TO HEDGE OR NOT TO HEDGE?

INVESTING IN FOREIGN BONDS: TO HEDGE OR NOT TO HEDGE? APRIL 2017 The asset manager for a changing world

Investing in foreign bonds: To hedge or not to hedge? I April 2017 I 3 I SUMMARY Many European institutional investors broaden their investment choices beyond their own domestic markets into other countries' financial markets. One of the crucial factors in assessing an investment opportunity in a foreign market is the hedging cost of the currency associated with that investment. In particular, European insurers who diversify their portfolios into assets denominated in currencies other than the euro, tend to hedge the underlying currency risk because the Solvency Capital Requirement (SCR) is higher for currency exposure and thus reduces the assets' attractiveness in terms of the SCR buffer requirements compared to the expected returns. Pension funds with future liabilities denominated in euros also tend to hedge the currency risks in their portfolios. An additional risk-related aspect that may be relevant to some investors is the tracking error against the portfolio's benchmark; simply put, how correlated the foreign investment is to its domestic counterpart within the benchmark. In this paper, we focus our currency hedging analysis on investments in foreign bonds, primarily for two reasons. Compared with equity investors, bond investors tend to hold their bonds for longer, thus the foreign exchange (FX) hedging cost for a potentially longer period of time becomes more pronounced. More importantly, as the total return from bonds is in general much smaller than that from equities, the FX hedging cost takes up a larger proportion of the total return from a bond investment than that from an equity investment. For eurozone and non-eurozone government bonds of the same duration and with the same sovereign risk, the only differences are their yields and their respective currencies. In such cases, we can estimate a breakeven between the 'pick-up' in yield vs. the currency hedging cost. In other words, how far the currency hedging cost can increase until an investor loses the yield 'advantage'. Koye Somefun, PhD Head of Retirement Solutions & Innovation, Multi Asset Solutions Raymond Yung, CFA Senior Portfolio Manager, Retirement Solutions & Innovation Multi Asset Solutions Here, we first look at the principal concept linking the foreign exchange forward price to fixed-income markets, as expressed in an equation. Next, we explain our recommended approach to estimating the FX hedging cost by applying this equation on swap and FX forward markets. We then use this approach on selected major government bond and credit markets. The results from our analysis show that currency hedging does indeed play a crucial role in the total return of a foreign bond investment. After taking into account the FX hedging cost, it appears that there are no compelling investment choices from the selected foreign government bonds, and that only Asian credit stands out as attractive among Asian, UK and US credit choices.

Investing in foreign bonds: To hedge or not to hedge? I April 2017 I 4 I I. CONCEPT OVERVIEW Our conceptual starting point from which we estimate the FX hedging cost is the theoretical pricing of the forward derived from its spot price, interest rate differential and investment horizon. We assume an arbitrage-free environment, as shown in Equation 1 below. Based on this equation, we can look at different financial markets to work out the "implied" FX hedging cost. There are two different approaches to estimating the hedging cost: 1) Deriving information from the current swap and FX forward markets. 2) Basing the estimate on the historical relationship between FX forward and spot prices. We review these separately using USD/EUR forwards (EUR into USD) as an example in the next section. We use the first method to proxy the hedging costs of a broad range of government bond and credit markets. The second method has the added benefit of allowing us to put our current estimated hedging cost into a historical context. Equation 1 (Pricing of USD/EUR forward price): FUSDEUR S e (r eur r USDEUR * usd)t This equation can also be written in the following way. ( r eur r usd ) T In(F USDEUR ) In(S USDEUR ) F USDEUR = Forward price (EUR into USD); S USDEUR = Spot price (EUR into USD); r eur = Risk-free EUR interest rate per annum r usd = Risk-free US interest rate per annum T = Investment horizon II. OUR APPROACH The swap and FX forward markets are two liquid financial markets where many institutional investors conduct their trades. We thus view them as valid sources from which we can derive the FX hedging cost. We first estimate the total FX hedging cost based on pricing information from the FX forward market. Next, we provide a breakdown of the two components of that total cost. The first component is the basic cost, which is the 3-month interestrate differentials of the two countries, as derived from the swap market. The second component is the additional cost charged by brokers (in excess of the 3-month interest-rate differential), which we define as the difference between the total cost and the basic cost. This breakdown of total cost gives us a clearer picture of the hedging cost. In addition, as much as the FX forward market is a reasonably liquid market, the liquidity of certain currency pairs (crosses) beyond three years may be difficult to foresee. We thus believe that it is more prudent to compute the first component (basic cost) as an anchor point and 'sanity check'. For illustrative purposes, we use the estimate of the hedging cost of EUR into USD as an example. 1) Total cost: We estimate the total cost of FX hedging by extracting information from the FX forward market as follows. First, we construct a term structure based on the current FX forward prices across different maturities. Then we derive the projected hedging returns (annualised return based on a rolling 3-month return) by extrapolating the return between two adjacent points of this 'synthetic' term structure. As the resultant estimate is derived from the forward rates in a risk neutral/arbitrage-free environment, we believe it offers a reasonable and representative estimation of the hedging cost. However, as with most estimates, the actual cost realised may deviate from an estimate made at a specific point of time with limited information. In Chart 1 on the next page, the solid blue line is the 3-month FX hedging of USD into EUR projected returns derived from the FX forward market (an average of -2.18%). 2) Basic cost: For this analysis, we assume hedging is carried out by rolling FX contracts on a 3-month basis. Alternatively, one could instead use different rebalancing frequencies, or construct a 'perfect' hedge where the maturity of the FX contract exactly matches that of a buy-and-hold investment in foreign government bonds. Our rationale for using the rolling 3-month approach is to mitigate the 'pin risk' (execution of hedging trade 'pinned' on pricing at a specific point of time) by re-hedging (rebalancing) every three months. The choice of rolling is based on our view that the 'pin risk' outweighs the reinvestment risk. Put another way, if the upward slope of the term structure of the swap curve holds in general, the cost by rolling the hedge will be lower than that of a one-off hedge for the entire period. By the same principle, hedging on a rolling 1-month basis could even be cheaper; however, from a regulatory point of view, rolling on a 3-month basis is the shortest frequency often accepted (e.g. under the Solvency II directive). As outlined in Equation 1, the hedging cost can be estimated from the 3-month interest-rate differential between the US dollar and the euro, derived from their respective current swap curves. In Chart 1 on the next page, the dotted green line is the 3-month FX hedging projected returns derived from the USD and EUR swap curves (an average of -1.78%).

Investing in foreign bonds: To hedge or not to hedge? I April 2017 I 5 I Chart 1: Projection of 3-month hedging returns (from swap and forward markets) 4% 3% 2% 1% 0% -1% -2% -3% Apr.-17 Apr.-19 Apr.-21 Apr.-23 Apr.-25 USD interest rate Return from swap market Data as of 30 December 2016 EUR interest rate Return from forward market 3) Additional cost (brokers' fees): The additional cost to the FX hedging relates to brokers' market-making, capital utilisation and trade facilitation fees. We proxy this as the difference between the hedging cost derived from the FX forwards (total cost) and swap curves (basic cost). In Chart 1, it is essentially the gap between the blue line (total cost) and the dotted green line (basic cost). The average of the estimated additional cost is 40 basis points (bp). 4) Reference: Historical returns of hedging USD (vs. EUR): As a point of reference, we compute the annualised historical returns of hedging the US dollar against the euro by extrapolating the interest-rate differential between the historical 3-month forwards and spots outlined in Equation 1. As shown in Chart 2, there are periods of high negative return on hedging the USD exposure. On the other hand, between 2000 and 2003, the opposite was true, with strong positive returns. The unpredictable outcome of ex-post hedging returns and its backward-looking nature (not incorporating what is currently priced in and projected by the market) are the main drawbacks of this historical approach. We observed only limited periods when the costs of hedging the US dollar against the euro reached 3%. Looking ahead, we would expect the returns on hedging to remain negative. Chart 2: Historical return on hedging USD exposure vs. EUR 3% 2% 1% 0% -1% -2% -3% -4% Feb.-99 Aug.-99 Feb.-00 Aug.-00 Feb.-01 Aug.-01 Feb.-02 Aug.-02 Feb.-03 Aug.-03 Feb.-04 Aug.-04 Feb.-05 Aug.-05 Feb.-06 Aug.-06 Feb.-07 Aug.-07 Feb.-08 Aug.-08 Feb.-09 Aug.-09 Feb.-10 Aug.-10 Feb.-11 Aug.-11 Feb.-12 Aug.-12 Feb.-13 Aug.-13 Feb.-14 Aug.-14 Feb.-15 Aug.-15 Feb.-16 Aug.-16 Data from 26 February 1999 to 30 December 2016 III. RESULTS AND ANALYSIS A. SELECTED MAJOR GOVERNMENT BOND MARKETS We applied our approach to selected major government bond markets. The results are shown in Table 1 on the next page. To illustrate how we computed the yield differential after accounting for the total FX hedging cost (in Table 1 on the next page under the row labelled "Net Yield Differential AFTER Total Hedging Return"), we take hedging USD into EUR as an example. Net Yield Differential AFTER Total Hedging Return (US vs. EUR) = US Yield Differential (vs. Core-EUR) Total Hedging Cost (FX Forward Market). Thus: Net Yield Differential AFTER Total Hedging Return (US vs. EUR) = (2.44% - 0.21%) 2.18% = 2.24% - 2.18% = 0.05%. By comparing the yield differential of each country vs. the Core-EUR and the respective hedging return it turns out that: 1) In the UK, Australia and Canada, the net yield differentials are not attractive. 2) In the US, the net yield differential is essentially flat. 3) For an investment in 10-year Japanese government bonds, our analysis shows that the yield differential is negative (5bp - 21bp = -16bp). The hedging return is positive (+68bp) and is large enough to more than offset the negative yield differential, thus giving rise to a net return (estimated) of 52bp (68bp 16bp = 52bp).

Investing in foreign bonds: To hedge or not to hedge? I April 2017 I 6 I Table 1: Government bond yields and FX hedging returns As of 30 December 2016 Core-EUR US UK JP AU CN 10-Year Government Yield 0.21% 2.44% 1.24% 0.05% 2.77% 1.72% Yield Differential (vs. Core-EUR) 1 NA 2.24% 1.03% -0.16% 2.56% 1.51% FX Hedging Return (vs. EUR) Basic Cost (FI Swap Market) -1.78% -0.61% 0.38% -2.36% -1.41% Additional Cost (Total Cost minus Basic Cost) -0.40% -0.44% 0.30% -0.69% -0.39% Total Hedging Cost (FX Forward Market) -2.18% -1.05% 0.68% -3.04% -1.79% Net Yield Differential AFTER Total Hedging Return 0.05% -0.02% 0.52% -0.48% -0.28% Historical Hedging Return (February 1999 - December 2016) Average -0.34% -1.06% 1.85% -2.68% -0.58% Maximum 1.95% 0.59% 4.92% 0.09% 2.66% Minimum -3.02% -2.91% -0.18% -4.54% -3.01% 1 For simplicity, we only use the yield to proxy the return on bond investments ignoring other aspects such as yield roll down and future changes in the yield curve. Data as of 30 December 2016 B. ASIAN, UK AND US CREDIT We extended our analysis to Asian, UK and US credit. The results are shown in Table 2 on the next page. As credit from different countries can have different durations, for a fair comparison, we made some yield adjustments with reference to the duration of euro-denominated credit as well as the euro bond yield curve. The duration-related yield adjustments are listed below. 1) Asian credit: a) For Asian A rated credit, 18bp are added to the Asianover-euro credit spread, as the Asian credit's duration is approximately half a year shorter than that of its euro counterpart. Eighteen basis points equates to approximately half of the average* yield difference between five and six-year A rated euro credit. b) For Asian BBB rated credit, 19bp are subtracted from the Asian-over-euro credit spread, as the Asian credit's duration is approximately half a year longer than that of its euro counterpart. Nineteen basis points equates to approximately half of the average* yield difference between five and six-year BBB rated euro credit. * Daily average for the month of December 2016. Source: Bloomberg, BNP Paribas Investment Partners 2) US credit: a) For US A rated credit, 36bp are subtracted from the US-over-Euro credit spread, as the US credit's duration is approximately two years longer than that of its euro counterpart. Thirty-six basis points is approximately half of the average* yield difference between five and seven-year A rated euro credit. b) For US BBB rated credit, 37bp are subtracted from the US-over-euro credit spread, as the US credit's duration is approximately two years longer than that of its euro counterpart. Thirty-seven basis points is approximately half of the average* yield difference between five and seven-year BBB rated euro credit. * Daily average for the month of December 2016. Source: Bloomberg, BNP Paribas Investment Partners 3) UK credit: a) For UK A rated credit, 69bp are subtracted from the UK-over-euro credit spread, as the UK credit's duration is approximately four years longer than that of its euro counterpart. Sixty-nine basis points is approximately half of the average* yield difference between five and nine-year A rated euro credit. b) For UK BBB rated credit, 37bp are subtracted from the UK-over-euro credit spread, as the UK credit's duration is approximately two years longer than that of its euro counterpart. Thirty-seven basis points is approximately half of the average* yield difference between five and seven-year BBB rated euro credit. * Daily average for the month of December 2016. Source: Bloomberg, BNP Paribas Investment Partners To illustrate how the credit spread figures are calculated in Table 2 on the next page, we use the spread between Asian A rated credit over euro credit as an example. "Spread: Asia over euro" A rated credit = Duration-adjusted Asian A rated credit yield Euro A rated credit yield "Spread: Asia over euro" A rated credit = [Asian A rated credit yield + duration adjustment to match euro A credit duration] euro A rated credit yield

Investing in foreign bonds: To hedge or not to hedge? I April 2017 I 7 I "Spread: Asia over euro" A rated credit = [3.9% (from Table 3) +18bp (from 1a stated above)] 0.7% (from Table 3) = 3.4% In addition, "Spread Asia over euro hedged (-2.2%)" A rated credit = "Spread: Asia over euro" - USD hedging cost of 2.2% (from Table 1 on the previous page) = 1.2% HIGHLIGHTS 1) Based on our analysis, investing in Asian credit remains attractive for euro-based investors. For Asian credit debts issued in US dollars, after accounting for the total FX hedging cost of 2.2% (see Table 1 on the previous page) derived from the current 3-month FX forward curve, an investment in an A rated corporate bond and hedged would still generate a spread of approximately 1.2% (compared to a similar credit investment in EUR). 2) For both US and UK credits, the net returns are close to flat after accounting for the hedging costs. The investment argument here is thus less compelling. Table 2: Credit spreads Rating A BBB 1a) Spread: Asia over Euro 1 3.4 3.2 1b) Spread Asia over Euro hedged (-2.2%) 1.2 1.0 2a) Spread: US over Euro 1 2.1 2.2 2b) Spread US over Euro hedged (-2.2%) -0.1 0.0 3a) Spread: UK over Euro 1 1.2 1.4 3b) Spread Asia over Euro hedged (-1.1%) 0.1 0.3 1 Duration adjusted to match that of Euro Data as of 30 December 2016 Table 3: Yield by rating & country Rating A BBB Euro 0.7 1.2 Yield Asia 3.9 4.6 US 3.2 3.8 UK 2.6 3.0 Data as of 30 December 2016 Table 4: Duration by rating and country Rating A BBB Euro 5.3 5.2 Asia 4.7 5.6 Duration US 7.5 7.2 UK 9.6 7.2 Data as of 30 December 2016 IV. CONCLUSION We have estimated the FX hedging returns (derived from swap and FX forward markets) in selected major government bond and credit markets from the perspective of a euro-based investor. From our analysis and assessment, after accounting for the FX hedging returns, there are few good opportunities in the selected major government bond markets. Of the three credit markets selected (Asia, the UK and the US), only Asian credit looks attractive. APPENDIX Historical correlation of Asian and US credit to their European counterpart As an estimate of the deviation of a foreign instrument from its domestic counterpart in the benchmark, we compute the correlation between each. The relevance of this correlation is highly dependent on the investment objective and mandate of an individual investor. For investors who have liabilitymatching requirements, a high correlation of the foreign instrument to the benchmark is probably more desirable. On the other hand, for investors seeking greater diversification in their portfolios, a lower correlation may be more appealing. Chart 3 below shows that the US Aggregate 1 has a high correlation to its euro counterparty 1 while Asian Credit 1 has a much lower correlation. Chart 3: Rolling 2-year correlation with Euro Aggregrate 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 Sept.-07 March-08 Sept.-08 March-09 Sept.-09 March-10 Sept.-10 March-11 Asian Credit & Euro Aggregate Sept.-11 March-12 Sept.-12 March-13 Sept.-13 March-14 Sept.-14 March-15 Sept.-15 March-16 US Aggregate & Euro Aggregate Source: Barclays, J.P. Morgan, DataStream, Bloomberg & BNP Investment Partners. Data from 31 October 2005 to 30 December 2016 1 Index Composition: a) Asia Credit (J.P. Morgan Composite Total Return); b) US Aggregate (proxy as 70% Barclays US Treasury & 30% Barclays US Corp. IG); c) Euro Aggregate (proxy as 70% Barclays Euro Aggregate Government Index & 30% Barclays Euro Aggregate Corporate 500MM (Euro)) Sept.-16

This material is issued by BNP Paribas Asset Management S.A.S. (BNPP AM)*, a member of BNP Paribas Investment Partners (BNPP IP)**. This material is produced for information purposes only and does not constitute: 1. an offer to buy nor a solicitation to sell, nor shall it form the basis of or be relied upon in connection with any contract or commitment whatsoever or 2. any investment advice. Opinions included in this material constitute the judgment of BNPP AM at the time specified and may be subject to change without notice. BNPP AM is not obliged to update or alter the information or opinions contained within this material. Investors should consult their own legal and tax advisors in respect of legal, accounting, domicile and tax advice prior to investing in the Financial Instrument(s) in order to make an independent determination of the suitability and consequences of an investment therein, if permitted. Please note that different types of investments, if contained within this material, involve varying degrees of risk and there can be no assurance that any specific investment may either be suitable, appropriate or profitable for a client or prospective client's investment portfolio. Given the economic and market risks, there can be no assurance that the Financial Instrument(s) will achieve its/their investment objectives. Returns may be affected by, amongst other things, investment strategies or objectives of the Financial Instrument(s) and material market and economic conditions, including interest rates, market terms and general market conditions. The different strategies applied to the Financial Instruments may have a significant effect on the results portrayed in this material. Past performance is not a guide to future performance and the value of the investments in Financial Instrument(s) may go down as well as up. Changes in currency exchange rates may affect the value of your investment. Investors may not get back the amount they originally invested. The performance data, as applicable, reflected in this material, do not take into account the commissions, costs incurred on the issue and redemption and taxes. *BNPP AM is an investment manager registered with the "Autorité des marchés financiers" in France under number 96002, a simplified joint stock company with a capital of 70.300.752 euros with its registered office at 1, boulevard Haussmann 75009 Paris, France, RCS Paris 319 378 832. www.bnpparibas-ip.com. ** BNP Paribas Investment Partners is the global brand name of the BNP Paribas group's asset management services. The individual asset management entities within BNP Paribas Investment Partners if specified herein, are specified for information only and do not necessarily carry on business in your jurisdiction. For further information, please contact your locally licensed Investment Partner. March 2017 - Design : - P1703065 The asset manager for a changing world