,
Tim Husson, Ph.D., FRMa, Craig McCann, Ph.D., CFAa,*

aSecurities Litigation and Consulting Group, Inc., 3998 Fair Ridge Drive, Suite 250, Fairfax, VA 22033, USA

Abstract

This article examines the properties and valuation of market-linked certificates of deposit (struc-
tured CDs). Structured CDs are similar to structured products—debt securities with payoffs linked to
market indexes—but while structured products have garnered significant interest in both the financial
media and in the academic literature, structured CDs have received relatively little attention. We
review the market for structured CDs in the United States and provide valuations for several common
product types. Using our methodology, we find significant mispricing of several common types of
structured CDs across multiple issuers, which is similar in magnitude to the well-documented
mispricing in the structured products market. In particular, we estimate that structured CDs are
typically worth �93% of the value of a contemporaneously issued fixed-rate CD. These results
suggest that unsophisticated investors may not understand the value, risks, and subtleties of these
ostensibly conservative investments. © 2014 Academy of Financial Services. All rights reserved.

Keywords: Certificate of Deposit; Structured Product; CD

1. Introduction

Structured certificates of deposit (structured CDs), also referred to as “market-linked
CDs,” “equity-linked CDs,” and “contingent interest CDs,” have existed since the late 1980s.
Structured CDs are FDIC insured deposits with interest payments that are contingent upon
changes in the levels of indexes, individual equities, and interest rates, or combinations of
indexes, individual equities, and interest rates.

* Corresponding author. Tel.: �1-703-246-9381; fax: �1-703-246-9387.
E-mail address: craigmccann@slcg.com (C. McCann)

Financial Services Review 23 (2014) 219–237

1057-0810/14/$ – see front matter © 2014 Academy of Financial Services. All rights reserved.



Because structured CDs are not securities registered with the SEC, the size of the
structured CD market is not clear, but has been estimated to be in the tens of billions of
dollars annually and growing.1 Investors in structured CDs have been described as “risk-
averse” and “conservative,” but “dissatisfied with the rates available in the traditional
fixed-income space.”2 A recent study conducted by the International Organization of Secu-
rities Commissions found that, at least in the United Kingdom, structured CD investors were
typically over 55 or under 35 with high household income.3 Market observers suggest that
banks are issuing increasing quantities of structured CDs to compete with the low yield on
traditional CDs and treasury securities. These reports suggest that structured CDs are
primarily being offered to individual investors.

Offering documents and marketing materials for structured CDs also suggest that they are
being sold through financial advisors to individual investors. Although they are often
marketed alongside structured notes and other retail investments, they can also be sold at or
through local banks, potentially bypassing regulatory requirements of brokers and financial
advisors. There remains a degree of regulatory uncertainty regarding structured CDs, which
could lead to unique risks and inappropriate sales to investors.

The academic and practitioner literature on structured CDs is sparse compared with the
considerable body of theoretical and empirical work for structured products.4 King and
Remolona (1987) showed that banks can typically hedge structured CDs efficiently using
exchange-traded or synthetic options matching the payout of the structured CD.5 Chance and
Broughton (1988) provided additional analysis and interpretation of these products. More
recently, Edwards and Swidler (2005) argue that equity-linked structured CDs do not have
equity-like returns. Unfortunately, these studies analyzed only a few structured CD types and
did not value a significant number of actual products.

Given these substantial risks and the lack of an established literature for this type of
investment, investors and financial advisors may not fully understand these highly complex
products. As with structured products, the customizability of structured CDs allows issuers
to charge additional gross margin into the product without explicitly stating their impact as
commissions and fees. Fees on structured CDs are often between 3% and 4%, and we have
observed all-in fees as high as 8%, but could effectively be higher because of mispricing of
the embedded options. Financial advisors should recognize the potential for this form of
mispricing and the features that can exacerbate it.

In addition, investors and advisors should recognize that structured CDs typically carry
FDIC insurance, but are not risk-free. The SEC has outlined a variety of risks in structured
CDs, including liquidity risk, market risk, call risk, as well as special tax considerations.6

FINRA is reportedly investigating how structured CDs are sold to investors given their
increasing complexity and market growth.7 The FDIC has also issued a short investor alert
on structured CDs,8 and the NYSE has provided guidance to issuers regarding sales practices
and disclosures.9 However, structured CDs remain relatively unexplored by analysts and
underreported by the financial media.

We extend this literature by (1) describing the market for structured CDs using a sample
of over 2,000 CDs; (2) providing valuations for four of the most common crediting formulas
in our sample; and (3) valuing over 300 products issued by a variety of banks and
documenting significant discounts to face value. Our findings suggest that structured CDs

220 G. Deng et al. / Financial Services Review 23 (2014) 219–237



tend to be worth less than face value or contemporaneously issued traditional CDs, have a
high probability of crediting the minimum return, and have death provisions that are often of
negative value to the investor. These properties would not be apparent to retail investors.

2. The market for structured CDs

2.1. Our sample of structured CD

We have collected a sample of 2,072 structured CDs issued in the past eight years.10 We
searched for CDs issued in the United States with more than $1,000,000 issued that reported
an underlying security to Bloomberg. The latter requirement confines our sample to struc-
tured CDs linked to equities, commodities, and other tangible assets, and therefore, excludes
structured CDs linked to interest rates, such as steepener and range accrual CDs. Although
we collected products from as early as 2005, our sample is heavily concentrated in 2009–
2013. Both the weighted average term—weighted by amount outstanding—and simple
average for the structured CDs in our sample is 5.8 years.

Table 1 summarizes our sample by issuing bank.11 Although other banks’ average issue
size ranges from $1.3 million to $16 million, Bank of America has the largest average issue
size ($81.6 million). Bank of America’s large average issue size is because of several large
deals including a billion dollar structured CD (CUSIP: 06051ACF7) issued in February
2008.

The great majority of the structured CDs in our sample are linked to baskets of commod-
ities, equities, currencies, indexes, or some combination thereof. Table 2 summarizes the
major underlying assets within our sample. Two of the top underlying assets are JP Morgan
proprietary indexes.12 All of the structured CDs linked to these indexes were issued by JP
Morgan. In fact, more than half of the JP Morgan structured CDs in our sample (by aggregate
amount outstanding) were linked to a JP Morgan proprietary index. JP Morgan’s use of its
proprietary index in its CDs adds to the information advantage issuers have over investors.

Table 1 Structured CD market sample by issuer

Issuer Number Total face value (MM)

HSBC 906 $ 5,827.6
JP Morgan Chase 499 $ 2,607.3
Bank of America 27 $ 2,202.1
Barclays 213 $ 1,444.6
SunTrust 134 $ 896.3
Wells Fargo 165 $ 615.6
Union Bank 23 $ 370.8
Citigroup 23 $ 283.5
BMO Harris Bank 30 $ 161.3
Other Issuers 52 $ 263.0
Total 2,072 $14,672.0

221G. Deng et al. / Financial Services Review 23 (2014) 219–237



2.2. How structured CDs are marketed and sold in the United States

We have also collected a number of marketing brochures and other sales material related
to structured CDs from a variety of issuers. In general, issuers claim that structured CDs
combine the safety of CDs with additional market-linked upside. Wells Fargo, for example,
claims that structured CDs “can provide a creative solution for investors looking to gain
access to the markets while reducing their exposure to market risk when held to maturity.”13

Our results in Section 4 suggest that the potential market-related gains in structured CDs are
minimal.

The offering documents for structured CDs and those for structured products are similar.14

Both disclose the terms and relevant pricing parameters used in the offering, and tend to
follow a similar format for each offering from a given issuer. Other marketing materials, such
as sales brochures and lists of available products, are also similar to those that exist for
structured products (and can often be found on the internet from brokerage firm websites).
Structured CD materials, of course, tend to prominently note FDIC insurance.

However, brokers who sell structured CDs, just like traditional CDs, do not have to be
registered nor licensed by any state or federal agency, though many may be associated with
banks or licensed broker-dealer institutions.15 According to the NYSE Rules 401 (“Business
Conduct”) and 405 (“Diligence as to the accounts”), structured CDs should be “priced at
market value on customer account statements, not at the purchase amount or at par unless that
is the actual market price.”16 Although structured products are now required to prominently
disclose a fair market value on offering materials, there is no such disclosure requirement for
structured CDs. As we will demonstrate below, the fair market value of structured CDs is
often much less than face value, reflecting a premium to the issuer that is similar in
magnitude to those seen in structured products.

In fact, not having to disclose fair market values might be a reason for banks to issue
structured CDs over structured products or traditional CDs. FDIC insured banks are limited
to offering CD rates that are no more than 75 basis points above the average of yields for CDs
of comparable terms within their local market without a waiver from the FDIC.17 FDIC
insured banks and deposit brokers may issue structured CDs to circumvent this limit. For
example, instead of issuing a traditional CD at a low fixed rate, a bank could market a
structured CD with a higher comparable yield than the FDIC mandated cap. An investor may

Table 2 Structured CD market sample by underlying asset

Underlying asset Number Total face value (MM)

Baskets 1,293 $ 9,346.5
S&P 500 Index 210 $ 1,924.1
JP Morgan ETF Efficiente 5 Index 111 $ 942.8
Dow Jones Industrial Average 136 $ 913.3
JP Morgan Optimax Market-Neutral Index 34 $ 260.7
Dow Jones-UBS Commodity Index 26 $ 204.2
Russell 2000 Index 37 $ 135.6
Other Underlying Assets 225 $ 944.7
Total 2,072 $14,672.0

222 G. Deng et al. / Financial Services Review 23 (2014) 219–237



believe that this product offers higher yield than a traditional CD, even though its fair value
might in fact be lower.

3. How to determine fair value

We present a Monte Carlo method for determining the fair value of some common
structured CDs. We assume throughout that the underlying asset returns follow geometric
Brownian motion (Black and Scholes, 1973; Merton, 1973), and simulate asset values with
a Monte Carlo framework in accordance with procedures outlined in (Glasserman, 2003).
The variance of that geometric Brownian motion is set to be the longest-term implied
volatility of the underlying asset available from Bloomberg. Investors in structured CDs are
not paid distributions and the underlying returns reflect only capital appreciation.

We compare our structured CD values to the present value of contemporaneously issued
traditional CDs of a similar maturity. Fig. 1 graphs the simple average of traditional CD rates
as reported by the FDIC between May 2009 and March 2013.18 For our analysis, we used the
“nonjumbo” rates (for deposit amounts under $100,000) as these are most likely to be FDIC
insured; however, the average difference between jumbo and nonjumbo rates over this period
was only two basis points according to the FDIC data.

The secondary market rates for short-term CDs as quoted by the Federal Reserve Board
are universally higher than those quoted for CDs in an initial offering by the FDIC, often by
50 basis points or more. We include the reported national average rates for five-year
traditional CDs from BankRate.com in Fig. 1 to show the magnitude of the discrepancy.19

We use the rates reported by the FDIC as a conservative estimate for the relative value of a
structured CD.

Fig. 1. National average rates on traditional five year CDs.

223G. Deng et al. / Financial Services Review 23 (2014) 219–237



3.1. Credit risk

Investors who purchase structured CDs bear credit risk. The FDIC made clear to investors
in the spring of 2012 that its insurance covers only principal repayment at maturity and
accrued interest, not any market-linked or contingent interest payments.20 Therefore, as
reflected in our valuation equation in Section 4, structured CDs are subject to the credit risk
of the issuer insofar as they are exposed to market fluctuations. As a result, a structured CD
that is issued by a less credit-worthy bank is worth less than an identical structured CD issued
by a more credit-worthy bank.

In addition, there are limits to the amount covered by FDIC insurance.21 The limit can be
made less restrictive by depositing funds at more than one FDIC insured institution, across
several ownership categories or through joint ownership of deposited funds. Investors should
also realize FDIC insurance does not protect against losses on early withdrawals, which
usually come with heavy penalties or forfeiture provisions.22 Any investment in a structured
CD over the FDIC limit would further expose investors to credit risk.

3.2. Liquidity and call risk

There are no public sources for market prices of structured CDs or any independent
market makers for structured CDs. Secondary markets for structured CDs, like those for
structured products, may be limited to the broker-dealers or banks who issued them, which
may in turn limit the price at which an investor can sell his or her investment before
maturity.23 We have not included in our valuations any discount to reflect the illiquidity of
structured CDs. Therefore, our valuations are conservative.

Some structured CDs include a call feature whereby the issuer may redeem the structured
CD at par at certain times before maturity. If a structured CDs’ underlying asset has increased
in value and payment at maturity is likely large, an issuing bank could in theory call back the
note at par, depriving the investor of the market-linked return despite having tied up his or
her money in the structured CD. Call risk is, therefore, a potentially significant factor in
structured CD valuation, though none of the products in our valuation sample are callable.

4. Valuing four common structured CD types

The following four subsections introduce common product structures found in our struc-
tured CD sample: contingent coupon, single-observation, ratchet, and average-value CDs.
Contingent coupon CDs pay periodic coupons based on the return of an underlying asset.
Single-observation CDs pay back principal at maturity plus an interest component that
reflects the return of the underlying asset over the entire term of the CD. Ratchet CDs also
payout only at maturity, but their interest component is based on returns calculated over
several periods. The interest component of average value CDs is based on the average of the
underlying asset level at several times before maturity. We present valuation methods for
each type, along with example valuations. Unless otherwise stated, we report the simple
average of the valuations for investors aged 25 to 85.

224 G. Deng et al. / Financial Services Review 23 (2014) 219–237



4.1. Contingent coupon CD

The most common type of structured CD in our sample is called a contingent coupon CD.
Contingent coupon CDs are typically linked to baskets of assets and pay periodic interest
payments contingent upon the return of the basket. Before each coupon payment date, the
prices of the basket elements are observed. Returns are calculated, capped and floored to
produce the component returns. If the weighted average of these component returns is larger
than the minimum coupon (usually zero), then a market-contingent coupon payment is paid
to the investor. Otherwise, the minimum coupon is paid to the investor, which could be zero.

Component returns for this type of CD are calculated in one of two ways: unbuffered or
buffered. An unbuffered component return is a continuous function of the basket element’s
return. With a buffered component return CD, if the basket element’s return is larger than the
buffer level, then the investor is credited with the maximum return (the return cap).

The payoff of unbuffered structured CDs includes payoffs to vanilla European options and
the payoff of buffered structured CDs includes payoffs to both vanilla and binary options. For
the same level of cap and floor, a buffered CD would be worth more than an unbuffered CD
(assuming the buffer is less than the cap). As a result, for a buffered CD and an unbuffered
CD to be similarly profitable, either the return cap or return floor must be lower on the
unbuffered CD, all else equal.

For an example valuation, we take the contingent coupon CD issued by Barclays Bank
Delaware on February 29, 2012, maturing on February 28, 2019. The CUSIP for this CD is
06740ARX9 and the aggregate issue size is $2,915,000. The CD was linked to an equally
weighted basket of 10 equities. Barclays paid coupons annually at a rate determined by a
floor of �25%, a cap of between 6.5% and 9.5% and a buffer of 0%.

We assumed a cap of 8% when valuing this buffered contingent coupon CD and found that
the CD was worth approximately $933.79 per $1,000 face value CD; �91% of the value of
a contemporaneously issued fixed-rate CD.24 We found that in two-thirds of the simulations,
investors received only the minimum return.

As an example of an unbuffered contingent coupon CD, we use the JP Morgan CDs issued
on February 29, 2012 with an aggregate face value of $4,241,000 (CUSIP: 48123Y6G6). The
products included a return cap of at least 6% and paid coupons annually with a minimum
coupon rate is 0.25% per annum. The return floor for this product is �30%. The CDs mature
in seven years on February 28, 2019.

Assuming a 6% cap, we value the JP Morgan unbuffered contingent coupon CD at
$933.88 per $1,000 face value CD; �91% of the value of a contemporaneously issued
fixed-rate CD. In this case, investors received the minimum return in �80% of the simula-
tions. The increase in this statistic relative to the buffered example can be traced to the fact
that this CD includes a nonzero minimum coupon. Fig. 2 illustrates the component return
results for a given basket element return as of each coupon date for both the unbuffered
example and buffered examples discussed above.

4.2. Ratchet CD

A second common type of structured CD is the “ratchet CD,” also known as a “cliquet
CD.” The issuer of a ratchet CD observes the underlying asset’s price or level at several

225G. Deng et al. / Financial Services Review 23 (2014) 219–237



points during the term of the CD and calculates returns between each observation date. A
local cap, and sometimes a local floor, is applied to each of the observed returns. At maturity,
the observed returns are summed, and a global floor is then applied; usually a minimum
return or 0% (corresponding to a return of principal). Products with a local cap but without
a local floor are exposed to the risk that a large decline in the underlying asset during one
period could wipe out many periods of capped positive returns.

This structure is similar to the “Simple Ratchet Equity Indexed Annuity” of Hardy (2003).
Hsieh and Chiu (2007) show that for an initial investment P in an equity-indexed annuity
(EIA) with guarantee ratio �, minimum guaranteed rate g and term T that the value (VEIA)
of the contract is given by

VEIA � E�T��Pe�rT max�1 � R, ��1 � g�T�� (1)

where r is the (assumed) constant and continuously compounded riskless rate and R is the
arithmetic sum of returns given by

R � �i�1
nT min �max ��Ri, Rf�, Rc� (2)

Here n is the number of times per year that the underlying asset is observed, Ri �
S(ti)/S(ti–1) � 1, Rf is the local return floor, Rc is the local return cap and � is the
participation rate.25 For structured CDs, principal protection sets � � 1 and typically there
is no local floor (f � �1) or leverage (� � 1). Closed-form solutions for this type of EIA,
in the absence of the maturity guarantee, were derived by Hsieh and Chiu (2007).

Although this formula is general and fits nicely into our context, it does not appropriately
apply credit risk or FDIC insurance. Including the CDS rate in the overall discount factor
would be inappropriate given the guarantees provided by FDIC insurance (� � 1). We alter
Eq. (1) to include the credit risk related to components that are sensitive to market returns
and obtain the value of the structured CD (VSCD)

Fig. 2. Example unbuffered contingent coupon CD with CUSIP 48123Y6G6 (a) and example buffered contingent
coupon CDs with CUSIP 06740ARX9 (b).

226 G. Deng et al. / Financial Services Review 23 (2014) 219–237



VSCD � E�T��Pe�rT � Pe��r�c�T max�R,gT��
� Pe�rT � E�T��Pe��r�c�T max�R,gT��

❘
Market/credit risk

(3)

where c is the continuously compounded CDS rate of the issuer.
An example of a ratchet CD is the Barclays “Certificates of Deposit Linked to the

Performance of the S&P 500 Index due October 27, 2015” issued in October 2010 (CUSIP:
06740AMD8). At maturity, Barclays pays investors a return equal to the sum of the S&P 500
quarterly returns, each capped at between 3% and 5%, subject to a minimum return between
0.25% and 1.00% per annum. The product was priced on October 22, 2010 and will mature
on October 27, 2015.

Although this Barclays structured CD has a local cap, it has no local floor. That means that
a large negative quarterly return during the term of the note could wipe out the several
accumulated capped returns. For example, if the S&P 500 increases at a rate of 10% per
quarter for two years, then the CD would have accumulated 3% capped returns for eight
quarters. If the S&P 500 subsequently decreases by 24% or more in the next quarter, the total
of the capped quarterly returns for this nine quarter period would be zero or less, even though
the S&P 500 would have increased over 60% during this period.

Using the volatilities and market rates as of the pricing date and the midpoint for the terms
of the CD (4% local cap and 0.625% minimum return per annum), we obtain a valuation of
$976.46 per $1,000 face-value CD. We find that there is �97.2% chance of investors earning
the minimum return and the average return above the minimum return is �2% per annum.
Using the contemporaneous national average rates on traditional CDs from the FDIC,
extrapolated to the appropriate term for this CD (1.66%), we find that this CD is worth �95%
of a traditional fixed rate CD.

We varied the applied minimum coupon to the structure within the range specified in the
preliminary offering document (between 0.25% and 1.00%). For each minimum coupon rate,
we determined the local cap that will result in an estimated value equal to the value of the
structured CD assuming a 25 year old investor and plotted the results in Fig. 3.

For this particular structured CD, we find that for each basis point increase in the
minimum coupon, Barclays would likely require a basis point decrease in the local cap level
to achieve the same level of profit.

4.3. Average return (Asian) CD

A third common type of structured CD in our sample resembles an average value option,
also known as an Asian option. The issuer of an “average return” CD observes the underlying
asset at several (n) points during the term of the CD and then calculates the average of those
levels (Li) to determine the underlying return. A floor (Rf), cap (Rc), and/or participation rate
(�) may then be applied to the observed average return to yield the return at maturity

R � min �max ��

n
�i	1

n Li � L0

L0
, Rf� , Rc� (4)

227G. Deng et al. / Financial Services Review 23 (2014) 219–237



where L0 is the initial level of the underlying asset.
As with the other types of structured CDs, we can write the value of the structured CD

(VSCD) as a component free of credit risk and a component exposed to market/credit risk:

VSCD � E�T��Pe�rT � Pe��r�c�TR�
� Pe�rT � E�T��Pe��r�c�TR�

❘
Market/credit risk

(5)

where P is the face-value of the CD.
As an example, on December 23, 2010 SunTrust issued $13,235,000 worth of Index-

Linked Certificates of Deposit linked to the Dow Jones Industrial Average (CUSIP:
86789VLY1) maturing on December 20, 2016. The returns were subject to a minimum return
of 8% and a cap of 30%. Although the structured CDs only pay out at maturity, SunTrust
observes the Dow Jones Industrial Average quarterly and the resulting 24 observed levels are
averaged. Based on our valuation methodology, we find that this CD was worth approxi-
mately $932.46 when it was issued in December 2010, �96% of the value of a contempo-
raneously issued fixed-rate CD.

Averaging periodic levels reduces the value of Asian options compared with traditional
European options. The type of crediting formula used in the structured CDs we study here
is known as an “averaging-in” procedure, as it uses periodically observed levels, as opposed
to an “averaging-out” procedure that would average several observation made closer to
maturity (Bouzoubaa and Osseiran, 2010). “Averaging-in” includes levels observed through-

Fig. 3. Contour plot for Barclays ratchet structured CD (CUSIP: 06740AMD8).

228 G. Deng et al. / Financial Services Review 23 (2014) 219–237



out the term of the note, meaning that their average will incorporate returns with widely
differing terms. If the expected return on the index is positive, than on average the
“averaging-in” procedure will reduce the value of the product relative to an “averaging-out”
style product, all else constant.

4.4. Single-observation CD

The simplest structured CDs do not pay coupons but pay a contingent return at maturity
if the index increases during the term of the note.26 The contingent payout can be modeled
as payoffs to long-term call options on the linked asset. The return on these CDs is capped
at a maximum return, which is equivalent to including the payoffs to a short out-of-the-
money call option. The participation rate effectively changes the number of call options the
investor is long and short.

Let the participation rate be given by � and the maximum (or capped) return be given by
Rc and minimum (floored) return RF. The value of the structured certificate of deposit VSCD

with face-value P and term T is given by

VSCD � Pe�rT � E�T��Pe��r�c�T min�max��R,RF�,Rc��
❘

Risky Component

(6)

In Eq. (6), r is the risk-free rate and c is the CDS rate of the issuer. Because the
market-contingent component is not paid until maturity and these CDs are typically long-
dated, a significant liability could build up with respect to such a CD (especially if it is not
callable).

Eq. (6) can be written in terms of European call options. Let C(K,T) represent the value
of a European call option with strike price K, expiring at time T. The value of the structured
CD is then given by

VSCD � Pe�rT �
P

S0
e�cT��C�S0,T� � C�S0�1 �

Rc

�
� ,T�� � PRFe��r�c�T

❘
Market/credit risk

(7)

where S0 is the initial price/level of the linked asset.
As an example, consider the Citibank, N.A. structured CD due January 29, 2016 linked to

the S&P 500 (CUSIP: 172986BF7). The CD was priced on January 26, 2010 and issued on
January 29, 2010. It pays no interest during the six-year term and, at maturity, pays the return
of the S&P 500 subject to a maximum return of 41% (5.89% annual percentage yield) and
a minimum return of 0%. The issue size was $1,952,000.

Citibank reported a comparable yield of 3.18% when the average rate for nonjumbo
deposits at this time was 2.4% and for jumbo deposits was 2.44%.27 The difference between
the comparable yield on this CD and the average yield for fixed-rate deposits is close to the
75 basis point FDIC cap.

Using the terms of the contract and observable market variables, we obtain a valuation of

229G. Deng et al. / Financial Services Review 23 (2014) 219–237



$938.67 per $1,000 investment in the Citibank structured CD. We find that the structured CD
was worth roughly 95% of the value of a contemporaneously issued traditional CD at the
national average rate.28

We find that as of the issue date there was a 58% chance that the Citibank CD would return
no additional payment beyond return of principal investment and a 20% chance that the
maximum return would be paid to the investor. To provide intuition for the sensitivity of the
structured CDs value to the issuer’s discretionary parameters, we show in Fig. 4 the
likelihood that an investor would realize the maximum return and minimum return as a
function of the minimum return.

There is a very high probability of obtaining the minimum return when holding the CD to
maturity. There is only a small probability of returns that would lead to direct market
exposure (i.e., neither capped nor floored), even for a minimum return of zero.

4.5. Mortality risk and structured CDs

Structured CDs, like traditional CDs, have defined payouts if the owner of the CD dies
before maturity, which may differ from the current market value of the CD. The products in
our sample simply pay the original principal amount to the investor’s beneficiary; essentially,
the issuer has sold a binary put option to “buy back” the CD at par upon the investor’s death,
known in the literature as the “death put.” A return of principal years after the investor makes
the CD will be worth less than the initial principal amount because of the time value of
money. Therefore, the death put can be of negative value to the investor. However, the effect
of the death put may differ between types of structured CDs.

Fig. 4. Probability of realizing the minimum return and maximum return as a function of the return floor (CUSIP:
172986BF7).

230 G. Deng et al. / Financial Services Review 23 (2014) 219–237



We have modeled the death put for each of our product types using the unisex Annuity
2000 Mortality Table from the Transactions of the Society of Actuaries. In our Monte Carlo
simulations, for each year of the product’s term we randomly select a number of simulated
paths to credit with the death put payment based on the conditional death probabilities for an
investor of an assumed age at purchase. We then rerun our simulations for ages between 25
and 85. Fig. 5 shows the average value of each of the four types of structured CDs described
above for investors of varying ages.

To clarify Fig. 5, an average single-observation structured CD is worth approximately
$96.55 to a 25 year old investor. To an 85 year old investor, the average single-observation
structured CD is worth approximately $95.47, �110 basis points less. Therefore, the curve
for the single-observation structured CD in Fig. 5 slopes downward from the 100% value to
the terminal value of 98.9%.

For the single-observation structured CDs, the value decreases as a function of age
because the payout at death is worth less than the crediting procedure would otherwise
return. However, for the ratchet, average and contingent coupon types, the value actually
increases with age. Effectively, crediting formulas in these products reduce investors’
value more strongly than the time value of money up to death. Therefore, a return of
principal before maturity would be a more favorable outcome from the investor’s point
of view. For all product types, however, the size of the effect is small.29 Our results
suggest that the death benefit is of little value to investors, and can be either positive or
negative depending on product type.

Fig. 5. Average structured CD value normalized by the average value of the structured CD to a 25 year old
investor.

231G. Deng et al. / Financial Services Review 23 (2014) 219–237



4.6. Summary of valuation results

We have valued a total of 303 structured CDs with aggregate face value of $1.8 billion.
This valuation sample accounts for roughly 15% of the structured CDs in our sample (12%
by aggregate face value). Our results indicate that, on average, investors receive �93.8 cents
in value for each dollar invested in structured CDs.

Table 3 summarizes our valuation results across product types. The size of the valuation
sample is limited both by quality of the data and the idiosyncrasies of individual products.
For example, many products reference proprietary indexes for which there is little or no
market information, such as those noted above from JP Morgan.

Table 3 also summarizes the structured CD valuation as a percentage of contemporane-
ously issued fixed-rate CDs. The rates assumed for fixed-rate CDs is the national average rate
given by the FDIC for the week that the structured CDs were priced. In the event that the
term of the structured CD did not match that of a traditional CD reported by the FDIC, linear
interpolation was used to determine the implied CD rate. With the exception of the average
CDs, the other structured CDs were priced at significant discounts to contemporaneously
issued fixed-rate CDs according to our valuation results.30

Table 4 summarizes the valuation results for our sample of structured CDs broken down
by issuing bank.31 Average initial valuation is written as a percentage of face value. In the
table, we also include the probability that an investor will realize the minimum return from
investing in the products.32

As discussed in the contingent coupon CD section, contingent coupon CDs can come in
two varieties: buffered and unbuffered. Table 5 decomposes the valuation results across these
two types.

Our analysis suggests that these two types of contingent coupon CDs are comparably
priced and that issuers likely lower the return cap to compensate for the presence of a buffer
in the buffered version of the CDs.

In summary, these results show that structured CDs are priced at significant discounts to
face value. As noted in the structured products literature, the difference between the fair
value and the issue price is effectively an undisclosed additional charge by the issuer.
Although the amount of this charge varies by product type, it is approximately as large as the
discounts observed in the structured product market.33 Although structured CDs are often
marketed as safe investments with additional market exposure, our results show that their

Table 3 Issue date weighted average fair value by product type

Product type Number Aggregate face
value (MM)

Average initial
valuation

Percentage of
traditional CD

Contingent coupon CD 230 $1,028.0 93.18%a 92.44%
Ratchet CD 31 $ 308.2 96.23%a 95.12%
Average CD 28 $ 338.2 92.54%a 93.28%
Single-observation CD 14 $ 122.1 96.35%a 97.03%
Total 303 $1,796.4 93.80%a 93.37%

aIndicates that the mean ratio of structured CD values to traditional CD values is statistically different from 1
at the 95% confidence level.

232 G. Deng et al. / Financial Services Review 23 (2014) 219–237



unfavorable crediting formulas lead to both little market exposure and a high probability of
below-market returns.

5. Discussion

In this article, we review the market for and common features of structured CDs. We use
a sample of products to provide aggregate data on this relatively obscure, yet very large

Table 4 Issuer structured CD valuation summary

Issuer Number Aggregate face
value (MM)

Initial
valuation

Probability of
minimum return

Contingent coupon CDs
JP Morgan Chase Bank 194 $ 787.6 93.28%a 69%
Barclays Bank Delaware 31 $ 175.3 94.10%a 65%
Citibank 3 $ 42.2 90.77%a 54%
HSBC 1 $ 21.5 86.86% 66%
American National Bank 1 $ 1.3 93.45% 66%
Total 230 $1,028.0 93.18%a 68%

Ratchet CDs
Barclays Bank Delaware 12 $ 64.7 94.43%a 96%
SunTrust Bank 11 $ 114.9 97.33%a 97%
Union Bank 5 $ 123.2 96.14%a 95%
JP Morgan Chase Bank 3 $ 5.4 96.35%a 98%
Total 31 $ 308.2 96.23%a 96%

Average return CDs
SunTrust Bank 17 $ 230.6 93.03%a 94%
JP Morgan Chase Bank 9 $ 26.6 89.81%a 70%
Wells Fargo Bank 1 $ 73.2 92.83% 99%
Citibank 1 $ 7.8 84.81% 53%
Total 28 $ 338.2 92.54%a 93%

Single-observation CDs
JP Morgan Chase Bank 6 $ 54.9 95.48%a 61%
SunTrust Bank 3 $ 25.6 98.65% 76%
Barclays Bank Delaware 2 $ 16.3 96.83% 62%
Citibank 2 $ 24.3 95.67% 60%
Wells Fargo Bank 1 $ 1.0 93.86% 98%
Total 14 $ 122.1 96.35%a 64%

aIndicates that the mean ratio of structured CD values to traditional CD values is statistically different from 1
at the 95% confidence level.

Table 5 Valuation results for our sample of contingent coupon structured CDs with and without buffers

Return
type

Number Aggregate face
value (MM)

Initial
valuation

Probability of
minimum return

Buffered 196 $ 805.1 93.02%a 68%
Unbuffered 34 $ 222.9 93.78%a 68%
Total 230 $1,028.0 93.18%a 68%

aIndicates that the mean ratio of structured CD values to traditional CD values is statistically different from 1
at the 95% confidence level.

233G. Deng et al. / Financial Services Review 23 (2014) 219–237



financial market. We provide valuation procedures for several structured CD types included
in our sample and value a variety of products based upon these procedures.

Our primary findings are that (1) structured CDs include payoffs to complex derivative
positions; (2) structured CDs are worth significantly less than their issue price; (3) this
mispricing is present across product types and issuers; (4) the crediting formulas in structured
CDs have a high probability of crediting the minimum return; and (5) the death put can be
of positive or negative value to the investor depending on product features and type. We find
that most of the apparent benefits of structured CDs are illusory and overwhelmed by their
inherent risks and embedded fees. Each of these results has important implications for
individual investors and financial advisors.

To our knowledge, our study is the first to examine the structured CDs market in detail,
but is limited by the few publicly available sources for offering documents. In addition, there
are a significant number of products (even in our limited sample) that reference proprietary
indexes for which there is little or no market data. This could represent an information
asymmetry or conflict of interest between investors and issuers of these CDs. We have also
noticed a variety of “exotic” structured CDs that have extremely complex payout formulas
and terms as long as 15 years. These may be fruitful topics for further research.

Because the structured CD market is entirely over-the-counter, customers may not be able
to compare different offerings and make informed decisions about the relative value of a
particular structured CD, even with information such as “comparable yields” or original
issuer discounts. Structured CDs can be extremely complex investments and, in many ways,
are just as complex as structured products, which are regulated much more stringently and
do not enjoy FDIC. We think it worth careful consideration whether such products should be
sold to retail investors, especially in an unregulated setting.

Until the regulatory framework for structured CDs becomes clear, financial advisors can
help fill this information gap between issuers and individual investors. In particular, investors
must appreciate the often limited nature of the market-linked exposure, the significance of
FDIC insurance, and the potential for significant mispricing. These issues are likely to be the
subject of continued debate as the market for structured CDs continues to expand and evolve.

Notes

1 See Bloomberg Structured Notes Brief (2011a), Bloomberg Structured Notes Brief
(2011c), Bloomberg Structured Notes Brief (2011d), and Bloomberg Structured
Notes Brief (2011e).

2 See Bloomberg Structured Notes Brief (2011b).
3 Regulation of Retail Structured Products, International Organization of Securities

Commissions, April 2013.
4 See for example Henderson and Pearson (2010), Deng et al. (2010), Deng et al.

(2011b), Deng et al. (2011a), and Deng et al. (2012).
5 The authors advocate the use of the underlying asset and futures contracts to

“dynamically hedge” the position because of high liquidity and low transaction costs.
6 Retrieved from http://www.sec.gov/answers/equitylinkedcds.html.

234 G. Deng et al. / Financial Services Review 23 (2014) 219–237



7 See Robinson (2012).
8 Retrieved from http://www.fdic.gov/consumers/consumer/news/cnspr12/market

linkedcds.html.
9 NYSE Information Memos No. 06-12, March 17, 2006.

10 This sample was collected mainly from Bloomberg using the SRCH function,
performed on March 13, 2013. The sample was augmented with additional structured
CDs found from other publicly available sources.

11 We include Wachovia structured CDs in the statistics of structured CDs issued by
Wells Fargo.

12 According to a July 2012 Strategy Guide, the JPMorgan ETF Efficiente 5 Index
represents the returns to a basket of twelve exchange traded funds (ETFs) that are
rebalanced each month to reflect the allocation that maximizes return for a given
level of volatility based on the previous six months of historical data. According to
a September 2009 Strategy Guide, the JPMorgan Optimax Market-Neutral Index
represents the returns to a basket of 18 to 24 commodity indexes (sub-indices of the
S&P GSCI). The allocation to each commodity is selected by a proprietary algorithm
that is “based on modern portfolio theory and momentum theory.”

13 Wells Fargo Securities, Market Linked Certificates of Deposit, accessed July 2, 2013.
14 Structured product offering documents are publically available from the SECs

EDGAR database, filed as form 424B2.
15 Retrieved from http://www.investor.gov/investing-basics/investment-products/

certificates-deposit-cds.
16 NYSE Information Memos No. 06-12, March 17, 2006.
17 FDIC Rules and Regulations §337.6.
18 Retrieved from http://www.fdic.gov/regulations/resources/rates/index.html.
19 The difference between the rates for jumbo and non-jumbo five year CDs is closer to

eight basis points according to BankRate.com data.
20 Market-Linked CDs: Don’t Let the Possibility of Higher Returns Cloud Your View of

the Potential Risks, FDIC Consumer News, Spring 2012.
21 The maximum deposit insurance amount was increased from $100,000 to $250,000

by the Emergency Economic Stabilization Act of 2008; see Public Law 110-343
§136(a)(1). Subsequently, the Helping Families Save Their Homes Act of 2009
extended this increase through December 31, 2013; see Public Law 111-22 §204(a)
(1) (A). In 2010, the Wall Street Reform and Consumer Protection Act made this
increase permanent; see Public Law 111-203 §335(a).

22 The FDIC has taken the position in the past that a potential decrease in principal
caused by the imposition of an early withdrawal penalty does not prevent a product
from qualifying as a ‘deposit’ for insurance purposes. FDIC Letter to Kevin P.
Murray, February 27, 2002.

23 Issuers have a disincentive to maintain such markets, as developed secondary mar-
kets may trigger more stringent regulatory requirements.

24 We note that for some periods of time, CD rates were above treasury rates of the
same term. Because we discount both structured and traditional CDs using treasury

235G. Deng et al. / Financial Services Review 23 (2014) 219–237



rates, the value of traditional CDs may exceed face value. On February 29, 2012,
implied seven year CD rates were higher than seven year treasury rates.

25 Eq. 2 is a trivial generalization of the formula in Hsieh and Chiu (2007) to include
more frequent observations.

26 A CD of this type was described by Edwards and Swidler (2005).
27 The FDIC currently reports rates on fixed-rate CDs out to a maturity of 60 months

on a weekly basis. We are using the rates observed by the FDIC on January 25, 2010
and using linear extrapolation to determine the implied rate on a 72 month CD.

28 Retrieved from http://www.fdic.gov/regulations/resources/rates/index.html.
29 The value for a 25 year old investor is only 1.1% higher for the single-observation

products, 0.4% lower for the ratchet products, 0.5% lower for the average products,
and 1.5% lower for contingent coupon CDs when compared with the value for an 85
year old investor.

30 Retrieved from http://www.fdic.gov/regulations/resources/rates/index.html.
31 For some smaller banks (e.g., Union Bank), CDS data is unavailable as of this

writing. To include the effects of credit risk for these banks, we averaged the CDS
rates for all banks that issued structured CDs within our sample.

32 For these calculations, we exclude investors who have died during the term of the
notes to estimate the probability of obtaining the minimum return (since these
investors do not hold the CDs to maturity).

33 Bloomberg Structured Notes Brief (2011f).

References

Black, F., & Scholes, M. (1973). The pricing of options and corporate liabilities. Journal of Political Economy,
81, 637–654.

Bloomberg Structured Notes Brief. (2011a). Bonacci of CD Funding Securities on popular kinds of structured
CDs, growth in market. Bloomberg Structured Notes Brief, page 6.

Bloomberg Structured Notes Brief. (2011b). David Fitzsimmons on choosing which structured CDs to sell and
working with issuers. Bloomberg Structured Notes Brief, page 4.

Bloomberg Structured Notes Brief. (2011c). Wells Fargo’s Ginsburg on selling structured CDs to crisis-weary
investors. Bloomberg Structured Notes Brief, page 4.

Bloomberg Structured Notes Brief. (2011d). Stoddart of Capital Investment Marketing on selling structured CDs
through bank branches. Bloomberg Structured Notes Brief, page 5.

Bloomberg Structured Notes Brief. (2011e). Barclays’s Wu and Gordon on commodities, CDs and selling in Latin
America. Bloomberg Structured Notes Brief, page 5.

Bloomberg Structured Notes Brief. (2013f). Banks are inconsistent on new sec disclosure rules. Bloomberg
Structured Notes Brief, pages 3–5.

Bouzoubaa, M., & Osseiran, A. (2010). Exotic options and hybrids: A guide to structuring, pricing and trading.
Wiley Finance.

Chance, D., & Broughton, J. (1988). Market index depository liabilities: Analysis, interpretation, and perfor-
mance. Journal of Financial Services Research, 1, 335–352.

Deng, G., McCann, C., & O’Neal, E. (2010). What TiVo and JP Morgan teach us about reverse convertibles.
SLCG working paper (available at http://www.slcg.com/).

Deng, G., Guedj, I., Mallett, J., & McCann, C. (2011a). The anatomy of absolute return barrier notes. The Journal
of Derivatives, 19, 61–70.

236 G. Deng et al. / Financial Services Review 23 (2014) 219–237



Deng, G., Mallett, J., & McCann, C. (2011b). Modeling autocallable structured products. Journal of Derivatives
4 Hedge Funds, 17, 326–340.

Deng, G., Dulaney, T., & McCann, C. (2012). Valuation of reverse convertibles in the Variance Gamma
economy. SLCG working paper (available at http://www.slcg.com/).

Edwards, M., & Swidler, S. (2005). Do equity-linked certificates of deposit have equity-like returns? Financial
Services Review, 14, 305–318.

Glasserman, P. (2003). Monte Carlo methods in financial engineering. Berlin: Springer.
Hardy, M. (2003). Ratchet equity indexed annuities. University of Waterloo Working Paper.
Henderson, B., & Pearson, N. (2010). The dark side of financial innovation: A case study of the pricing of a retail

financial product. The Journal of Financial Economics, 100, 227–247.
Hsieh, M., & Chiu, Y. (2007). Monte Carlo methods for valuation of ratch equity indexed annuities. Proceedings

of the 2007 Winter Simulation Conference.
King, S., & Remolona, E. (1987). The pricing and hedging of market index deposits. Federal Reserve Bank New

York Quarterly Review, 11, 9–20.
Merton, R. (1973). Theory of rational option pricing. Bell Journal of Economics and Management Science, 4,

141–183.
Robinson, M. (2012). FINRA examines CDs tied to derivatives as sales surge to record. Bloomberg News,

February 7, 2012.

237G. Deng et al. / Financial Services Review 23 (2014) 219–237