Bloomberg Barclays Indices
A Bloomberg Professional Service Offering
INTRODUCTION TO BLOOMBERG BARCLAYS MIRROR FUTURES AND
DURATION HEDGED BENCHMARK INDICES
Duration Hedged Indices: Motivation and Overview ........................................... 3
MIRROR FUTURES INDEX (MFI) METHODOLOGY ........................................... 5
MFI Construction and Returns for a Single Currency Index ................................. 5
MFI Construction for a Multi-Currency Index......................................................... 8
Assessing MFI Performance .................................................................................... 14
DURATION HEDGED INDEX (DHI) METHODOLOGY ...................................... 19
DHI Construction ..................................................................................................... 19
DHI Tracking Performance ..................................................................................... 20
Partially Hedged DHI ............................................................................................... 21
Other Customized MFI and DHI ............................................................................ 23
Conclusion ............................................................................................................... 24
APPENDIX: ACCESSING MIRROR FUTURES AND DURATION HEDGED
INDICES ................................................................................................................... 25
15 January 2020
Bloomberg Barclays Mirror Futures and Duration Hedged Indices
2
Introduction to Bloomberg Barclays Mirror Futures and
Duration Hedged Benchmark Indices
Mirror Futures Indices (MFI) and Duration Hedged Indices (DHI) are used by investors
seeking to adjust the duration of their fixed income benchmarks while preserving the broad
coverage and diversification of their existing fixed income investment set. An MFI is a basket
of Treasury futures contracts designed to match the duration exposure of a Bloomberg
Barclays Index. A DHI combines a cash index with a short MFI position, so as to reduce the
Treasury duration exposure of the cash index. These new indices may be used as
replacements for existing portfolio benchmarks, reference indices for various replication
strategies, or informational measures of interest rate duration-hedged (fully or partially) bond
market returns.
• Mirror Futures Index: An MFI is an index whose return reflects a funded set of Treasury
futures contracts, weighted so as to match closely the beginning-of-the-month option-
adjusted duration (OAD) profile of an underlying standard bond index. For example, the
US Aggregate MFI will include five US Treasury futures contracts weighted to match the
OAD profile and market value of the cash US Aggregate Index, plus a cash investment (a
“funding component”) in US T-bills. We chose to include a funding component in the MFI
to make its returns more comparable with those of a traditional cash index, which assumes
a funded investment in the index bonds.
• Duration Hedged Index: A funded index whose return reflects the return on the
underlying cash index, with its OAD exposure hedged (fully or partially) using its MFI. For
example, the US Aggregate DHI is the US Aggregate less its MFI, plus the MFI’s funding
component added back.
MFI and DHI returns will be available on currency unhedged and hedged bases. We publish
them for select standard Bloomberg Barclays Indices, but the methodology discussed in this
publication can be applied to almost any standard or customized Bloomberg Barclays Index.
Current coverage includes:
• USD Indices: US Aggregate Index, US Government/Credit Index, US Credit Index, US
Corporate Index, US Treasury Index, US Fixed-Rate MBS Index, and US TIPS Index (Series-
L).
• Global Indices: Global Aggregate Index, Global Corporate Index, Global Aggregate x-
USD Index, Global Aggregate Ex-JPY Index, Global Corporate Ex-USD Index, Global
Corporate Ex-JPY Index, Global Treasury Index, and Global Credit Index.
• Pan Euro Indices: Euro Aggregate Index, Euro Corporate Index, Sterling Aggregate
Index, and Sterling Corporate Index.
Clients seeking an index that retains some interest rate duration exposure can also request a
DHI that scales back the degree of duration hedging but does not fully hedge it. For example,
a client may prefer to remove 50% of the duration exposure of the US Aggregate and request
a US Aggregate 50% DHI. Additionally, constant duration benchmarks (e.g., a US Aggregate
Duration of 3 Index) can be created using the MFI/DHI methodology.
Bloomberg Barclays Mirror Futures and Duration Hedged Indices
3
Duration Hedged Indices: Motivation and Overview
Exposure to the Treasury yield curve is a dominant driver of total returns for most fixed-income
indices. For example, the US Aggregate Index posted year-to-date total return of 8.52% (as of
30 September 2019) with reported excess (spread) return of 1.11%, which implies curve return
of 7.41% (total return minus the excess return).
In some circumstances, investors may prefer a standard index with its Treasury term structure
return component removed. These are the DHI. For example, those worried about increases
in Treasury yields may want a benchmark without Treasury term structure exposure while
continuing to capture spread carry and the spread change component of returns.
Advantages of a DHI vs. a Shorter-Maturity Cash Index
Of course, investors worried about duration risk could simply change to a shorter duration
standard benchmark, say, move from the US Aggregate to the 1-3y US Aggregate. However,
there are certain advantages to a DHI. First, it will target a zero duration, whereas the 1-3y
Aggregate still has meaningful Treasury duration exposure (2.48y as of 30 September 2019).
Second, a DHI does not disrupt the existing portfolio management process. A portfolio
manager who was originally managing against the US Aggregate can continue to do so, with
the same broad range of index issues from which to select (approximately 10,900 in the
Aggregate vs. 2,200 in the 1-3y Aggregate), the same ability to overweight/underweight
duration, and the same ability to overweight/underweight sector allocation and spread
duration. In addition, with a DHI, the asset owner, presumably happy with the manager’s
performance, can continue to profit from the manager’s full set of skills while knowing that the
Treasury baseline duration of the portfolio is approximately zero. The asset owner does not
have to search for a new short-duration manager. Similarly, the investment manager can avoid
losing satisfied, but duration-concerned, clients to other low duration or alternative products
(e.g., cash or equities).
Third, DHI returns contain the tracking error of using a basket of Treasury futures (i.e., the MFI)
to remove an index’s return component arising from changes in the Treasury curve. For the
Treasury sector of an index, the Treasury component of return is that on the bonds themselves.
However, for non-Treasury bonds, the index defines this component as the return on a KRD-
matched portfolio of hypothetical par Treasury bonds. Depending on the set of available
Treasury futures and the stability of the cash-futures basis, there is bound to be some tracking
error between the return on the duration-matched futures basket and the index’s Treasury
component of returns.
For investment managers, the advantage of a DHI benchmark is that any tracking error
between futures and the index Treasury component of return is part of the benchmark’s
return. The manager is not responsible for how well the futures track the Treasury component
of returns. Instead of specifying a benchmark duration and making the manager bear the
tracking error of adjusting an index to a client’s target duration, the DHI benchmark will reflect
that change. This means that as long as the Aggregate manager mimics the futures basket
overlay in the DHI, his or her performance can continue to be measured cleanly versus the
Aggregate index, while the asset owner enjoys a benchmark with its desired duration but
bears the tracking error volatility arising from using futures to reduce duration.
Investors worried about
increases in Treasury yields
may want a benchmark
without Treasury term
structure exposure while
continuing to capture spread
carry and the spread change
component of returns.
Bloomberg Barclays Mirror Futures and Duration Hedged Indices
4
DHI Design Overview
As discussed, to construct a DHI, we must first identify and remove the Treasury component
of a cash index’s return. To do so, we use the MFI for the cash index. MFIs build on a long
history of research on index replication using futures and are similar in concept to our Mirror
Swap Indices.
An MFI does not synthetically replicate the total returns of a cash index, only the Treasury
component. It holds a funded (using T-bills) basket of Treasury futures contracts weighted to
match the Treasury OAD exposure of the underlying cash index. The MFI futures basket is
rebalanced monthly, just like the underlying cash bond index. As we will show, an MFI follows
a straightforward construction methodology and is easily replicable.
A DHI is defined as the total return on the underlying cash index less the return of its
corresponding MFI, with the MFI funding component added back. A DHI is a funded index,
just like its underlying cash index. However, since the MFI is also a funded index, subtracting
its return from the cash index return would make the DHI an unfunded index. This is why we
add back the MFI funding component of return to the DHI. Without this, the DHI should closely
track the reported excess returns of the cash index, which are calculated as the difference
between two funded returns (the index return and the duration-matched Treasury portfolio
return).
1
Other Uses for Mirror Futures Indices
In addition to its use to construct a DHI, an MFI is also a reference index on its own for
access/replication products. For example, in total return swap form, a manager could
receive/pay the total return on the replicable MFI basket (and pay/receive a short funding rate
(e.g., 1m Libor +/- a spread). Portfolio managers may find such swaps, which relieve them from
having repeatedly to size the individual futures positions and rebalance, to be particularly
useful. For example, a TIPS manager wishing to make an explicit breakeven view, without the
nominal term structure exposure, could overlay a pay-TIPS MFI swap on his or her cash TIPS
index portfolio. Or a hedge fund could combine its alpha return with a receive-Aggregate MFI
swap to compete for traditional FI mandates. Similarly, a credit fund could overlay its credit
alpha on a receive-Credit MFI swap to compete for corporate mandates.
We first present the MFI methodology, an intuitive and rules-based approach for matching
duration exposure. For several cash indices, we then show how well an MFI has been able
historically to track the Treasury component of an index’s returns. We then illustrate the DHI
construction process and show how DHI returns (adjusted for funding) have tracked the
underlying cash index’s excess returns.
1
Reported index excess returns equal the return difference between the cash index and a portfolio of hypothetical Treasury bonds that match the
KRD exposure of the index. Since the cash index and the duration-matched Treasury portfolios are both funded, the excess return is an unfunded
return.
An MFI does not synthetically
replicate the total returns of a
cash index, only the Treasury
component of the returns.
The MFI design makes it a
prime candidate as a
reference index for
access/replication products.
Bloomberg Barclays Mirror Futures and Duration Hedged Indices
5
Mirror Futures Index (MFI) Methodology
We first describe the MFI construction process for a single currency domestic index (e.g., the
US Aggregate). Then we describe MFI construction for global indices containing multiple
currencies (e.g., the Global Aggregate).
Bloomberg Barclays MFIs are designed to be straightforward and intuitive for a wide variety
of investors, while remaining flexible and adaptable enough to offer bespoke solutions to
meet an investor’s specific constraints, preferences, or guidelines. As such, the underlying
index mechanics and methodology adhere to design principles required for any good
benchmark:
• Representative of the market being measured and the desired risk exposures of its users;
• Replicable, without unnecessary turnover and transaction costs; and
• Objective and transparent, with clearly defined and objective rules and visibility into
current index composition and future composition during rebalancing periods.
MFI Construction and Returns for a Single Currency Index
1. Index OAD bucketing and futures contract selection
MFI construction starts with sorting the underlying cash bond index (Statistics Universe) at the
beginning-of-the-month (BOM) into various non-overlapping OAD buckets. OAD refers to the
option adjusted analytical treasury duration, as computed and reported by Bloomberg
Barclays index analytics.
• For an index containing bonds denominated in a single currency, each available and
liquid treasury futures contract (front-month contract in the quarterly cycle) is assigned to
one (and only one) OAD bucket.
2
• The underlying single currency cash index will be divided into OAD buckets depending
on the number of available contracts in the currency. Breakpoints are selected so that the
OAD of each bucket is close to that of its matching futures contract and there is a non-
trivial amount of index market value in the bucket.
2. Weight Calculations – Single Currency MFI
We then calculate the market equivalent value of futures contracts needed in the MFI so as to
match the contribution to the OAD of the futures position with that of the corresponding
index’s OAD bucket.
3
• For each eligible futures contract, we use Bloomberg analytics for its estimated treasury
duration exposures.
4
These are computed in a manner consistent with Bloomberg’s
duration calculation for cash bonds.
2
An index’s OAD bucketing scheme will remain constant unless there are dramatic changes in a futures contract deliverable or the availability of
new liquid futures contract markets.
3
A futures position at initiation has no market value. So to measure the contribution to OAD of a futures position, we use its hypothetical market
value (or, its market equivalent value). A futures position’s market equivalent value equals the futures price multiplied by the contract size. For
example, the contract size for the US Treasury bond futures contract is $100,000. If the futures price is 112-00, then the market equivalent value of
one futures contract is assumed to be $112,000. If the OAD of the futures contract were 9.0 and there was a 10bp increase in the yield of the
underlying bond, then the estimated market equivalent value of the contract would decline to $110,992.
4
Prior to August 16, 2018, Barclays POINT® analytics were used for futures contracts. Prior to October 2013, we used the sum of the futures KRDs
to estimate the OAD for the futures contract for data completeness.
Bloomberg Barclays Mirror Futures and Duration Hedged Indices
6
• The MFI weights are then fixed for the ensuing month.
• At the next monthly rebalancing, we repeat this exercise. If the next month is the futures
contracts’ expiration month, we roll the futures contracts to the next quarterly expiration
month. For example, a contract expiring in December will be rolled to the contract
expiring the following March at the end of the last business day of November.
Example: US Aggregate MFI
• For USD-denominated indices, we have identified five liquid Treasury futures contract
markets (TU, FV, TY, US, and WN)
5
. Figure 1 shows the OAD break points for each of these
contracts and the total market value and market-value weighted OAD for each OAD bucket
of the US Aggregate (Statistics Universe) as of the last business day in September 2019.
FIGURE 1
US Aggregate Mirror Futures Index Constituents, 30 September 2019
Bucket
Index
OAD MV ($mn) MV %
Cont. to
OAD
Futures
Contract
Futures
OAD
MFI
Weight
Total 5.776 23,144,356 100.00% 5.776
OAD 0-3y 2.044 9,051,448 39.11% 0.799 TUZ9 1.903 42.00%
OAD 3-5y 3.840 6,012,604 25.98% 0.997 FVZ9 4.185 23.83%
OAD 5-7.5y 6.162 3,070,685 13.27% 0.818 TYZ9 6.402 12.77%
OAD 7.5-15y 10.625 2,350,339 10.16% 1.079 USZ9 12.518 8.62%
OAD 15y+ 18.123 2,659,279 11.49% 2.082 WNZ9 18.954 10.99%
US T-bills Stub Position
1.79%
100.00%
Note: each futures contract’s weight is a funded position, where principal is invested at the same weight in short
maturity US T-bills
Source: Bloomberg
• The US Aggregate Index (Statistics Universe) had 39.11% of its market value in the 0-3y
OAD bucket (as of 30 September 2019).
• The OAD of this index bucket was 2.044, for a contribution to OAD of 0.799. We assign
the 2y Treasury futures contract (TU) to this bucket. We use December expiration futures
contracts up until the end of November.
• Since the contract had an estimated OAD of 1.903, the MFI needs a market equivalent value
percentage weight in the futures contract (i.e., futures price × # of contracts × notional size
of the futures contract / market value of cash index) so that the contribution to OAD of the
futures position matches that of the index’s 0-3y OAD bucket.
• As of 30 September 2019, the weight of the 2y futures contract position in the US
Aggregate MFI was 42.00%:
MFI
TUZ9 weight
= ContrOAD
OAD 0-3y bucket
/ OAD
2y futures
= 0.799 / 1.903
= 0.4200
• Each futures position in the MFI is a funded position. In other words, the market equivalent
value weight of each futures position in an MFI is paired with an identical market value
5
These are Bloomberg tickers. For historical analysis, the relatively new Ultra-Long futures contract market was not added to the MFI until January
2010.
Bloomberg Barclays Mirror Futures and Duration Hedged Indices
7
weight investment in short-maturity Treasury bills.
6
Consequently, 42.00% of the US
Aggregate MFI’s market value weight, corresponding to the 39.11% market value weight
of the 0-3y OAD bucket of the US Aggregate, contains long positions in TUZ9 futures
contracts and short-maturity Treasury bills.
7
• We repeat this exercise for all OAD buckets. If the market value weights of the five MFI
positions do not sum to 100%, we then add/subtract a “stub” investment in short-maturity
Treasury bills, with an assumed OAD of zero, so that not only does the duration of the MFI
equal that of the index, but also that the market value percentages of all MFI positions
sums to 100%.
3. Return Calculations and Funding Assumptions – Single Currency MFI
• The month-to-date return of an MFI futures position is a price return calculated by dividing
the current closing futures price by the BOM closing futures price, and subtracting 1.
8
This is
the percentage P&L (mark-to-market) per futures position, to which we refer as the “unfunded”
futures return, as there is no initial cash investment.
• To this unfunded return we then add the MTD return on the short-maturity Treasury bill
position to arrive at a MTD “funded” total return for this futures position. These funded
returns are then summed across the MFI futures positions, using the BOM MFI weights, to
compute the MTD MFI total return (Figure 2).
• We chose to make the MFI a funded index to be similar to our usual family of indices, as well
as our mirror swap indices. Also, subtracting the MFI from its cash index produces an excess
return that can be directly compared with our reported index excess returns.
FIGURE 2
Month-to-Date US Aggregate MFI Total Return, 31 October 2019
Futures Funding (T-bills)
Contract Weight MTD Return (%) Weight MTD Return (%)
TUZ9 42.00% 0.047 42.00% 0.156
FVZ9 23.83% 0.046 23.83% 0.156
TYZ9 12.77% -0.012 12.77% 0.156
USZ9 8.62% -0.578 8.62% 0.156
WNZ9 10.99% -1.124 10.99% 0.156
Stub Position
1.79% 0.156
MTD MFI Return 100.00% 0.012
Note: Weights of the unfunded futures positions do not necessarily sum up to 100%.
Source: Bloomberg
6
Presently we use a composite index of short maturity Treasury bills to represent the funding component of the MFI. If Treasury bills are not
available, we use short maturity LIBOR or deposit rates.
7
We assume the short-maturity Treasury bill position has zero duration.
8
The closing futures price is the Exchange’s official 2pm (local time) closing price. Unlike our swap index returns, the futures return does not
subtract 1bp/m.
Bloomberg Barclays Mirror Futures and Duration Hedged Indices
8
MFI Construction for a Multi-Currency Index
The basic design for an MFI on a multi-currency benchmark index follows that of a single
currency index. However, to hedge the interest rate exposure of local currency debt markets
that do not have active futures markets, further adjustments need to be made.
1. Index OAD bucketing and futures contract selection
Local Currency Mappings to Markets with Liquid Futures Contracts
The Global Aggregate Index contains 25 different currencies (as of 30 September 2019)
9
.
However, only a few of these have liquid Treasury futures contracts. To construct an MFI for
multi-currency indices, we map the market value of a “non-futures” currency to a “futures”
currency (i.e., a currency with a liquid Treasury futures market).
• We identify six currencies with liquid Treasury futures markets: the USD, GBP, JPY, CAD,
AUD and EUR.
10
Figure 3 identifies the mapping of “non-futures” currencies to “futures”
currencies for the standard Global Aggregate MFI.
11
Figure 4 provides details of the
Global Aggregate MFI market value mapped from “non-futures” currencies to the six
“futures” currencies as of 30 September 2019.
12
• Approximately 6.8% of the Global Aggregate’s market value is mapped to a non-native
currency (as of 30 September 2019).
13
FIGURE 3
Currency Mapping of “Non-futures” Currencies to “Futures” Currencies for the Global
Aggregate MFI
“Futures” Currency Mapped Global Agg Eligible “Non-Futures” Currency
USD
CLP, MXN
EUR
CHF, CZK, DKK, HUF, ILS, NOK, PLN, RUB, SEK
JPY
CNY, HKD, IDR, KRW, MYR, SGD, THB
AUD
NZD
CAD
GBP
Source: Bloomberg
9
Though broad-based in its coverage of local currency debt markets, ~92% of the overall universe is denominated in USD, EUR, JPY, GBP, or CAD
on a market value weighted basis.
10
Any new local currency debt markets that are added to the Global Aggregate through the annual index governance review will be mapped at
that time with Americas mapping to USD, EMEA mapped to EUR, and Asia mapped to JPY. Additional non-investment-grade local currency EM
markets would follow a similar mapping.
11
This currency mapping is the same as that for our standard Replicating Bond Index (RBI) baskets.
12
We assume that an investor excluding a specific local currency bond market from a cash bond index does not want any exposure to that same
futures market arising from currency mapping. Therefore, if the underlying cash bond index excludes one of the four futures currencies that have
other currencies mapped to it (USD, EUR, JPY, and AUD), we re-map to another futures currency. The market value of Global Agg currencies
mapped to JPY, EUR, and AUD will be remapped to USD if those underlying futures currencies are not benchmark eligible. Currencies mapped to
USD will be remapped to EUR.
13
Because many issuers borrow in non-native hard currencies, the mapping of issuer country of risk to futures is less close, including non-Germany
Eurozone countries that are mapping to German futures (e.g., France’s 5.4% market value weight is mapped to the EUR futures contract markets).
Only 65.6% of the Global Aggregate market value (30 September 2019 – Statistics Universe) is mapped to a country whose treasury instruments
underlie the treasury futures contract markets for its currency.
Bloomberg Barclays Mirror Futures and Duration Hedged Indices
9
FIGURE 4
Currency Mapping Details from “Non-futures” Currencies to “Futures” Currencies,
Global Aggregate MFI, 30 September 2019
“Futures”
Currency Currency
MV% in
Global Aggregate
MV% of “Futures”
Currency in
Global Aggregate
USD USD 45.44% 45.79%
CLP 0.06%
MXN 0.28%
EUR EUR 23.59% 25.53%
CHF 0.54%
CZK 0.09%
DKK 0.23%
HUF 0.08%
ILS 0.13%
NOK 0.10%
PLN 0.20%
RUB 0.17%
SEK 0.41%
JPY JPY 15.82% 20.24%
CNY 2.08%
HKD 0.02%
IDR 0.31%
KRW 1.16%
MYR 0.29%
SGD 0.19%
THB 0.36%
AUD AUD 1.20% 1.31%
NZD 0.11%
GBP GBP 4.56% 4.56%
CAD CAD 2.58% 2.58%
Source: Bloomberg
Selecting Eligible Contracts for Each Futures Currency
• Figure 5 shows the specific futures contracts selected for the six “futures” currencies used
in Global Aggregate MFIs.
14
In total, we use 14 different futures contracts for the Global
Aggregate: five for USD
15
; four for EUR; two for AUD; and one each for GBP, JPY, and
CAD. The number of contracts for each “futures” currency equals the number of duration
14
The list of eligible futures contracts and mapping of non-futures currencies is governed and reviewed
annually (as of 31 May each year) by the
Bloomberg Index Group. Any changes to the eligible contract list or currency mappings will be announced on INP <Go> and take effect as of the 30
June rebalancing for all MFI. Eligible contracts and new contract candidates are evaluated on a case-by-case basis for their relative liquidity and
investability using metrics including but not limited to trading volume and open interest. In between annual rebalancing dates, the Bloomberg Index
group may remove or replace an existing contract if there has been a sustained market disruption or impairment to liquidity that affects an investor’s
ability to replicate, particularly in a roll month. Any substitutions based on market disruptions will be announced on INP <Go> and made at the ensuing
month-end rebalancing with a minimum of five business days’ notice.
15
The Ultra-Long USD futures contract was used beginning in February 2010. Prior to that date, we used only four contracts/buckets with all
exposure in the 7.5y+ bucket mapping to the Long Bond futures contract.
Bloomberg Barclays Mirror Futures and Duration Hedged Indices
10
buckets applied to the corresponding underlying cash bond index component of the
Global Aggregate.
• While some “futures” currencies have additional Treasury futures contract markets,
trading volume and open interest are often too low and are therefore not selected to be
a constituent of an MFI.
• We map only the market value of the “non-futures” currency to the “futures” currency. For
example, we add the “non-futures” NZD market value to the “futures” AUD market value.
We do not map the OAD profile of the NZD market to the OAD profile of the AUD market.
Effectively, both the AUD market value and term structure in the MFI are augmented by
the NZD market value.
FIGURE 5
Eligible Contracts for each “Futures” Currency, 30 September 2019
Currency Bloomberg Id
Futures Contract
Market Description
Open
Interest
Assigned
Duration
Bucketing
USD
TUZ9 2y Treasury Note
3,658,352
0-3y
FVZ9 5y Treasury Note 4,149,985 3-5y
TYZ9 10y Treasury Note
3,580,823
5-7.5y
USZ9 30y Treasury Bond 971,388 7.5-15y
WNZ9 Ultra-Long Treasury Bond 1,139,701 15y+
EUR
DUZ9 Euro Schatz 1,432,903 0-3y
OEZ9 Euro Bobl
1,073,740
3-6y
RXZ9 Euro Bund 1,392,674 6-13y
UBZ9 Euro Buxl 221,330 13y+
JPY
JBZ9 10y JGB 99,891
GBP
G Z9 Long Gilt 636,088
AUD
YMZ9 3y T-Bond
1,139,844
0-3y
XMZ9 10y T-Bond 1,303,368 3y+
CAD
CNZ9 10y Canadian Gov’t Bond 571,131
Source: Bloomberg, CME, EUREX, TSE, NYSE EURONEXT (LIFFE), ASX, and TMX
Funding Instruments for “Futures” Currencies
For the six “futures” currencies, we identify their short-maturity funding instruments.
• For the G4 currencies (USD, EUR, GBP, JPY), we represent these funding instruments with
short-term treasury bills issued by the corresponding countries.
16
For the “futures”
16
The T-Bill universe used for this calculation is a rules-based index that may include one or more securities and is designed to achieve a target
maturity of 1m, while ensuring that there is at least one eligible security for the length of the index history. Some investors may prefer a single, large
on-the-run T-Bill issue to be used over a 1m T-Bill basket for replication purposes. This is being evaluated as a potential future enhancement, but is
not available as of the index launch. Any futures changes to the funding assumptions will be made on a forward basis and announced in advance.
Bloomberg Barclays Mirror Futures and Duration Hedged Indices
11
currency AUD, we use the 1m AUD deposit rate. For CAD, we use the Canada Bankers
Acceptances 1m rate. Figure 6 shows the current funding definition for each of the six
“futures” currencies.
FIGURE 6
Funding Definition for Each “Futures” Currency in the Global Aggregate MFI
“Futures” Currency
Funding Instrument
USD US short-term treasury bills
EUR German short-term treasury bills
GBP UK short-term treasury bills
JPY Japan short-term treasury bills
CAD Canada Bankers Acceptances 1m rate
AUD Australian 1m deposit rate
Source: Bloomberg
2. Weight Calculations for Multi-Currency Index
Having identified the six “futures” currencies, their liquid futures contracts, the OAD bucketing
scheme of the cash index for each currency, and the futures funding definition, Figure 7 show
the constituents and weighting mechanics (which follow a similar approach to that used in the
single currency index) of the Global Aggregate MFI as of 30 September 2019.
Bloomberg Barclays Mirror Futures and Duration Hedged Indices
12
FIGURE 7
Global Aggregate MFI Constituents, 30 September 2019
Cash Index Mirror Futures Index
USD Mapped OAD bucket
Currency Bucket
MV%
Glob Agg
MV%
OAD Eligible Futures Futures OAD
Weight
OAD 0-3y 38.54% 17.65% 2.04 2y 1.90 18.98%
OAD 3-5y 25.84% 11.83% 3.85 5y 4.19 10.90%
OAD 5-7.5y
13.95% 6.39%
6.18 10y
6.40 6.14%
OAD 7.5-15y
10.62% 4.86%
10.65 30y
12.52 4.07%
OAD 15y+
11.05% 5.06%
18.04 Ultra
18.95 4.85%
45.79%
USD T-bills (stub)
0.86%
EUR Mapped OAD bucket
Currency Bucket
MV%
Glob Agg
MV%
OAD Eligible Futures Futures OAD
Weight
OAD 0-3y 23.11% 5.90%
2.00
Euro Schatz
1.95 5.90%
OAD 3-6y 31.07% 7.93%
4.46
Euro Bobl
4.81 7.33%
OAD 6-13y 31.16% 7.95%
8.54
Euro Bund
8.86 7.76%
OAD 13y+ 14.66% 3.74%
19.08
Euro Buxl
20.74 3.52%
25.53%
DEM T-bills (stub)
1.03%
GBP OAD bucket
Currency Bucket
MV%
Glob Agg
MV%
OAD Eligible Futures Futures OAD
Weight
OAD 100% 4.56% 11.86 Long Gilt 8.51 6.35%
GBP T-bills (stub)
-1.79%
JPY Mapped OAD bucket
Currency Bucket
MV%
Glob Agg
MV%
OAD Eligible Futures Futures OAD
Weight
OAD 100% 20.24% 9.80 JGB 10y 7.21 27.52%
JPY T-bills (stub)
-7.28%
CAD OAD bucket
Currency Bucket
MV%
Glob Agg
MV%
OAD Eligible Futures Futures OAD
Weight
OAD 100% 2.58% 8.22 Canadian 10y 7.95 2.67%
CBA 1m Rate (stub)
-0.09%
AUD Mapped OAD bucket
Currency Bucket
MV%
Glob Agg
MV%
OAD Eligible Futures Futures OAD
Weight
OAD 0-3y 21.80% 0.28% 2.01 Australia 3y 2.79 0.20%
OAD 3y+ 78.20% 1.02% 7.25 Australia 10y 8.14 0.91%
1.31%
AUD 1m Depo (stub)
0.19%
Total
100.00%
100.00%
Note: each futures contract’s weight is a funded position. Prior to November 2015, EUR Mapped OAD buckets were 0-3y, 3-5y, 5-7.5y, and 7.5y+.
Source: Bloomberg
3. MFI Return Calculations and Funding Assumptions for Multi-Currency
Index
MFI Returns for Multi-Currency Indices (FX Unhedged)
• We compute the month-to-date return (in USD, unhedged) of the Global Aggregate MFI
by first calculating the MTD return (in USD, unhedged) of each individual futures
positions. To do so, we divide the current futures price by the BOM futures price and
Bloomberg Barclays Mirror Futures and Duration Hedged Indices
13
subtract 1, then multiply this difference by the percentage change in the FX rate (per USD).
The result is the percentage P&L in USD per futures contract, or the unfunded MTD return.
• This unfunded return is then added to the MTD unhedged return (in USD) on the funding
position to arrive at the MTD funded futures return.
MFI Returns for Multi-Currency Indices (FX Hedged)
• FX-hedged funded futures returns are calculated as follows: we assume the expected
return on a futures position is zero, so there is no hedging at the BOM for each futures
position. However, the funding component is hedged following standard index
conventions. The fully funded MTD hedged return (in USD) of a futures position is
therefore equal to the unfunded MTD return of the futures (or percentage P&L, in USD)
plus the MTD hedged return (in USD) of the funding component.
• With the individual MTD unhedged (hedged) MFI returns, we then sum across the Global
Agg MFI futures positions, using the BOM MFI weights, to produce the MTD unhedged
(hedged) Global Agg MFI return.
Figure 8 shows a sample MTD return (in USD, unhedged) for the Global Aggregate MFI.
FIGURE 8
Monthly Global Aggregate MFI Total Return (in USD, unhedged), 31 October 2019
Source: Bloomberg
Currency
Futures Funding
Contract Weight
MTD Return (%)
Weight MTD Return (%)
USD
2y (TUZ9) 18.98% 0.047 18.98% 0.156
5y (FVZ9) 10.90% 0.046 10.90% 0.156
10y (TYZ9) 6.14% -0.012 6.14% 0.156
30y (USZ9) 4.07% -0.578 4.07% 0.156
Ultra (WNZ9) 4.85% -1.124 4.85% 0.156
Stub (USD T-bills) 0.86% 0.156
EUR
Schatz (DUZ9) 5.90% -0.237 5.90% 2.273
Bobl (OEZ9) 7.33% -0.777 7.33% 2.273
Bund (RXZ9) 7.76% -1.462 7.76% 2.273
Buxl (UBZ9) 3.52% -3.548 3.52% 2.273
Stub (EUR T-bills) 1.03% 2.273
GBP
Long Gilt (G Z9) 6.35% -1.095 6.35% 5.078
Stub (GBP T-bills) -1.79% 5.078
JPY
JGB 10y (JBZ9) 27.52% -0.690 27.52% -0.046
Stub (JPY T-bills) -7.28% -0.046
CAD
Canadian 10y (CNZ9) 2.67% -0.396 2.67% 0.904
Stub (CBA 1m Rate) -0.09% 0.904
AUD
Australia 3y (YMZ9) 0.20% -0.285 0.20% 2.229
Australia 10y (XMZ9) 0.91% -1.068 0.91% 2.229
Stub (AUD 1m Depo) 0.19% 2.229
MFI MTD Return 100.00% 0.272
Bloomberg Barclays Mirror Futures and Duration Hedged Indices
14
Assessing MFI Performance
MFIs are a component of DHI, as they are the instrument that removes the desired amount of
Treasury duration from a particular cash index. How well a DHI tracks an index’s excess returns
will depend in large part on how well MFIs track the cash treasury market. We will quantify
how well DHIs allow investors to capture index excess returns. However, before doing so, it is
instructive to show how well MFIs track the treasury market. In addition, although MFIs are not
expected to track indices with a large spread component precisely, especially over long
periods, we illustrate how well they can serve as a hedging and replication instrument for
indices with a large spread component.
We present historical MFI performance for several notable USD and global indices. Unlike an
RBI, which is designed to track closely the total return of an underlying index, including its
spread component of return, an MFI is designed to track only the treasury component of the
underlying index’s return. To illustrate that the MFI tracks cash treasury market returns well,
we first show how the US Treasury MFI tracks the US Treasury cash index. If the MFI is doing a
good job, we would expect the short- and long-term average monthly tracking error and TEV
to be very low.
Sources of MFI Tracking Error
An MFI should not be expected to match exactly an index’s treasury component of return, for
two reasons. First, the MFI contains only futures contracts, whereas the treasury component
of an index’s return is computed using cash treasury data. So any cash-futures basis volatility
will lead to an MFI tracking error (i.e., the return difference between the MFI and treasury
component of a cash index’s return).
Second, the treasury component of an index’s return is calculated by matching the six KRD
exposures of each index bond to a set of six hypothetical par treasury bonds at the beginning
of each month. The index calculates the price of each hypothetical treasury by pricing the
bond off the treasury spline curve. So even if the treasury-futures basis is stable, the MFI return
may not match the reported treasury component of return if the MFI’s OAD matching
methodology leads to a different treasury return than does the KRD matching methodology.
This is more likely to be the case in markets with sharp movements in the shape of the yield
curve or for indices where the constituents have amortizing principal (e.g., MBS).
17
Next, we turn our attention to indices with spread components (e.g., the US Aggregate). As
mentioned at the outset, the main source of an index’s return volatility is from its exposure to
the treasury term structure. Consequently, we would expect an index’s MFI also to track its
total returns broadly, except for periods of significant spread widening/tightening and for the
spread carry component of return. For a wide variety of indices, the MFI captures the great
majority of a cash index’s total return. These results should encourage investors to consider
using MFI as a hedge for index returns driven by exposure to the term structure.
US Treasury MFI
As an initial check on the accuracy of the MFI (and, hence, the DHI) methodology, how well
does the US Treasury MFI match the total returns of the US Treasury Index? As shown in Figure
10, except for the volatile cash-futures basis period of September 2008-March 2009, when
17
We do not use KRD matching when constructing the MFI primarily for simplicity, but also because this can lead to large, and sometimes short,
Treasury futures positions. For example, although the MFI positions are all expected to be long futures positions, KRD matching may lead to, say, a
short 5y futures position, offset by a somewhat larger long 10y futures position. This situation can arise because the 5y and 10y futures contracts
may have 5y KRD exposure. If a cash index loads heavily on the 10y KRD and less so on the 5y KRD, a short 5y futures position may be necessary to
get the 5y KRD exposure in alignment with the cash index. We believe such a pattern of futures weights would be confusing to users.
Any cash-futures basis
volatility will lead to MFI
tracking error (i.e., return
difference between the
MFI and Treasury
component of a cash
index’s return)
Bloomberg Barclays Mirror Futures and Duration Hedged Indices
15
the normal cash-futures arbitrage broke down, the monthly return differences between the
cash Treasury index and its MFI have been small.
For May 1999-October 2019, the average monthly tracking error has been -0.9bp with a TEV
of 10.6bp. Figure 9 shows the index values of the US Treasury Index and its MFI. The largest
monthly tracking error was 87.3bp in November 2008.
US Aggregate MFI
We would not expect the US Aggregate MFI to track the Aggregate Index very closely, due to
the relatively large spread component of the Aggregate. However, the Aggregate MFI can
serve a useful, low-cost hedging and replication role. To evaluate the hedging usefulness of
the US Aggregate MFI, we compare its returns with those of the Aggregate Index.
The average monthly tracking error has been -2.7bp, with a TEV of 46.8bp, over May 1999-
October 2019 (Figure 12). Figure 11 shows the index values of the US Aggregate MFI vs US
Aggregate Index’s curve return (total return minus excess return). As expected, the largest
monthly tracking errors occurred during fall 2008. For November 2008, the monthly tracking
error was 256.5bp (i.e., the MFI outperformed), due to the large widening experienced by the
spread sectors of the Aggregate. Much of this outperformance, however, was recouped in
spring 2009. Excluding September 2008-March 2009, the average monthly tracking error has
been -4.8bp, with a TEV of 38.5bp and a maximum monthly tracking error of -172.2bp (April
2009). Over long periods, we would expect the MFI to underperform the cash index, given
the lack of spread carry in the former.
FIGURE
9
US Trea
sury MFI vs. Cash Index, Index Values,
May 1999
-October 2019
FIGURE
10
Realized Monthly TE of US Treasury
MFI vs. Cash Index,
May 1999
-October 2019, bp
Source:
Bloomberg
Source:
Bloomberg
Bloomberg Barclays Mirror Futures and Duration Hedged Indices
16
For investors managing against the Aggregate, initiating a short or long Aggregate MFI
overlay (without the funding component of the MFI) may be a good way to adjust Treasury
exposure quickly. Another use of the Aggregate MFI is for an alpha-beta recombination (i.e.,
portable alpha) strategy. Managers with a good source of alpha can overlay the (unfunded)
Aggregate MFI to produce a synthetic Aggregate portfolio.
Global Treasury G7 MFI
As we did with the US Treasury MFI, an initial check on the accuracy of the MFI methodology in
a global context is to examine how well the Global Treasury G7 MFI tracks the Global Treasury
G7 Index, which as of September 2019 contained treasuries from seven countries (Canada,
France, Germany, Italy, Japan, UK and US), denominated in their local currencies.
The G7 Treasury MFI contains futures contracts representing all five currencies. However,
approximately 12% of the index market value (France (6%) and Italy (6%)) is not represented
by a futures market, as the EUR-denominated futures markets represent the German treasury
market. So some tracking error is expected to the extent that French and Italian treasury
markets diverge from the German one.
Figures 13 and 14 show the tracking performance of the Global Treasury G7 MFI versus the
cash index. For January 2008-October 2019, the average monthly tracking error between the
cash index and its MFI was -0.5bp, with a TEV of 24.6bp. As expected, the largest monthly
tracking error (93.5bp) occurred during the height of the European sovereign crisis (July 2011)
when the Italian, and to a lesser extent, the French treasury markets diverged from the German
treasury market. As the sovereign crisis subsided, the cumulative performance of the cash
index and its MFI rejoined each other.
Despite the severity of the sovereign crisis, however, the MFI tracked the global treasury
markets with a relatively low tracking error volatility.
FIGURE
11
US Aggregate
MFI vs. US Aggregate Curve Return,
Index Values
, May 1999-October 2019
FIGURE
12
Realized Monthly TE of US Aggregate
MFI vs. US
Aggregate Total Return
, May 1999-October 2019, bp
Source:
Bloomberg
Source:
Bloomberg
Bloomberg Barclays Mirror Futures and Duration Hedged Indices
17
For comparison, consider the Global Treasury MFI. The average monthly TE and TEV for the
Global Treasury Index and its MFI are 0.5bp and 38.4bp, respectively. As of September 2019,
the Global Treasury Index contained treasuries from 41 countries, representing 25 currencies
while the Global Treasury MFI contained futures contracts from only six countries representing
six currencies. Approximately 30% of the Global Treasury’s market value is not represented
by a domestic futures contract market. In addition, the Global Treasury Index once contained
countries (e.g., Greece) that were subsequently removed at a time of extreme
underperformance as a result of the European sovereign crisis. Consequently, we would
expect the Global Treasury MFI to track its index less well than either the G7 Treasury MFI or
the US Treasury MFI.
Global Aggregate MFI
As with the US Aggregate MFI, we would not expect the Global Aggregate MFI to track the
cash Global Aggregate Index as well as the Global Treasury MFI tracked its index.
Nevertheless, the Global Aggregate MFI can be very useful. It can serve as a useful hedging
tool as it will remove the largest source of return volatility. Also, as mentioned above, it can
serve as a synthetic beta source for alpha-beta recombination strategies.
The Global Aggregate MFI has an average monthly tracking error versus the Global
Aggregate of -3.6bp, with a TEV of 49.9bp over January 2008-October 2019. Figure 15 shows
the index values of the Global Aggregate MFI vs. the cash index. The largest monthly tracking
error was 184.3bp (November 2008) (Figure 16).
FIGURE
13
Global Treasury G7 Index vs. Global Treasury G7 MFI,
(%, USD unhedged),
Index Values, January 2008-
October 2019
FIGURE
14
Realized Monthly TE of Global Treasury G7
MFI vs.
Global Treasury G7
Index, January 2008-October 2019,
bp, USD unhedged
Source:
Bloomberg
Source:
Bloomberg
Bloomberg Barclays Mirror Futures and Duration Hedged Indices
18
FIGURE
15
Global
Aggregate MFI vs. Global Aggregate Index, (in
USD unhedged),
Index Values, January 2008-October
2019
FIGURE
16
Realized Monthly TE of Global
Aggregate MFI vs.
Global
Aggregate Index, January 2008-October 2019,
bp, USD unhedged
Source:
Bloomberg
Source:
Bloomberg
Bloomberg Barclays Mirror Futures and Duration Hedged Indices
19
Duration Hedged Index (DHI) Methodology
We describe how DHIs are calculated using cash bond indices and MFIs, and then compare
DHI returns with published excess returns for a number of flagship Bloomberg Barclays
Indices.
DHI Construction
A DHI is defined as a cash index minus its MFI. Since the MFI is a funded index, we add back
the MFI funding component to the DHI so that the latter is also a funded index.
For example, to calculate the US Aggregate DHI return (Figure 17), we start with the US
Aggregate Index total return and subtract the US Aggregate MFI total return. However, since
both of these are “funded” indices, the difference between the two is a return on a zero net
capital investment (i.e., an unfunded return). So we add back the MFI “funding” return
component of the Aggregate MFI to generate the Aggregate DHI return. This makes the latter
a funded return, which is how most benchmarks are expressed. Both FX-hedged and
unhedged DHI returns are available.
FIGURE 17
US Aggregate DHI Return Calculation, 31 October 2019
US Aggregate Mirror Futures Index
OAD Bucket
Futures Funding (T-bills)
Contract Weight MTD Returns (%) Weight MTD Return (%)
0-3y TUZ9 42.00% 0.047 42.00% 0.156
3-5y FVZ9 23.83% 0.046 23.83% 0.156
5-7.5y TYZ9 12.77% -0.012 12.77% 0.156
7.5-15y USZ9 8.62% -0.578 8.62% 0.156
15y+ WNZ9 10.99% -1.124 10.99% 0.156
Stub Position
1.79% 0.156
MFI MTD Total Return
0.012
Cash Index MTD Total Return
0.301
DHI MTD Total Return
0.445
US Aggregate DHI MTD Total Return
≡ US Agg Index MTD Total Return
– US Agg MFI MTD Total Return
+ MFI Funding MTD Total Return
= 0.301 – 0.012 + 0.156
= 0.445
Source: Bloomberg
Bloomberg Barclays Mirror Futures and Duration Hedged Indices
20
DHI Tracking Performance
A DHI less its associated MFI funding component should closely track the underlying index’s
reported excess returns. We consider several indices to examine the historical tracking
performance of DHIs.
US Aggregate DHI
For May 1999-October 2019, the average monthly tracking error of the US Aggregate DHI (less
MFI funding) versus the excess returns of the US Aggregate was -0.4bp, with a TEV of 12.0bp
(Figure 18). The largest monthly tracking error was -93.5bp (November 2008).
FIGURE 18
US Aggregate DHI (less funding) vs. US Aggregate Excess Returns, May 1999-October 2019, bp
Source: Bloomberg
Global Aggregate DHI
For January 2008-October 2019, the average monthly tracking error of the Global Aggregate
DHI (less MFI funding) versus the excess returns of the Global Aggregate, both USD
unhedged, was -0.2bp with a TEV of 16.8bp (Figure 19). The largest monthly tracking error was
-69.2bp (November 2008).
FIGURE 19
Global Aggregate DHI (less funding) vs. Global Aggregate Excess Returns, January 2008-
October 2019, bp
Source: Bloomberg
Bloomberg Barclays Mirror Futures and Duration Hedged Indices
21
As shown, the DHI, less funding, does not exactly track the cash index’s excess returns. There
are many sources for this tracking error. First, a DHI uses OAD bucketing, not the KRD matching
that is used when computing index excess returns. Second, it uses Treasury futures to hedge
Treasury curve exposure, not hypothetical cash par Treasury bonds as used for index excess
returns. Consequently, the tracking error between DHI and excess returns will be influenced by
fluctuations in the cash-futures basis. Finally, for global indices, the DHI is limited to using liquid
futures contracts from a limited set of markets, which requires mapping “non-futures” currencies
to “futures” currencies. In contrast, index excess returns are calculated for each currency market.
Return differences between “non-futures” and “futures” currencies will produce tracking error
between DHI and excess returns.
Partially Hedged DHI
Rationale
While some investors may seek a benchmark that hedges interest rate duration completely,
others may still want to (or are required to) retain some exposure to interest rates as a part of
their overall portfolio allocations, while keeping a similar spread duration profile as their
existing index.
The residual interest rate exposure being sought may be expressed as a specific fixed
duration target (e.g., 2y) or as a percentage-based scaling down of their benchmark duration
vis-à-vis a standard Bloomberg Barclays Index (e.g., 50% hedged, 75% hedged). In fact, some
investors in this camp may be constrained by explicit policy guidelines or investment
management agreements that specify (in percentage terms) how large a duration mismatch
they are allowed to maintain in an active portfolio.
Those who seek a shorter duration cash bond benchmark have deployed a number of
strategies. Some have expanded their universe to include shorter duration bonds (such as
floating rate notes or fixed rate bonds with less than 1y to maturity), but the size of these two
universes is generally much smaller than the one measured by existing broad-based
Bloomberg Barclays Indices. Others have sought to lower duration by excluding longer
maturity bonds from their benchmark and measure themselves against only the short end of
their true investment universe (e.g., 1-3y, 1-5y, 1-10y), but this greatly reduces the size of the
investment choice set, scales back spread duration considerably, and limits the diversification
options for those whose expertise lies in credit analysis and issuer selection.
While these approaches may lower overall benchmark duration, they also dramatically alter
the investment universe and may be viewed as a significant change in benchmark design,
even if mandated by the various stakeholders that govern benchmark selection (consultants,
boards, CIOs, etc.).
Therefore, a partially hedged DHI may be more appealing for investors who are strategically
bearish on interest rates but are more neutral-to-bullish on spreads and would therefore want
to retain a comparable spread duration exposure and diversification potential to their existing
broad-based benchmark index. We are able to offer partially hedged versions of MFI and DHI
that retain a desired residual interest rate duration exposure by scaling back (or even leveraging
up) the size of the futures replication to match a specific target duration target.
Bloomberg Barclays Mirror Futures and Duration Hedged Indices
22
Mechanics
The weighting and return mechanics of the partially hedged DHI are nearly identical to a fully
hedged version and track the same full universe as the underlying cash bond index. However,
instead of using 100% of the MFI weight, the weighting of the MFI is scaled up or down by a
fixed percentage to achieve the desired duration exposure.
For example, for a US Aggregate 50% DHI, 50% of the US Aggregate MFI return will be
subtracted from the US Aggregate Index return (and only 50% of the MFI funding return will
be added back).
For example, the return for the October 2019 US Aggregate DHI – 50% hedged is as follows:
≡ US Agg Index MTD Total Return
– 50% (US Agg MFI MTD Total Return)
+ 50% (MFI Funding MTD Total Return)
= 0.301 – (0.012 / 2) + (0.156 / 2)
= 0.373
Figures 20 compares the size, sector exposure, durations, and spread durations of the US
Aggregate 1-5y Index and the US Aggregate as an example. Over the past 10 years, the US
Aggregate 1-5y Index has had an average OAD exposure of ~49% of the full US Aggregate
18
,
serving as reasonable shorter duration proxy for the US Agg with approximately 50%
duration. While having similar durations over time, the US Aggregate DHI 50% Hedged offers
investors (in particular, active ones) a much larger universe to choose from (an average of
4,700 more bonds in the broader index over time); far better issuer diversification; and a
closer spread duration, sector weight profile, and level of spread carry to the cash bond
universe.
FIGURE 20
Comparison of US Aggregate vs. US Aggregate 1-5 Year Cash Bond Indices, 30 September 2019
US Aggregate US Aggregate 1-5y
OAD 5.78 2.54
OASD 6.04 2.95
Market Value ($bn) $23,144 $12,259
# Bonds 10,904 4,789
Treasury 257 150
Government-Related 1,022 555
Corporate 6,094 2,249
Securitized 3,531 1,835
Market Value [%] 100% 100%
Treasury 39.8% 42.3%
Government-Related 5.8% 6.1%
Corporate 25.2% 16.9%
Securitized 29.3% 34.7%
Source: Bloomberg
18
Ranging between 43% and 55% of the US Aggregate’s duration during that window.
Bloomberg Barclays Mirror Futures and Duration Hedged Indices
23
Isolating the non-Treasury duration return component of both indices reveals that the DHI
tracks the broader US Agg excess return better than the US Agg 1-5y excess return (Figure 21
and 22), making it a potentially more appealing option for investors seeking a shorter duration
benchmark, but with comparable spread risk exposure of their existing index.
Bloomberg publishes 50% DHI versions in the DHI family only, but different duration hedge ratio
options are available upon request as bespoke solutions.
Other Customized MFI and DHI
The MFI and DHI methodology permits a high degree of customization. Some requests may
be more complex than others, but the platform offers considerable flexibility to address a
number of special cases. Those that cannot currently be supported can be researched further
and may become available at a future date.
Variables that can be adjusted include:
• Reference Bloomberg Barclays cash bond index: Most existing standard and bespoke
Bloomberg Barclays bond indices can be made available as MFI or DHI, which can also
be paired with a new custom cash bond index request.
• Eligible futures contracts: Clients can request the exclusion of some markets from the
standard MFI. For example, they can request the removal of the USD Ultra-Long bond
futures contract from the US Aggregate MFI.
• Funding assumptions: Other funding assumptions can also be used in the definition of
custom MFI and DHI.
• Additional currency tracking: Clients can also request that more foreign exchange risk
be hedged via the choice of the MFI/DHI constituents. For example, a short-maturity NZD
or CHF deposit could be added to the MFI basket to track these currency movements
better versus the MFI’s base currency. As discussed earlier, the standard MFI maps CHF
FIGURE
21
US Agg
DHI (less funding) vs. US Agg Excess Returns,
May
1999-October 2019, bp
FIGURE
22
US Agg
1-5 Yr Excess Returns vs. US Agg Excess Returns,
May
1999-October 2019, bp
Source:
Bloomberg
Source:
Bloomberg
Bloomberg Barclays Mirror Futures and Duration Hedged Indices
24
to EUR. Consequently, the MFI’s performance versus the cash index is exposed to
CHF/EUR exchange rate movements. Adding a short-maturity CHF deposit (and
correspondingly reducing the short-maturity EUR deposit) to the MFI basket would
remove this exchange rate risk.
• Constant duration benchmarks: The MFI/DHI methodology can also be employed to
create constant duration indices. For example, an investor may wish to have a US
Aggregate Duration = 3 benchmark for risk management or asset allocation purposes.
• Alternative duration measures: Some clients may prefer a Treasury futures hedge
based on a different duration measure than the Bloomberg Barclays analytical duration
measure. For example, a TIPS portfolio manager may want a TIPS MFI that is constructed
using an empirical duration measure. This type of customization is currently not available
in the platform. However, if an investor is interested in such customizations, please contact
the Bloomberg Index group.
Conclusion
Views on interest rates are core to fixed income portfolio management for both active and
passive investors. After a generational bull market for rates, many investors concerned about
absolute return objectives find themselves repositioning for a rising rate environment over
changing time horizons. While some may do this tactically as an overlay strategy, others may
prefer their strategic rates view to be reflected in their benchmarks.
Bloomberg Barclays Mirror Futures Indices (MFI) and Duration Hedged Indices (DHI) are index
solutions that measure the same fixed income choice set as existing Bloomberg Barclays Indices,
but with a fully or partially hedged interest rate duration exposure that is constructed using liquid
futures contracts. As opposed to existing published excess returns on Bloomberg Barclays
Indices that hedge out interest rate exposure using hypothetical Treasuries, MFI and DHI returns
are intended to be replicable and more seamless solutions for those considering a benchmark
switch that preserves many of the characteristics of their existing benchmark. The intuitive, rules-
based, and objective design of this offering makes them a viable alternative.
Bloomberg Barclays Mirror Futures and Duration Hedged Indices
25
Appendix: Accessing Mirror Futures and Duration
Hedged Indices
Bloomberg Terminal
Index tickers that display the total return index levels are available on Bloomberg. A list of
initial total return tickers for major indices can be found in Figure 23.
FIGURE 23
Bloomberg Tickers for MFI and DHI versions of flagship Bloomberg Barclays Indices
USD Total Return EUR Total Return GBP Total Return
Index Unhedged Hedged Unhedged Hedged Unhedged Hedged
Global Global Aggregate Mirror Futures MGAGTRUU MGAGTRUH MGAGTREU MGAGTREH MGAGTRGU MGAGTRGH
Global Aggregate Duration Hedged
DGLATRUU DGLATRUH DGLATREU DGLATREH DGLATRGU DGLATRGH
Global Aggregate Duration Hedged - 50% Hedged DGA5TRUU DGA5TRUH DGA5TREU DGA5TREH DGA5TRGU DGA5TRGH
Global Aggregate ex USD Mirror Futures MGXUTRUU MGXUTRUH MGXUTREU MGXUTREH MGXUTRGU MGXUTRGH
Global Aggregate ex USD Duration Hedged
DGXUTRUU DGXUTRUH DGXUTREU DGXUTREH DGXUTRGU DGXUTRGH
Global Aggregate ex USD Duration Hedged - 50%
Hedged
DGX5TRUU DGX5TRUH DGX5TREU DGX5TREH DGX5TRGU DGX5TRGH
Global Aggregate Corporate Mirror Futures
MGCOTRUU MGCOTRUH MGCOTREU MGCOTREH MGCOTRGU MGCOTRGH
Global Aggregate Corporate Duration Hedged DGLCTRUU DGLCTRUH DGLCTREU DGLCTREH DGLCTRGU DGLCTRGH
Global Aggregate Corporate Duration Hedged -
50%
Hedged
DGC5TRUU DGC5TRUH DGC5TREU DGC5TREH DGC5TRGU DGC5TRGH
US US Aggregate Mirror Futures MUAGTRUU MUAGTRUH MUAGTREU MUAGTREH MUAGTRGU MUAGTRGH
US Aggregate Duration Hedged
DUAGTRUU DUAGTRUH DUAGTREU DUAGTREH DUAGTRGU DUAGTRGH
US Aggregate Duration Hedged - 50% Hedged DUA5TRUU DUA5TRUH DUA5TREU DUA5TREH DUA5TRGU DUA5TRGH
US Treasury Mirror Futures MUTRTRUU MUTRTRUH MUTRTREU MUTRTREH MUTRTRGU MUTRTRGH
US Treasury Duration Hedged
DUSTTRUU DUSTTRUH DUSTTREU DUSTTREH DUSTTRGU DUSTTRGH
US Treasury Duration Hedged - 50% Hedged DUT5TRUU DUT5TRUH DUT5TREU DUT5TREH DUT5TRGU DUT5TRGH
US Corporate Mirror Futures MUCOTRUU MUCOTRUH MUCOTREU MUCOTREH MUCOTRGU MUCOTRGH
US Corporate Duration Hedged
DUSCTRUU DUSCTRUH DUSCTREU DUSCTREH DUSCTRGU DUSCTRGH
US Corporate Duration Hedged - 50% Hedged DUC5TRUU DUC5TRUH DUC5TREU DUC5TREH DUC5TRGU DUC5TRGH
US Credit Mirror Futures MUCTTRUU MUCTTRUH MUCTTREU MUCTTREH MUCTTRGU MUCTTRGH
US Credit Duration Hedged DUCRTRUU DUCRTRUH DUCRTREU DUCRTREH DUCRTRGU DUCRTRGH
US Credit Duration Hedged - 50% Hedged
DUR5TRUU DUR5TRUH DUR5TREU
DUR5TREH DUR5TRGU DUR5TRGH
US TIPS Mirror Futures MTIPTRUU MTIPTRUH MTIPTREU MTIPTREH MTIPTRGU MTIPTRGH
US TIPS Duration Hedged DTIPTRUU DTIPTRUH DTIPTREU DTIPTREH DTIPTRGU DTIPTRGH
US TIPS Duration Hedged - 50% Hedged
DTI5TRUU DTI5TRUH DTI5TREU DTI5TREH DTI5TRGU DTI5TRGH
Euro Euro Aggregate Mirror Futures MEUATRUU MEUATRUH MEUATREU MEUATREH MEUATRGU MEUATRGH
Euro Aggregate Duration Hedged DEAGTRUU DEAGTRUH DEAGTREU DEAGTREH DEAGTRGU DEAGTRGH
Euro Aggregate Duration Hedged - 50% Hedged
DEA5TRUU DEA5TRUH DEA5TREU DEA5TREH DEA5TRGU DEA5TRGH
Euro Corporate Mirror Futures MECOTRUU MECOTRUH MECOTREU MECOTREH MECOTRGU MECOTRGH
Euro Corporate Duration Hedged DECPTRUU DECPTRUH DECPTREU DECPTREH DECPTRGU DECPTRGH
Euro Corporate Duration Hedged - 50% Hedged
DEP5TRUU DEP5TRUH DEP5TREU DEP5TREH DEP5TRGU DEP5TRGH
Source: Bloomberg
15 January 2020
New York
+1
-212-318-2000
London
+44
-20-7330-7500
Singapore
+65
-6212-1000
Hong Kong
+852
-2977-6000
Tokyo
+81
-3-3201-8900
Sydney
+61
-2-9777-8600
Disclaimer
BLOOMBERG is a trademark and service mark of Bloomberg Finance L.P. BARCLAYS is a
trademark and service mark of Barclays Bank PLC, used under license. Bloomberg Finance L.P. and
its affiliates, including Bloomberg Index Services Limited, as administrator of the BLOOMBERG
BARCLAYS INDICIES (collectively, “Bloomberg”), or Bloomberg’s licensors own all proprietary
rights in the BLOOMBERG BARCLAYS INDICES. Neither Bloomberg nor Barclays Bank PLC or
Barclays Capital Inc. or their affiliates (collectively "Barclays") guarantee the timeliness, accuracy or
completeness of any data or information relating to BLOOMBERG BARCLAYS INDICES or make
any warranty, express or implied, as to the BLOOMBERG BARCLAYS INDICES or
any data or values
relating thereto or results to be obtained therefrom, and each of Bloomberg and Barclays
expressly disclaims all warranties of merchantability and fitness for a particular purpose with
respect thereto. It is not possible to invest directly in an index. Back-tested performance is not
actual performance. Past performance is not an indication of future results. To the maximum extent
allowed by law, Bloomberg and its licensors, and their respective employees, contractors, agents,
suppliers and vendors shall have no liability or responsibility whatsoever for any injury or damages
- whether direct, indirect, consequential, incidental, punitive or otherwise - arising in connection
with BLOOMBERG BARCLAYS INDICES or any data or values relating thereto - whether arising
from their negligence or otherwise. This document constitutes the provision of factual information,
rather than financial product advice. Nothing in the BLOOMBERG BARCLAYS INDICES shall
constitute or be construed as an offering of financial instruments or as investment advice or
investment recommendations (i.e., recommendations as to whether or not to “buy,” “sell,” “hold” or
enter into any other transaction involving a specific interest) by Bloomberg or its affiliates or
licensors or a recommendation as to an investment or other strategy. Data and other information
available via the BLOOMBERG BARCLAYS INDICES should not be considered as information
sufficient upon which to base an investment decision. All information provided by the
BLOOMBERG BARCLAYS INDICES is impersonal and not tailored to the needs of any specific
person, entity or group of persons. Bloomberg and its affiliates express no opinion on the future or
expected value of any security or other interest and do not explicitly or implicitly recommend or
suggest an investment strategy of any kind. In addition, Barclays is not the issuer or producer of the
BLOOMBERG BARCLAYS INDICES and has no responsibilities, obligations or duties to investors in
these indices. While Bloomberg may for itself execute transactions with Barclays in or relating to
the BLOOMBERG BARCLAYS INDICES, investors in the BLOOMBERG BARCLAYS INDICES do not
enter into any relationship with Barclays and Barclays does not sponsor, endorse, sell or promote,
and Barclays makes no representation regarding the advisability or use of, the BLOOMBERG
BARCLAYS INDICES or any data included therein. Customers should consider obtaining
independent advice before making any financial decisions. ©2020
Bloomberg. This document and
its contents may not be forwarded or redistributed without the prior consent of Bloomberg.
The BLOOMBERG TERMINAL service and Bloomberg data products (the “Services”) are owned
and distributed by Bloomberg Finance L.P. (“BFLP”) except (i) in Argentina, Australia and certain
jurisdictions in the Pacific islands, Bermuda, China, India, Japan, Korea and New Zealand, where
Bloomberg L.P. and its subsidiaries distribute these products, and (ii) in Singapore and the
jurisdictions serviced by Bloomberg’s Singapore office, where a subsidiary of BFLP distributes
these products.
