This is not a product of Barclays Research. This is a product of Barclays Sales and Trading.

For Professional and Institutional Investors Only. Not for Retail Distribution.
Investors in specific Asia Pacific countries should review important disclosures at the end of this document.

9th February 2018

QIS Insights
Style Investing in Rates Markets INVESTMENT STRATEGIES

 While factor-based investing is a well-established concept in equities, supported by Charles Fattouche

more than two decades of research and literature, a comparable investment +44 (0)20 3134 8059
charles.fattouche@barclays.com
approach in rates markets remains in its nascent stages.

 We explore ways to capture risk premia returns from interest rate markets and
www.barclays.com
emphasize three systematic investment styles constructed long-short in the cross
section of sovereign bond futures: carry, value and momentum. These alternative
return sources historically delivered attractive risk-adjusted returns, net of costs,
with low duration exposure.

 To avoid biases associated with the three-decade long rally in bonds and in
particular to cover periods of rising yields, we extend data sets and backtests to the
early 1960s and investigate the behaviour of rates factors conditional on different
yield environments.

 We highlight that, while each single factor historically exhibited attractive

performance characteristics, a multi-factor approach provides an attractive solution
aiming to address essential investors’ needs in the current environment: capturing
excess returns in both rising and falling yield environments.
Factor-based investing looks beyond traditional asset class labels and seeks to target the underlying primary
drivers of risk and return within a market segment or asset class. These factors can take the form of macro risk
factors representing non-diversifiable risks such as economic growth or inflation, or style risk factors aiming to
capture a risk premium or a systematic behavioural bias. In equity markets, style investing is a well-established
concept supported by more than two decades or research and literature, and includes common systematic factors
such as value, size, momentum, quality and low risk. Despite the importance and size of the fixed income markets
as a whole, a comparable investment approach in this asset class remains in its nascent stage. This may be
attributed, among others, to limited access to reliable data, little academic literature and liquidity constraints. In
this paper, we explore factor-based investing in rates markets implemented in a systematic fashion and
emphasize that it can benefit institutional investors looking for yield without taking additional duration, credit or
liquidity risk.

Achieving returns in bond markets for the last three decades was in hindsight relatively straightforward. Bond
markets indeed saw a rally since the mid-1980s driven by the secular decline in yields. However, at current yield
levels, the outlook for long-only bond exposure may be less promising, whether yields are on the rise or remain
anchored at current historical lows. In the subsequent sections, we explore ways to capture risk premia returns
from interest rate markets and define three systematic investment styles in interest rate markets aiming to deliver
excess returns with low duration exposure in both rising and falling yield environments.

Data

To make this study relevant for implementation, and since we are interested only in excess returns above risk-free
rates, we used primarily sovereign bond futures data where and when available. In particular, we used U.S.,
German, U.K., Japanese, Canadian and Australian sovereign bond futures, i.e. the most liquid markets, since they
started trading. Additionally, to avoid biases associated with the long recent period of falling yields and in
particular to cover the period of rising yields in the 1960s, 1970s and early 1980s, we extended data sets back
before bond futures data were available using proxies calculated from the corresponding government bond total
returns minus returns of a 3-month cash rate.

Defining Styles

Carry
An asset’s “carry” is defined as its return assuming that market conditions stay the same, i.e. carry is the income
earned if the asset’s price remains constant over the holding period. In the context of a bond whose returns can
be decomposed into yield income earned over time and capital gain or loss due to yield changes, carry is no more
than the former return component. For a bond future, carry is therefore the difference between the relevant bond
yield and the short-term financing rate, hence reflects curve steepness. Such measure is sometimes called excess
yield or term spread.

We use 10-year sovereign bond futures across six countries, as described previously, and compute their carry
measures as follows:

Value
One of the most studied risk factors in the stock market is the value factor, describing the relation between a
stock’s return and the ratio of its market value relative to its fundamental (or book) value. A value strategy
typically aims to harvest excess returns by investing in undervalued assets and shorting relatively overvalued
assets. Given a valuation model, one can easily test the presence of a value premium in any market. For bond

This is not a product of Barclays Research. This is a product of Barclays Sales and Trading.
For Professional and Institutional Investors Only. Not for Retail Distribution.
Investors in specific Asia Pacific countries should review important disclosures at the end of this document.
2
markets, we choose to rely on standard economic theory which suggests that the yield on a risk-free government
bond should approximately equal the rate of growth in the economy, plus the rate of inflation. To illustrate that
relationship, we consider the U.S. economy over the last 50 years: economic growth averaged 2.8%1 over that
period while annual inflation realized at 4.1%2, suggesting an average level for “normal” bond yields of 6.9%. This
is more or less in line with U.S. generic 10 year government bond yields that averaged 6.5% over that same period.
The historical evolution of that relationship is provided in Appendix 1 for the six markets considered in this paper.

We use 10-year sovereign bond futures across six countries, as described previously, and compute their value
measures as follows:

Note that we lagged the nominal GDP growth rate by three months to ensure point-in-time data.

Momentum
Momentum refers to persistence in asset returns: past winners tend to keep winning and past losers tend to keep
losing relative to their peers. While momentum has no adequate risk-based explanation and is a challenge to the
theory of financial market efficiency, a plethora of academic papers3 found that momentum is present in major
financial markets. A number of behavioural-based explanations have been suggested, e.g. under-reaction to new
information or the so-called disposition and bandwagon effects.

In the literature, momentum is interchangeably used to describe two fairly different strategies: time-series
momentum i.e. trend-following strategies; and cross-sectional momentum which always takes long-short
exposure with the aim of having at any point in time a low market beta exposure. For the avoidance of doubt,
momentum in this paper refers to cross-sectional momentum i.e. constructed long-short consistently with the
previously described rates carry and value factors.

Profitability of cross-sectional momentum is typically attributed to the cross-sectional dispersion in underlyings’

returns, itself being inversely related to the underlyings’ pairwise correlation. With sovereign bond futures at the
10-year point being historically correlated at around 70%4, there is little dispersion for a momentum strategy to
capture decent excess returns5. We introduce sovereign bond futures at the 2-year and 5-year points in countries
where such instruments are liquid, namely the U.S. and Germany, and duration-weight them versus their
respective 10-year bond future6. Enlarging the universe of underlyings with shorter dated bond futures enables
the momentum strategy to capture returns from the second principal component of yield curves, i.e. from
duration-neutral steepening and flattening movements.

We therefore use sovereign bond futures across six countries and three tenors: 10-year, 5-year and 2-year futures
for U.S. and Germany, and 10-year futures for U.K., Japan, Canada and Australia. Momentum measures are
computed as follows:

1 Average of US GDP Chained 2009 Dollars YoY SA, released quarterly, between Dec 1967 and Sep 2017
2 Average of US CPI Urban Consumers YoY NSA, released monthly, between Dec 1967 and Sep 2017
3 Jegadeesh N. et al (1993 and 2001), Asness C. S. et al (2013), Hurst B. (2014)
4 Average pairwise correlation between U.S., German, U.K., Canadian and Australian 10-year sovereign bond futures has been 76%

historically since August 1999 (based on monthly returns), and has been 61% when adding the Japan 10-year bond futures.
5 We acknowledge results from Jamil Baz et al. (2015) in their paper on “Dissecting Investment Strategies in the Cross Section and Time

Series” showing that a cross-sectional momentum strategy, using as underlyings fourteen interest rate swap contracts all at the 10-year
point, has historically generated negative excess returns. We effectively observe consistent results when using a universe of 10-year
sovereign bond futures across six countries.
6 2-year and 5-year US Treasury futures are duration weighted versus 10-year US Treasury futures, while Schatz and Bobl futures are

duration-weighted versus Bund futures.

This is not a product of Barclays Research. This is a product of Barclays Sales and Trading.
For Professional and Institutional Investors Only. Not for Retail Distribution.
Investors in specific Asia Pacific countries should review important disclosures at the end of this document.
3
Assessing Styles

Before implementing any strategy, we focus on assessing the predictive power of each measure, both on a cross-
sectional and time-series basis.

 Cross-sectional: Are the measures of carry, value and momentum predictive of whether a bond future is likely
to perform better or worse than other bond futures in the near future?
 Time-series: Are the measures of carry, value and momentum predictive of subsequent excess returns for each
bond future independently?

While only the former is relevant in the context of cross-sectional long-short strategies a la Fama-French, we also
test predictive power on a time-series basis to gather further insights into the forecasting ability of the previously
defined carry, value and momentum measures.

Assessing Styles: Cross-Sectional

In order to assess the validity of the carry, value and momentum measures as predictors of subsequent relative
performance, we run the following test:
(1) Rank on a monthly basis the sovereign bond futures according to their carry, value or momentum measures
(2) Calculate the average7 subsequent monthly futures return (annualised) and Sharpe ratio of the sorted bond
futures

Carry

Figure 1 provides evidence of a strong positive relationship between carry measures and subsequent bond futures
performance, whereby higher relative carry measures tend to translate into higher subsequent returns and Sharpe
ratios, and vice-versa.

Figure 1: Relationship between carry measures and average subsequent monthly futures
returns (annualised) and Sharpe ratio
4.5% 0.7
Futures Return (LHS)
Average Subsequent Sharpe Ratio
4.0% 0.6
Average Subsequent Monthly

Sharpe Ratio (RHS)

3.5%
0.5
3.0%
Return p.a.

2.5% 0.4

2.0% 0.3
1.5% 0.2
1.0%
0.1
0.5%
0.0% 0.0

-0.5% -0.1
1 2 3 4 5 6
Rank
Determined using monthly data from Jan 1960 to Jun 2017. Source: Bloomberg, Barclays
Historical and hypothetical/simulated performance is not indicative of future performance or future results.

Value

The validity of the previously defined value measure is assessed by running a test identical to that for the Rates
Carry Factor. Figure 2 emphasizes that the relationship between value measures and subsequent bond futures
performance is nearly monotonic: undervalued bond futures tend to deliver higher returns and risk-adjusted
returns over time, and vice-versa.

7 An identical analysis based on subsequent median returns as opposed to mean returns produces very similar results.

This is not a product of Barclays Research. This is a product of Barclays Sales and Trading.
For Professional and Institutional Investors Only. Not for Retail Distribution.
Investors in specific Asia Pacific countries should review important disclosures at the end of this document.
4
 Figure 2: Relationship between value measures and average subsequent monthly futures
returns (annualised) and Sharpe ratio
4.5% 0.8
Futures Return (LHS)

Average Subsequent Sharpe Ratio

4.0% 0.7

Average Subsequent Monthly

Sharpe Ratio (RHS)
3.5% 0.6
3.0%
Return p.a. 0.5
2.5%
0.4
2.0%
0.3
1.5%
1.0% 0.2

0.5% 0.1

0.0% 0.0
1 2 3 4 5 6
Rank
Determined using monthly data from Apr 1961 to Jun 2017. Source: Bloomberg, Barclays
Historical and hypothetical/simulated performance is not indicative of future performance or future results.

Momentum

While the predictive ability of the momentum measure seems less strong than for carry or value, there is still a
clear pattern between momentum measures and subsequent relative performance of bond futures as represented
on Figure 3.

Figure 3: Relationship between momentum measures and average subsequent monthly

futures returns (annualised) and Sharpe ratio
4.0% 0.6
Futures Return (LHS)

Average Subsequent Sharpe Ratio

3.5%
Average Subsequent Monthly

Sharpe Ratio (RHS) 0.5

3.0%
0.4
2.5%
Return p.a.

2.0% 0.3
1.5%
0.2
1.0%
0.1
0.5%
0.0% 0.0
1 2 3 4 5 6 7 8 9 10
Rank
Determined using monthly data from Jan 1960 to Jun 2017. Source: Bloomberg, Barclays
Historical and hypothetical/simulated performance is not indicative of future performance or future results.

We would like to refer the interested reader to our work on “Tactical Allocation to Alternative Risk Premia”8 which
provides further details on the predictive ability of such measures, in particular carry and value measures, for
subsequent performance of relevant factors.

Assessing Styles: Time-Series

In order to test the absolute forecasting ability of carry, value and momentum measures for bond futures returns,
time-series regressions of each measure on subsequent excess returns are run for each market and over different
horizons, ranging from one month to three years. Figure 4 reports the regression coefficients (β), associated p-
values and r-squared.

8Fattouche C. (2016). Please refer as well to Barclays Alternative Risk Premia Scores published monthly and indicating on a scale of 0 to
10 how rich or cheap the relevant risk premia currently is compared to its history.

This is not a product of Barclays Research. This is a product of Barclays Sales and Trading.
For Professional and Institutional Investors Only. Not for Retail Distribution.
Investors in specific Asia Pacific countries should review important disclosures at the end of this document.
5
Carry

Figure 4 reveals that relationships between ex-ante carry measures and subsequent bond futures excess returns
are highly statistically significant for all markets and horizons up to two years, while explanatory power is highest
for the one year horizon.

Figure 4: Regression of carry measures on subsequent returns of bond futures over different horizons

USD 10Y EUR 10Y GBP 10Y JPY 10Y CAD 10Y AUD 10Y
p- p- p- p- p- p-
R R R R R R
2 2 2 2 2 2
β β β β β β
value value value value value value
Next
0.16 3.9E-03 1.2% 0.17 3.3E-08 4.3% 0.17 9.8E-05 2.2% 0.23 3.1E-04 2.5% 0.22 2.5E-08 4.4% 0.32 5.3E-09 5.8%
Month ** *** *** *** *** ***
Next
0.70 2.6E-11 6.3% 0.50 8.0E-14 7.8% 0.46 3.9E-07 3.7% 0.62 3.0E-07 5.0% 0.65 2.8E-15 8.7% 0.78 3.7E-15 10.3%
Quarter *** *** *** *** *** ***
Next 6
1.40 6.0E-20 11.5% 0.99 2.1E-20 11.8% 0.82 5.0E-10 5.5% 1.03 6.8E-11 8.1% 1.19 1.1E-22 13.2% 1.16 5.3E-15 10.2%
Months *** *** *** *** *** ***
Next
2.57 7.8E-30 17.4% 1.55 2.0E-20 11.9% 1.20 1.1E-10 6.0% 1.53 2.8E-12 9.3% 1.87 8.2E-25 14.5% 1.79 2.5E-16 11.3%
Year *** *** *** *** *** ***
Next 2
2.56 3.2E-13 7.7% 1.17 1.7E-05 2.8% 1.11 2.6E-05 2.6% 1.79 8.7E-08 5.7% 2.26 8.4E-16 9.3% 1.52 6.5E-06 3.6%
Years *** *** *** *** *** ***
Next 3
1.91 1.6E-05 2.8% 0.39 2.7E-01 0.2% 0.68 3.9E-02 0.7% 1.85 1.5E-05 3.9% 2.13 3.2E-09 5.2% 1.57 2.3E-04 2.5%
Years *** * *** *** ***
Statistical significance codes: ‘***’ p-value < 0.001; ‘**’ p-value < 0.01; ‘*’ p-value < 0.05; ‘.’ p-value < 0.1
Determined using monthly data from Jan 1960 to Jun 2017. Source: Bloomberg, Barclays
Historical and hypothetical/simulated performance is not indicative of future performance or future results.

Value

While carry measures seem to best explain one-year subsequent returns of bond futures, value measures tend to
have, perhaps unsurprisingly, larger explanatory power for longer term horizons. Indeed, explanatory power tends
to increase monotonically with the window used for subsequent returns. Additionally, relationships between ex-
ante value measures and subsequent bond futures excess returns are highly statistically significant for all markets
and horizons beyond six months.

Figure 5: Regression of value measures on subsequent returns of bond futures over different horizons

USD 10Y EUR 10Y GBP 10Y JPY 10Y CAD 10Y AUD 10Y
p- p- p- p- p- p-
R R R R R R
2 2 2 2 2 2
β β β β β β
value value value value value value
Next
0.07 2.7E-03 1.3% 0.04 1.2E-02 0.9% 0.07 1.3E-03 1.5% 0.01 6.9E-01 0.0% 0.05 7.5E-04 1.7% 0.09 6.6E-05 2.7%
Month ** * ** *** ***
Next
0.28 5.4E-09 5.0% 0.13 5.1E-05 2.4% 0.22 4.1E-07 3.8% 0.08 7.0E-02 0.6% 0.17 4.4E-08 4.4% 0.28 3.3E-11 7.4%
Quarter *** *** *** . *** ***
Next 6
0.55 6.4E-15 8.7% 0.28 3.3E-08 4.5% 0.48 8.9E-15 8.6% 0.18 1.2E-03 2.1% 0.32 3.2E-12 7.0% 0.53 6.2E-18 12.3%
Months *** *** *** ** *** ***
Next
0.98 1.6E-20 12.2% 0.55 1.4E-11 6.7% 0.90 1.1E-24 14.7% 0.35 4.4E-06 4.1% 0.55 8.6E-16 9.3% 0.93 6.5E-25 17.3%
Year *** *** *** *** *** ***
Next 2
2.19 6.2E-47 27.3% 0.92 1.2E-12 7.5% 1.04 5.9E-17 10.2% 0.70 1.4E-09 7.3% 1.12 6.1E-28 16.9% 1.44 4.1E-26 18.4%
Years *** *** *** *** *** ***
Next 3
3.17 4.9E-67 37.5% 1.28 2.3E-14 8.8% 1.47 3.8E-22 13.7% 0.84 1.4E-08 6.6% 1.47 2.3E-30 18.6% 1.64 2.6E-21 15.4%
Years *** *** *** *** *** ***
Statistical significance codes: ‘***’ p-value < 0.001; ‘**’ p-value < 0.01; ‘*’ p-value < 0.05; ‘.’ p-value < 0.1
Determined using monthly data from Apr 1961 to Jun 2017. Source: Bloomberg, Barclays
Historical and hypothetical/simulated performance is not indicative of future performance or future results.

This is not a product of Barclays Research. This is a product of Barclays Sales and Trading.
For Professional and Institutional Investors Only. Not for Retail Distribution.
Investors in specific Asia Pacific countries should review important disclosures at the end of this document.
6
Momentum

Being by nature a much more dynamic strategy, we only observe statistical significance for much shorter horizons
than for carry and value, on average for horizons up to six months. The momentum measure tends to best explain
subsequent bond futures return also over short horizons. It is worth noting that r-squared for these regressions
are a lot smaller than for carry and value, reflecting the much higher variability of momentum measures.

Figure 6: Regression of momentum measures on subsequent returns of bond futures over different horizons

USD 10Y EUR 10Y GBP 10Y JPY 10Y CAD 10Y AUD 10Y
p- p- p- p- p- p-
R R R R R R
2 2 2 2 2 2
β β β β β β
value value value value value value
Next
0.03 2.0E-03 1.4% 0.04 3.2E-08 4.4% 0.02 6.7E-02 0.5% 0.01 4.9E-01 0.1% 0.03 5.0E-05 2.4% 0.03 1.4E-02 1.1%
Month ** *** . *** *
Next
0.05 2.4E-03 1.4% 0.09 8.2E-09 4.8% 0.01 6.5E-01 0.0% 0.00 9.1E-01 0.0% 0.07 1.3E-04 2.2% 0.05 8.4E-03 1.2%
Quarter ** *** *** **
Next 6
0.08 3.0E-03 1.3% 0.13 8.0E-08 4.2% -0.02 4.8E-01 0.1% 0.01 6.6E-01 0.0% 0.10 9.0E-05 2.3% 0.08 4.8E-03 1.4%
Months ** *** *** **
Next
0.08 5.2E-02 0.6% 0.17 1.2E-05 2.8% -0.03 4.6E-01 0.1% 0.07 1.2E-01 0.5% 0.14 2.9E-04 2.0% 0.11 1.2E-02 1.2%
Year . *** *** *
Next 2
0.01 8.8E-01 0.0% 0.10 1.1E-01 0.4% -0.10 6.9E-02 0.5% 0.03 6.8E-01 0.0% 0.13 2.4E-02 0.8% 0.05 4.2E-01 0.1%
Years . *
Next 3
0.10 1.8E-01 0.3% -0.13 1.1E-01 0.4% -0.29 1.5E-05 2.9% -0.01 9.1E-01 0.0% 0.11 1.4E-01 0.3% -0.04 6.2E-01 0.0%
Years ***

USD 2Y USD 5Y EUR 2Y EUR 5Y

p- p- p- p-
R R R R
2 2 2 2
β β β β
value value value value
Next
0.01 1.0E-02 1.0% 0.02 2.3E-02 0.8% 0.01 6.2E-07 3.6% 0.03 2.6E-07 3.9%
Month * * *** ***
Next
0.02 8.3E-03 1.0% 0.03 2.0E-02 0.8% 0.02 1.4E-06 3.4% 0.06 2.6E-07 3.9%
Quarter ** * *** ***
Next 6
0.03 3.1E-04 1.9% 0.05 7.1E-03 1.1% 0.03 8.2E-05 2.3% 0.07 1.6E-05 2.7%
Months *** ** *** ***
Next
0.04 2.9E-04 2.0% 0.06 2.4E-02 0.8% 0.01 2.5E-01 0.2% 0.05 4.4E-02 0.6%
Year *** * *
Next 2
0.02 3.3E-01 0.1% -0.01 7.9E-01 0.0% -0.09 4.8E-07 3.8% -0.13 1.5E-03 1.5%
Years ** **
Next 3
0.01 7.3E-01 0.0% -0.01 8.0E-01 0.0% -0.08 2.5E-05 2.7% -0.15 1.3E-03 1.6%
Years ** **
Statistical significance codes: ‘***’ p-value < 0.001; ‘**’ p-value < 0.01; ‘*’ p-value < 0.05; ‘.’ p-value < 0.1
Determined using monthly data from Jan 1961 to Jun 2017. Source: Bloomberg, Barclays
Historical and hypothetical/simulated performance is not indicative of future performance or future results.

Empirical Results

The performance of bond futures sorted on their ex-ante carry, value or momentum measures suggest that
building carry, value or momentum factors taking long positions in high yielding, undervalued or past performing
bond futures and short positions in low yielding, overvalued or past underperforming ones, may extract attractive
risk-adjusted returns over the long run with low beta exposure to the rates market.

As a simple approach, we build each of the Rates Carry, Rates Value and Rates Momentum factors by allocating
each month 50% weight to each of the top two ranked bond futures, and -50% weight to each of the bottom two
ranked futures, and account for all transaction costs i.e. futures roll costs, rebalancing costs and margining cost
estimates.

This is not a product of Barclays Research. This is a product of Barclays Sales and Trading.
For Professional and Institutional Investors Only. Not for Retail Distribution.
Investors in specific Asia Pacific countries should review important disclosures at the end of this document.
7
Carry Factor

From Figure 1, we anticipate that the annualised excess return of such strategy should be around 3.5%9 provided
that the bulk of the returns stems from carry differential as opposed to market moves. Figure 7 shows the
evolution of such strategy and confirms that the captured excess return is in line with expectations while the
correlation with a long-only equally weighted basket of bond futures is close to zero.

Figure 7: Rates Carry Factor

800 Rates Carry
Jan 1960 - Jun 2017
Factor
700

600 Annualised Excess Return 3.54%

500 Annualised Volatility 5.02%

400
Sharpe Ratio 0.70
300

200 Maximum Drawdown -15.1%

100 Rates Carry Factor
Correlation with Long-Only Basket 4.1%
0
Correlation with Global Treasury
1960

1984

2008
1963
1966
1969
1972
1975
1978
1981

1987
1990
1993
1996
1999
2002
2005

2011
2014
2017
9.1%
Aggregate
Source: Bloomberg, Barclays. Determined using monthly data from Jan 1960 to Jun 2017 for all statistics except the correlation with Global Treasury
Aggregate which was determined using data from Jan 1973 to Jun 2017. Returns for underlying sovereign bond futures are FX-translated into USD.
Long-Only Basket is computed as an equally-weighted basket of 10-year bond futures across U.S., Germany, U.K., Japan, Canada and Australia,
rebalanced monthly. Global Treasury Aggregate means the Bloomberg Barclays Global Treasury Total Return Index Hedged USD (BTSYTRUH Index)
from Jan 1987 onwards, and the Bloomberg Barclays US Treasury Total Return Index (LUATTRUU Index) before that.
Historical and hypothetical/simulated performance is not indicative of future performance or future results.

Value Factor

We adopt the same approach as for the Rates Carry Factor and build a Rates Value Factor by allocating each
month 50% weight to each of the two most undervalued bond futures, and -50% weight to each of the most
overvalued ones, and account for all transaction costs. As highlighted in Figure 8, this relatively simple approach
historically generated an excess return net of costs of more than 2.6% per annum since the early 1960s, with a
Sharpe ratio in excess of 0.5 and a correlation to a long-only basket of bond futures close to zero.

Figure 8: Rates Value Factor

500 Rates Value
Apr 1961 - Jun 2017
450 Factor
400 Annualised Excess Return 2.65%
350
300 Annualised Volatility 4.70%
250
200 Sharpe Ratio 0.56
150
Maximum Drawdown -14.9%
100
Rates Value Factor
50
Correlation with Long-Only Basket -5.3%
0
Correlation with Global Treasury
1967

1994
1961
1964

1970
1973
1976
1979
1982
1985
1988
1991

1997
2000
2003
2006
2009
2012
2015

6.7%
Aggregate
Source: Bloomberg, Barclays. Determined using monthly data from Apr 1961 to Jun 2017 for all statistics except the correlation with Global Treasury
Aggregate which was determined using data from Jan 1973 to Jun 2017. Returns for underlying sovereign bond futures are FX-translated into USD.
Long-Only Basket is computed as an equally-weighted basket of 10-year bond futures across U.S., Germany, U.K., Japan, Canada and Australia,
rebalanced monthly. Global Treasury Aggregate means the Bloomberg Barclays Global Treasury Total Return Index Hedged USD (BTSYTRUH Index)
from Jan 1987 onwards, and the Bloomberg Barclays US Treasury Total Return Index (LUATTRUU Index) before that.
Historical and hypothetical/simulated performance is not indicative of future performance or future results.

9 Average of blue bars corresponding to #1 and #2 ranks, minus average of bars corresponding to #5 and #6 ranks

This is not a product of Barclays Research. This is a product of Barclays Sales and Trading.
For Professional and Institutional Investors Only. Not for Retail Distribution.
Investors in specific Asia Pacific countries should review important disclosures at the end of this document.
8
Momentum Factor

We construct a Rates Momentum Factor using the exact same methodology as for the previous two factors. As
typically observed in other asset classes, momentum strategies tend to have fairly large turnover and high
volatility. Average annual turnover of Rates Momentum Factor is indeed larger than that for Carry or Value factors
by a factor of 1.8, while its historical realised volatility is higher by a factor of around 1.5.
Figure 9: Rates Momentum Factor
350 Rates Momentum
Jan 1961 - Jun 2017
Factor
300
Annualised Excess Return 1.61%
250

200 Annualised Volatility 7.20%

150 Sharpe Ratio 0.22

100
Maximum Drawdown -32.3%
50 Rates Momentum Factor Correlation with Long-Only
-3.2%
0 Basket
Correlation with Global
1963

1972
1960

1966
1969

1975
1978
1981
1984
1987
1990
1993
1996
1999
2002
2005
2008
2011
2014
2.4%
Treasury Aggregate
Source: Bloomberg, Barclays. Determined using monthly data from Jan 1961 to Jun 2017 for all statistics except the correlation with Global Treasury
Aggregate which was determined using data from Jan 1973 to Jun 2017. Returns for underlying sovereign bond futures are FX-translated into USD.
Long-Only Basket is computed as an equally-weighted basket of 10-year bond futures across U.S., Germany, U.K., Japan, Canada and Australia,
rebalanced monthly. Global Treasury Aggregate means the Bloomberg Barclays Global Treasury Total Return Index Hedged USD (BTSYTRUH Index)
from Jan 1987 onwards, and the Bloomberg Barclays US Treasury Total Return Index (LUATTRUU Index) before that.
Historical and hypothetical/simulated performance is not indicative of future performance or future results.

Multi-Factor

We define four multi-factor strategies which will be useful in emphasizing the complementary nature of the
different single-style strategies: carry & value, value & momentum, carry & momentum, and carry, value &
momentum. For the purpose of simplicity, we construct the multi-factor strategies as equally-weighted
combinations of the underlying single factors, and account for netting benefit from potential offsetting positions.
Figure 10 illustrates such multi-factor strategy combining carry, value and momentum. Its Sharpe ratio net of
costs has historically been in excess of 0.7, while its correlation to a long-only basket of bond futures has been
close to zero.
Figure 10: Rates Multi Factor (Carry, Value & Momentum)
500 Rates Multi
Apr 1961 - Jun 2017
450 Factor
400 Annualised Excess Return 2.75%
350
300 Annualised Volatility 3.78%
250
200 Sharpe Ratio 0.73
150
Maximum Drawdown -12.4%
100
Rates Multi Factor (Carry, Value & Momentum)
50
Correlation with Long-Only Basket -1.5%
0
Correlation with Global Treasury
1964

1991

2003
1961

1967
1970
1973
1976
1979
1982
1985
1988

1994
1997
2000

2006
2009
2012
2015

8.6%
Aggregate
Source: Bloomberg, Barclays. Determined using monthly data from Apr 1961 to Jun 2017 for all statistics except the correlation with Global Treasury
Aggregate which was determined using data from Jan 1973 to Jun 2017. Returns for underlying sovereign bond futures are FX-translated into USD.
Long-Only Basket is computed as an equally-weighted basket of 10-year bond futures across U.S., Germany, U.K., Japan, Canada and Australia,
rebalanced monthly. Global Treasury Aggregate means the Bloomberg Barclays Global Treasury Total Return Index Hedged USD (BTSYTRUH Index)
from Jan 1987 onwards, and the Bloomberg Barclays US Treasury Total Return Index (LUATTRUU Index) before that.
Historical and hypothetical/simulated performance is not indicative of future performance or future results.

This is not a product of Barclays Research. This is a product of Barclays Sales and Trading.
For Professional and Institutional Investors Only. Not for Retail Distribution.
Investors in specific Asia Pacific countries should review important disclosures at the end of this document.
9
Performance Attribution

A bond’s returns can be decomposed into yield income earned over time, i.e. carry returns, and capital gain or loss
due to yield changes. This translates in the sovereign bond futures markets into roll-down along the futures curve,
which we call carry returns, and market returns due to outright market moves. Figure 11 shows the historical
return decomposition of each factor strategy into these two components. Unsurprisingly, a carry factor earns the
bulk of its returns from the carry component. Additionally, the evolution of this carry component confirms that the
carry factor always earns positive carry over time. Conversely, a value factor earns very little from the carry
component and this component has historically been volatile, revealing that a value strategy may often take
negative carry positions. Its market return component has been steadily earning positive returns historically.
Finally, a momentum factor seems to be halfway between carry and value in terms of performance attribution
profile. It historically earned positive carry returns but in a less steady fashion than the carry factor, while its
market returns have historically also contributed positively. It is not surprising to see the momentum factor
capturing large carry returns, as ranking bond futures on past performance tends to favour those bonds with high
carry.

Figure 11: Performance Attribution

Carry Factor Value Factor Momentum Factor

700% 700% 400% 200%
180%
600% 600% 350%
160%
500% 300%
500% 140%
400% 250% 120%
400%
200% 100%
300%
300%
150% 80%
200% 60%
200%
100%
100% 40%
100% 50% 20%
0%
0% 0% 0%
1962

1974

1986

1997

2009
1960

1965
1968
1971

1977
1980
1983

1989
1992
1995

2000
2003
2006

2012
2015
1960
1964
1969
1974
1979
1984
1989
1993
1998
2003
2008
2013

1961
1966
1970
1975
1980
1985
1989
1994
1999
2004
2008
2013

1961
1965
1970
1975
1980
1984
1989
1994
1999
2003
2008
2013
Carry Returns Market Returns
Determined using monthly data from Jan 1960 to Jun 2017. Source: Bloomberg, Barclays
Long-Only Basket is computed as an equally-weighted basket of 10-year bond futures across U.S., Germany, U.K., Japan, Canada and Australia,
rebalanced monthly. Historical and hypothetical/simulated performance is not indicative of future performance or future results.

Diversification Benefit

The above performance attribution analysis highlights that sources of returns are different for each factor. Figure
12 provides a consistent picture and emphasizes that the three rates factors exhibit fairly low correlation among
themselves over the full sample period. It is worth noting that value and momentum are negatively correlated
which is a reflection of the fact that value is a mean-reversion play. Bond futures that experienced the highest
uptrend in the recent past, away from their value measure, will tend to be selected as shorts and vice-versa, hence
the opposite profile to a momentum factor.

This is not a product of Barclays Research. This is a product of Barclays Sales and Trading.
For Professional and Institutional Investors Only. Not for Retail Distribution.
Investors in specific Asia Pacific countries should review important disclosures at the end of this document.
10
 Figure 12: Correlations of single factor strategies

Carry Factor Value Factor Momentum Factor

Carry Factor 100%

Value Factor 30% 100%

Momentum Factor 22% -2% 100%

Determined using monthly data from Feb 1962 to Jun 2017. Source: Bloomberg, Barclays
Historical and hypothetical/simulated performance is not indicative of future performance or future results.

Low correlations tend to suggest diversification benefit at the multi-factor level. While the carry factor on its own
delivered attractive risk-adjusted returns, combinations of two or three factors historically benefitted, in some
cases, from improved performance statistics. For instance, as illustrated on Figure 13, combining carry and value
delivered, over the full sample period, a Sharpe ratio of 0.80 versus 0.72 and 0.56 for each of carry and value
factors. Additionally, carry and value combined exhibited a positive skew higher than any of the two underlying
strategies, while its maximum drawdown was over the full sample period -13.3% compared to -15.1% and
-14.9% for each of the carry and value factors, respectively.

Figure 13: Diversification benefit

Value & Carry & Carry, Value
Momentum Carry & Value
Carry Factor Value Factor Momentum Momentum & Momentum
Factor Factor
Factor Factor Factor
Annualised
3.7% 2.7% 1.7% 3.1% 2.2% 2.7% 2.7%
Excess Return
Annualised
5.0% 4.7% 7.2% 3.9% 4.3% 4.8% 3.8%
Volatility

Sharpe Ratio 0.72 0.56 0.23 0.80 0.53 0.56 0.73

Maximum
-15.1% -14.9% -32.3% -13.3% -14.3% -20.1% -12.4%
Drawdown

Skew 0.33 0.45 -1.06 0.65 0.00 0.03 0.48

Determined using monthly data from Apr 1961 to Jun 2017. Source: Bloomberg, Barclays
Historical and hypothetical/simulated performance is not indicative of future performance or future results.

Macro Properties

Our factors were constructed using a notional10 weighting approach, allocating each month 50% weight to each
of the top two ranked bond futures, and -50% weight to each of the bottom two ranked futures. One may see
such an approach, which is agnostic to duration considerations, as prone to creating unintended beta or
directionality in each factor. We argue that duration weighting across countries makes little sense, as it implicitly
assumes a one-for-one parallel shift across all country curves which is far from realistic. For instance, the volatility
of 10-year JGB futures have been historically about 50% that of 10-year US Treasury futures, and only 35% for the
three years ending in December 2017. Similarly, over the past three years, when 10-year US Treasury yields
moved by 1bp, 10-year JGB yields were more likely to move by around 0.25bp. Duration weighting only makes
sense in our view across tenors within a given curve, as parallel shift i.e. the first principal component is
responsible for more than 90% of the yield curve variability. This reasoning leads us to conclude that volatility
weighting or even beta weighting versus some form of benchmark should be worth considering. These more

10 Dollar equivalent notional

This is not a product of Barclays Research. This is a product of Barclays Sales and Trading.
For Professional and Institutional Investors Only. Not for Retail Distribution.
Investors in specific Asia Pacific countries should review important disclosures at the end of this document.
11
advanced weighting schemes have their benefits and drawbacks11, but this is beyond the scope of this short
study.

In this section, we investigate the behaviour of rates factors conditional on different yield environments, in order to
give us comfort that such observed historical returns were not merely a reflection of the very biased yield
environment experienced over the past three decades. In particular, we sort monthly changes in yields of the
Bloomberg Barclays Global Treasury Total Return Index12 in quartiles and show corresponding average returns for
each factor strategy.

While carry showed, over the full sample period, a good balance across different yield environments, this has been
less the case for momentum and value. In particular, momentum clearly exhibited negative beta to global yields.
This sensitivity has been most prominent in the bond bull period over the last three decades, as a momentum
factor naturally allocated long positions to bond futures with highest beta while taking short positions on those
with lowest beta, hence having a net positive beta to the bond market. Value, being a mean-reversion play,
exhibited the opposite profile to momentum, albeit with much smaller absolute sensitivity to global bond yields.

An obvious benefit of combining factor strategies into multi-factor portfolios has been to capture returns in a
more stable fashion through different yield environments. This is illustrated on Figure 15 with the conditional
profiles of each multi-factor strategy showing some balance across the different quartiles.

Figure 14: Annualised monthly returns (annualised) of single rates factors conditional on
global yield environments sorted by quartiles
5.0%

4.0%

3.0%

2.0%

1.0%

0.0%

-1.0%

-2.0%
Carry Value Momentum

Δyields Q1 (Falling Yields) Δyields Q2 Δyields Q3 Δyields Q4 (Rising Yields)

Determined using monthly data from Feb 1962 to Jun 2017. Source: Bloomberg, Barclays
Δyields Q1 to Q4 represent monthly changes in treasury yields sorted by quartiles, where treasury yields means
the Bloomberg Barclays Global Treasury Yield to Worst (BTSYYW Index) from Jan 1987 onwards, and US Generic
Government 10Y Yield (USGG10YR Index) before that.
Historical and hypothetical/simulated performance is not indicative of future performance or future results.

11 For instance, an unconstrained volatility weighting approach may lead to large concentration in 10-year JGB futures for the same
reason as mentioned previously.
12 Bloomberg Barclays Global Treasury Yield to Worst (BTSYYW Index) from Jan 1987 onwards, and US Generic Government 10Y Yield

(USGG10YR Index) before that.

This is not a product of Barclays Research. This is a product of Barclays Sales and Trading.
For Professional and Institutional Investors Only. Not for Retail Distribution.
Investors in specific Asia Pacific countries should review important disclosures at the end of this document.
12
 Figure 15: Annualised monthly returns (annualised) of rates multi-factors conditional on
global yield environments sorted by quartiles
5.0%

4.0%

3.0%

2.0%

1.0%

0.0%

-1.0%

-2.0%
Carry + Value Value + Momentum Carry + Momentum Carry + Value +
Momentum

Δyields Q1 (Falling Yields) Δyields Q2 Δyields Q3 Δyields Q4 (Rising Yields)

Determined using monthly data from Feb 1962 to Jun 2017. Source: Bloomberg, Barclays
Δyields Q1 to Q4 represent monthly changes in treasury yields sorted by quartiles, where treasury yields means
the Bloomberg Barclays Global Treasury Yield to Worst (BTSYYW Index) from Jan 1987 onwards, and US Generic
Government 10Y Yield (USGG10YR Index) before that.
Historical and hypothetical/simulated performance is not indicative of future performance or future results.

Concluding Remarks

In this study, we defined and examined three systematic investment styles in interest rate markets, namely carry,
value and momentum. These were constructed cross-sectionally taking equal notional long and short exposure,
hence reducing the impact of rates movements and reducing the positive bias that time-series strategies typically
have to bond rally environments. While each single factor has historically delivered attractive risk-adjusted returns,
we emphasized that a multi-factor approach, combining single factors with equal weights, provides an attractive
solution aiming to address essential investors’ needs in the current environment: capturing excess returns in both
rising and falling yield environments.

This is not a product of Barclays Research. This is a product of Barclays Sales and Trading.
For Professional and Institutional Investors Only. Not for Retail Distribution.
Investors in specific Asia Pacific countries should review important disclosures at the end of this document.
13
Appendix 1 - Value Measure
As described earlier in this study, economic theory suggests that the yield on a risk-free government bond should
approximately equal the rate of growth in the economy plus the rate of inflation, in other words the rate of nominal
growth. This relationship should hold over the long run and for long-term yields on which central banks’ monetary
policy typically has no direct influence. Below figures represent this historical relationship for 10-year yields and for
the six markets considered in this paper.

Figure 16: 10-year nominal yields versus nominal GDP growth rates
20% 16% y = 0.6315x + 0.0268
US Govt Bonds 10Y Yield
R² = 0.355
US Nominal GDP Growth Rate YoY

US Nominal GDP Growth Rate YoY

15% 12%

8%
10%

4%
5%

0%
0% 0.00% 5.00% 10.00% 15.00% 20.00%
1985

2013
1961
1965
1969
1973
1977
1981

1989
1993
1997
2001
2005
2009

-4%
-5% US Govt Bonds 10Y Yield

20% Germany Govt Bonds 10Y Yield 22% y = 0.8496x + 0.0042

Germany Nominal GDP Growth Rate YoY R² = 0.3104
Germany Nominal GDP Growth Rate

18%
15%
14%
10%
10%
YoY

5% 6%

0% 2%
1977

2005
1961
1965
1969
1973

1981
1985
1989
1993
1997
2001

2009
2013

-5.00% -2%0.00% 5.00% 10.00% 15.00%

-5%
-6%
-10% Germany Govt Bonds 10Y Yield

30% UK Govt Bonds 10Y Yield 30% y = 1.1327x - 0.0062

UK Nominal GDP Growth Rate YoY R² = 0.6045
UK Nominal GDP Growth Rate YoY

25% 26%
22%
20%
18%
15%
14%
10% 10%

5% 6%
2%
0%
-2%0.00%
1977

2005
1961
1965
1969
1973

1981
1985
1989
1993
1997
2001

2009
2013

5.00% 10.00% 15.00% 20.00%

-5%
-6%
-10% UK Govt Bonds 10Y Yield

This is not a product of Barclays Research. This is a product of Barclays Sales and Trading.
For Professional and Institutional Investors Only. Not for Retail Distribution.
Investors in specific Asia Pacific countries should review important disclosures at the end of this document.

14
30% Japan Govt Bonds 10Y Yield y = 1.2539x - 0.013
Japan Nominal GDP Growth Rate YoY 24%
R² = 0.6274

Japan Nominal GDP Growth Rate

25%
20%
20%
16%
15%
12%

YoY
10%
8%
5%
4%
0%
0%
1976

2012
1979
1982
1985
1988
1991
1994
1997
2000
2003
2006
2009

2015
-5% -5.00% 0.00% 5.00% 10.00%
-4%
-10% -8%
-15% Japan Govt Bonds 10Y Yield

25% Canada Govt Bonds 10Y Yield y = 0.68x + 0.027

Canada Nominal GDP Growth Rate YoY 24%
R² = 0.2414

Canada Nominal GDP Growth Rate

20% 20%

15% 16%
12%
10%
YoY

8%
5% 4%

0% 0%
0.00% 5.00% 10.00% 15.00% 20.00%
1981

2009
1961
1965
1969
1973
1977

1985
1989
1993
1997
2001
2005

2013

-4%
-5%
-8%
-10% Canada Govt Bonds 10Y Yield

30% Australia Govt Bonds 10Y Yield

24% y = 0.7115x + 0.0284
Australia Nominal GDP Growth Rate YoY
Australia Nominal GDP Growth Rate

R² = 0.2981
25% 20%

20% 16%
12%
15%
YoY

8%
10% 4%

5% 0%
0.00% 5.00% 10.00% 15.00% 20.00%
-4%
0%
-8%
1971
1974
1978
1981
1984
1988
1991
1994
1998
2001
2004
2008
2011
2014

-5% Australia Govt Bonds 10Y Yield

Source: Bloomberg, Barclays. Determined using monthly data from Jan 1960 to Jun 2017.
Historical and hypothetical/simulated performance is not indicative of future performance or future results.

This is not a product of Barclays Research. This is a product of Barclays Sales and Trading.
For Professional and Institutional Investors Only. Not for Retail Distribution.
Investors in specific Asia Pacific countries should review important disclosures at the end of this document.

15
Appendix 2 – Index Information
RISK PREMIUM TICKER INDEX NAME INDEX LIVE DATE INDEX BASE DATE

Rates Carry BXIIRCUE Barclays Rates Carry Factor Index Nov 2017 Sep 1999

Rates Value BXIIRVUE Barclays Rates Value Factor Index Nov 2017 Jun 1999

Rates Momentum BXIIRMUE Barclays Rates Momentum Factor Index Nov 2017 Oct 2000

BXIIRCVU Barclays Rates Carry & Value Factor Index Nov 2017 Oct 2000

BXIIRCMU Barclays Rates Carry & Momentum Factor Index Nov 2017 Oct 2000
Rates Multi Factor
BXIIRVMU Barclays Rates Value & Momentum Index Nov 2017 Oct 2000

BXIIRCVM Barclays Rates Carry, Value & Momentum Index Nov 2017 Oct 2000

This is not a product of Barclays Research. This is a product of Barclays Sales and Trading.
For Professional and Institutional Investors Only. Not for Retail Distribution.
Investors in specific Asia Pacific countries should review important disclosures at the end of this document.

16
References
Asness C. S., Moskowitz T. J. and Pedersen L. H. (2013), “Value and Momentum Everywhere”

Baz J., Granger N., Harvey C. R., Le Roux N and Rattray S. (2015), “Dissecting Investment Strategies in the Cross
Section and Time Series”

Desclee A., Polbennikov S. and Katenko E. (2017), “Carry Strategies in Global Rates Markets”

Fattouche C., (2016), “Barclays QIS Insights: Tactical Allocation to Alternative Risk Premia”

Hurst B., Hua Ooi Y. and Pedersen L. H. (2014), “A Century of Evidence on Trend-Following Investing”

Ilmanen A. (2011), “Expected Returns” - Wiley

Ilmanen A., Israel R., Moskowitz T. J. (2012), “Investing with Style - The Case for Style Investing”

Jacoby P., (2017), “Barclays QIS Insights: Alternative Risk Premia: Macro Factor Analysis”

Jegadeesh N. and Titman S. (1993), “Returns to buying winners and selling losers: Implications for stock market
efficiency”

Jegadeesh N. and Titman S. (2001), “Profitability of momentum strategies: An evaluation of alternative

explanations”

Koijen R. S.J., Moskowitz T. J., Pedersen L. H. and Vrugt E B. (2013), “Carry”

Rennison G., Davis J. and Dorsten M. (2016). “The Carry and Value Pendulum”

This is not a product of Barclays Research. This is a product of Barclays Sales and Trading.
For Professional and Institutional Investors Only. Not for Retail Distribution.
Investors in specific Asia Pacific countries should review important disclosures at the end of this document.

17
Disclaimer
BARCLAYS This communication has been prepared by Barclays.

“Barclays” means any entity within the Barclays Group of companies, where “Barclays Group”
means Barclays Bank PLC, Barclays PLC and any of their subsidiaries, affiliates, ultimate holding
company and any subsidiaries or affiliates of such holding company.

CONFLICTS OF BARCLAYS IS A FULL SERVICE INVESTMENT BANK. In the normal course of offering investment
INTEREST banking products and services to clients, Barclays may act in several capacities (including
issuer, market maker and/or liquidity provider, underwriter, distributor, index sponsor, swap
counterparty and calculation agent) simultaneously with respect to a product, giving rise to
potential conflicts of interest which may impact the performance of a product.

NOT RESEARCH This communication is from a Barclays Sales and/or Trading desk and is not a product of the
Barclays Research department. Any views expressed may differ from those of Barclays
Research.

Any views and/or commentary in this communication are of the Barclays Sales and/or Trading
desk from which it originates (the “Authors”). This communication has not been produced,
reviewed or approved by Barclays’s Research department, and is not subject to any prohibition
on dealing ahead of the dissemination of research. Any views expressed herein may not be
objective or independent of the interests of the Authors or other Barclays Sales and/or Trading
desks, who are active participants in the markets, investments or strategies referred to in this
communication. This communication is not a personal recommendation and do not take into
account whether any product or transaction is suitable for any particular investor.

BARCLAYS POSITIONS Barclays may at any time acquire, hold or dispose of long or short positions (including hedging
and trading positions) and trade or otherwise effect transactions for their own account or the
account of their customers in the products referred to herein which may impact the
performance of a product.

PRIVATE INFORMATION BARCLAYS MAY HAVE PRIVATE INFORMATION ABOUT ANY PRODUCT AND/OR THE
UNDERLYING ASSETS REFERENCED BY THE PRODUCT. It is not obligated to disclose any such
information to investors or counterparties.

FOR INFORMATION THIS COMMUNICATION IS PROVIDED FOR INFORMATION PURPOSES ONLY AND IT IS
ONLY SUBJECT TO CHANGE. IT IS INDICATIVE ONLY AND IS NOT BINDING.

NO OFFER Barclays is not offering to sell or seeking offers to buy any product or enter into any transaction.
Any offer or entry into any transaction requires Barclays’ subsequent formal agreement which
will be subject to internal approvals and execution of binding transaction documents.

NO LIABILITY Neither Barclays nor any of its directors, officers, employees, representatives or agents, accepts
any liability whatsoever for any direct, indirect or consequential losses (in contract, tort or
otherwise) arising from the use of this communication or its contents or reliance on the
information contained herein, except to the extent this would be prohibited by law or
regulation.

NO ADVICE Barclays is acting solely as principal and not as fiduciary. Barclays does not provide, and has not
provided, any investment advice or personal recommendation to you in relation to the
transaction and/or any related securities described herein and is not responsible for providing
or arranging for the provision of any general financial, strategic or specialist advice, including
legal, regulatory, accounting, model auditing or taxation advice or services or any other services
in relation to the transaction and/or any related securities described herein. Accordingly
Barclays is under no obligation to, and shall not, determine the suitability for you of the

This is not a product of Barclays Research. This is a product of Barclays Sales and Trading.
For Professional and Institutional Investors Only. Not for Retail Distribution.
Investors in specific Asia Pacific countries should review important disclosures at the end of this document.

18
 transaction described herein. You must determine, on your own behalf or through independent
professional advice, the merits, terms, conditions and risks of the transaction described herein.

THIRD PARTY Barclays is not responsible for information stated to be obtained or derived from third party
INFORMATION sources or statistical services.
DISTRIBUTION All laws and regulations in any relevant jurisdiction(s) must be complied with when offering,
marketing or selling a Product or distributing offering materials.

PAST & SIMULATED Any past or simulated past performance including back-testing, modelling or scenario analysis
PAST PERFORMANCE contained herein is no indication as to future performance or future results.

No representation is made as to the accuracy of the assumptions made within, or

completeness of, any modelling, scenario analysis or back-testing.

OPINIONS SUBJECT TO All opinions and estimates are given as of the date hereof and are subject to change. The value
CHANGE of any investment may also fluctuate as a result of market changes. Barclays is not obliged to
inform the recipients of this communication of any change to such opinions or estimates.

NOT FOR RETAIL This communication is being directed at persons who are professionals and is not intended for
retail customer use.

REGULATORY Information relating to an investment may be disclosed when required by regulators or other
DISCLOSURE authorities, including tax authorities.

TAX DISCLOSURE All discussions and any related materials relating to the tax treatment or tax structure of any
transactions described in this document (including any attachments) may be disclosed without
limitation. This authorisation of tax disclosure supersedes anything to the contrary contained
in this document or otherwise communicated.

IMPORTANT For important regional disclosures you must read, click on the link relevant to your region.
DISCLOSURES Please contact your Barclays representative if you are unable to access.

EMEA EMEA Disclosures

APAC APAC Disclosures

U.S. US Disclosures

CONFIDENTIAL This communication is confidential and is for the benefit and internal use of the recipient for
the purpose of considering the securities/transaction described herein, and no part of it may be
reproduced, distributed or transmitted without the prior written permission of Barclays.

ABOUT BARCLAYS Barclays offers premier investment banking products and services to its clients through
Barclays Bank PLC. Barclays Bank PLC is authorised by the Prudential Regulation Authority and
regulated by the Financial Conduct Authority and the Prudential Regulation Authority and is a
member of the London Stock Exchange. Barclays Bank PLC is registered in England No.
1026167 with its registered office at 1 Churchill Place, London E14 5HP.

Barclays Capital Inc. is a US registered broker/dealer affiliate of Barclays Bank PLC and a
member of SIPC, FINRA and NFA. Barclays Capital Inc. operates out of 745 Seventh Avenue,
New York, NY 10019. Where required pursuant to applicable US laws, rules and/or regulations,
Barclays Capital Inc. accepts responsibility for the distribution of this document in the United
States to U.S. Persons. Where a communication is being directed at persons who are
professionals, it is directed at institutional investors in the U.S. as defined by FINRA Rule
2210(a)(4).

In Canada, Barclays Capital Inc. relies on the International Dealer Exemption in various
provinces to engage in securities trading with Institutional clients who qualify as Permitted
Clients under National Instrument 31-103.

This is not a product of Barclays Research. This is a product of Barclays Sales and Trading.
For Professional and Institutional Investors Only. Not for Retail Distribution.
Investors in specific Asia Pacific countries should review important disclosures at the end of this document.

19
 Barclays Securities Japan Limited is a joint-stock company incorporated in Japan with registered
office of 6-10-1, Roppongi, Minato-ku, Tokyo 106-6131, Japan. It is a subsidiary of Barclays
Bank PLC and a registered financial instruments firm regulated by the Financial Services Agency
of Japan. Registered Number: Kanto Zaimukyokucho (kinsho) No. 143.
This material is prepared and distributed to persons in Australia, it may be distributed by
Barclays Bank plc or by Barclays Capital Inc., Barclays Capital Securities Limited or Barclays
Capital Asia Limited. None of Barclays Bank plc, nor any of the other referenced Barclays group
entities, hold an Australian financial services licence and instead they each rely on an
exemption from the requirement to hold such a licence. This material is intended to only be
distributed to “wholesale clients” as defined by the Australian Corporations Act 2001.

Barclays Bank PLC (DIFC Branch) (Registered No. 0060) is regulated by the Dubai Financial
Services Authority (DFSA). Barclays Bank PLC (DIFC Branch) may only undertake the financial
services activities that fall within the scope of its existing DFSA licence. Related financial
products or services are only available to Market Counterparties, as defined by the Dubai
Financial Services Authority. Principal place of business in the DIFC: Dubai International
Financial Centre, The Gate Village Building No. 4, Level 4, PO Box 506504, Dubai, U.A.E.

COPYRIGHT © Copyright Barclays Bank PLC, 2018 (all rights reserved).

© Copyright Barclays Securities Japan Limited, 2018 (all rights reserved).

The Shiller Barclays CAPE® Index Family has been developed in part by RSBB-I, LLC, the research principal of which is
Robert J. Shiller. RSBB-I, LLC is not an investment advisor, and does not guarantee the accuracy or completeness of the
Shiller Barclays CAPE® Index Family, or any data or methodology either included therein or upon which it is based.
Neither RSBB-I, LLC nor Robert J. Shiller shall have any liability for any errors, omissions, or interruptions therein, and
makes no warranties, express or implied, as to performance or results experienced by any party from the use of any
information included therein or upon which it is based, and expressly disclaims all warranties of merchantability or
fitness for a particular purpose with respect thereto, and shall not be liable for any claims or losses of any nature in
connection with the use of such information, including but not limited to, lost profits or punitive or consequential
damages, even if RSBB-I, LLC is advised of the possibility of same.

Important Disclosures for APAC Countries

For below Asia Pacific countries, please note the relevant important disclosures:

AUSTRALIA “Wholesale client” as defined by the Australian Corporations Act 2001.

Please further note that the Australian Securities and Investments Commission (ASIC) has
provided certain exemptions to Barclays Capital Asia Limited 巴克萊亞洲有限公司 (BCAL),
Barclays Capital Securities Limited (BCSL) and Barclays Capital Inc. (BCI) under paragraph
911A(2)(l) of the Corporations Act 2001 from the requirement to hold an Australian financial
services licence (AFSL) in respect of financial services provided to Australian wholesale clients,
on the basis that:

• BCAL is regulated by the Hong Kong Securities and Futures Commission (HK SFC) under
Hong Kong laws.
• BCSL is authorised by the Prudential Regulation Authority of the United Kingdom (PRA)
and regulated by the Financial Conduct Authority (FCA) of the United Kingdom and the
PRA under United Kingdom laws.
• BCI is regulated by the United States Securities and Exchange Commission (US SEC) under
United States laws.

All of the abovementioned jurisdictions have laws which differ from Australian laws. When
providing financial services to Australian wholesale clients, BCAL, BCSL and BCI rely on the

This is not a product of Barclays Research. This is a product of Barclays Sales and Trading.
For Professional and Institutional Investors Only. Not for Retail Distribution.
Investors in specific Asia Pacific countries should review important disclosures at the end of this document.

20
 relevant exemption from the requirement to hold an AFSL. Accordingly, BCAL, BCSL and BCI do
not hold an AFSL.

Barclays Bank PLC has relinquished its AFSL and going forward will provide financial services to
Australian clients by relying on certain exemptions that are available to it in relation to the
requirement to hold an AFSL.

SINGAPORE For Accredited, Institutional and Professional Investors Only. Not for Further Distribution or for
Distribution to Retail Customers

SOUTH KOREA For Locally Licensed Investment Dealer or Investment Broker Only. Not for Retail Distribution.

INDIA The strategies may not be permitted in India and may have to be amended in line with India
regulations.

HONG KONG “Professional investor” as defined by the Securities and Futures Ordinance (“SFO”) for the
Hong Kong Special Administrative Region.

JAPAN “Professional investor” as defined by the Financial Instruments and Exchange Laws (“FIEL”) for
the Japan region. It is not intended for persons (i) who do not fall within the definition of
professional investor as specified in Article 2, paragraph 31 of the FIEL and the relevant cabinet
ordinances thereof or (ii) who are professional investors but regarded as not being professional
investors due to application of Article 34-2, paragraph 5 of the FIEL.

NEW ZEALAND For Wholesale Clients Only. Not for Further Distribution or for Distribution to Retail Customers.

BARCLAYS BANK PLC IS NOT A REGISTERED BANK IN NEW ZEALAND

IMPORTANT DISCLOSURE FOR NEW ZEALAND

This document and the information contained in or accompanying this document are not, and
are under no circumstances to be construed as, an offer of financial products for issue
requiring disclosure to an investor under Part 3 of the Financial Markets Conduct Act 2013 (the
“FMCA”). This document and the information contained in or accompanying this document
have not be registered, filed with or approved by any New Zealand regulatory authority or
under or in accordance with FMCA. This document and the information contained in or
accompanying this document is not a disclosure document under New Zealand law and do not
contain all the information that a disclosure document is required to contain under New
Zealand law. Any offer or sale of any product described in these materials in New Zealand will
be made only in accordance with FMCA:

(a) to a person who is an investment business as specified in FMCA; or

(b) to a person who meets the investment activity criteria specified in FMCA; or

(c) to a person who is large as defined in the FMCA; or

(d) to a person who is a government agency as defined in the FMCA; or

(e) in other circumstances where there is no contravention of the FMCA (or any statutory
modification or re-enactment of, or statutory substitution for, the FMCA).

This is not a product of Barclays Research. This is a product of Barclays Sales and Trading.
For Professional and Institutional Investors Only. Not for Retail Distribution.
Investors in specific Asia Pacific countries should review important disclosures at the end of this document.

21