巴黎银行-全球-外汇策略-2019年欧元兑美元基本面展望-20190220-28页.pdf

Please refer to important information at the end of this report EUR/USD xccy basis 2019 OUTLOOK February 2019 Camille de Courcel,Senior Rates Strategist|BNP Paribas London Branch 2 Key messages Our analysis of various factors converge towards the same conclusion:the EUR/USD xccy basis has room to rise further in 2019:The potential for tighter USD OIS/BOR spreads;The early end to Fed balance sheet normalisation which,relative to the end of ECB QE,supports tighter xccy basis;Our credit analysts view that US credit will outperform EUR in 2019.Relative funding costs are back to neutral.Seasonality analysis points to a widening(more negative)pattern in the two weeks starting 11 February.However,the unusual widening that occurred in January reduces the likelihood that Februarys typical seasonal widening will be seen this year.The following typical seasonal pattern is a tightening(rise)towards the end of March.In contrast to previous years,cross-border investment flows normalised sharply in 2018 and we expect them to continue to normalise in 2019,supporting tighter(higher)xccy basis.The benchmark reform and ISDA fallback methodologies for derivatives could spur a widening at the longer end.Trade ideas:against this backdrop,we maintain our long-term normalisation trade idea.Pay 3y2y EUR/USD xccy basis versus receive 10y20y(entered at 32bp on 5 September 2018;target:15bp;carry:+3.5bp/year;current:29bp).3 01 Sensitivity analysis 4 Summary:Most drivers suggest a further tightening ahead USD OIS/BOR:After falling sharply since the US money market fund reform,the OIS/OIS xccy basis sensitivity to USD OIS/BOR has returned to mean.Arbitrage between domestic USD unsecured funding and secured funding(xccy swap)is back,but we still expect the OIS/OIS basis to remain insensitive to risk-off.Under our untwist base case scenario for the end of the Feds balance sheet normalisation,we see scope for front-end USD OIS/BOR spreads to be supported,potentially even tighten.This,as well as the prospect of declining T-bill issuance in 2019,favours tighter(higher)front-end xccy bases,in our view.ECB excess liquidity and new TLTRO:If the ECB introduces a new TLTRO as we expect,excess liquidity is likely to remain around current levels,suggesting no change in xccy basis.More important is how the new TLTRO would impact credit markets.Early end to Fed balance sheet normalisation:The dovish shift in Fed rhetoric makes us feel more confident with our risk scenario that the end of balance sheet normalisation could take place as soon as this summer,when reserves fall into a USD1.2-1.4trn range.With the end of ECB QE while the Fed slows down its balance sheet normalisation,the resulting relative change in the two central banks balance sheets will shift,supporting a tightening in xccy basis.Credit:Asset markets sensed a problem with quantitative tightening,with credit weakening substantially in late Q4 2018 as a result.We expect the end of QT to be bullish for credit.In our view,US credit performance relative to EUR will remain the main driver of xccy basis this year.Our credit strategy team expects US credit to outperform EUR in 2019 as the Fed shifts to a more dovish tone regarding both rates and the balance sheet while the ECB is more in wait and see mode.5 USD OIS/BOR:Sensitivity returns to mean 1y EUR/USDOIS/OIS xccy(y)vs USDOIS/BOR(x)Source:BNP Paribas(bp)Observed vs fitted 1y EUR/USDOIS/OIS xccy Source:BNP Paribas(1y rolling fit vs.USDOIS/BOR;bp)Xccy sensitivity to USDOIS/BOR:back to mean Source:BNP Paribas(1-year rolling)Arbitrage between USD domestic unsecured funding(USDBOR)and secured funding(xccy swap)Pre-versus post-MMF reform and structural changes in USD-funding patterns=less sensitive to risk-off Our view:USD OIS/BOR skewed tighter with a drop in volatility.We see net T-bill supply falling to USD165bn from USD383bn last year.Under our untwist base case scenario for the end of Fed balance sheet normalisation,we expect MBS prepayments to be reinvested in T-bills from 2020 onwards,with the possibility of an earlier start,possibly in Q2 this year.These factors could favour tighter USD OIS/BOR spreads.Spot 3m Libor/SOFR fallback has averaged 24-26bp over the past five years,arguing for the USDOIS/BOR curve to settle around 26-28bp(under the assumption of a 5-year lookback window)which is where the curve currently is:1y at 26.5bp,5y at 28bp,10y at 29.5bp,and 30y at 30bp.bp bp 6 ECB:Questions on a potential new TLTRO Excess liquidity:the pace matters more Sources:BNP Paribas,Bloomberg From one regime to another Sources:BNP Paribas,Bloomberg Three regimes?Sources:BNP Paribas,Bloomberg EUR/USD xccy basis versus ECB excess liquidity:different regimes?Is EUR 1trn excess liquidity a critical threshold?Central bank responses to credit developments versus the credit response to changes in central bank policy MMF reform and subsequent change in USD funding behaviour Excess liquidity:the pace matters more Our view:No expected impact from a new TLTRO alone;more important will be how a new TLTRO impacts credit markets.If the ECB introduces a new TLTRO as we expect,excess liquidity is likely to remain around current levels,suggesting no change in xccy.In the absence of a new TLTRO,excess liquidity could decline by EUR 200bn we estimate,consistent with a 5bp rise in the basis.7 FOMC:An early end to balance sheet normalisation Xccy vs Fed/ECB relative pace of expansion/reduction Sources:BNP Paribas,Bloomberg Xccy basis vs global central bank balance sheets Sources:BNP Paribas,Bloomberg Xccy basis versus global central bank balance sheets:global balance sheet expansion appears to favour a wider xccy basis,while a contraction seems to favour tighter levels.Relative pace of expansion/reduction matters more:a faster expansion/reduction at the Fed,relative to the ECB,would favour tightening/widening pressure,in our view.Conversely,faster expansion/reduction at the ECB,relative to the Fed,would likely favour widening/tightening pressure.Our view:Basis skewed higher as Fed balance sheet normalisation slows and the ECB pauses We expect an early end to the Feds balance sheet normalisation(see US:After the rain(normalisation),comes the sun(reinvestment):we are more confident with our risk scenario that the Fed will normalise until reserves reach USD1.2-1.4trn(USD1.1trn would imply LCR of 100%)from the current USD1.65trn,meaning an end as soon as this summer.This would imply a balance sheet of around USD3.75trn from current USD4trn.Meanwhile,the ECB will keep reinvesting bonds purchased under QE at full rate and we expect a new TLTRO which,overall,will likely keep its balance sheet unchanged.We therefore expect global liquidity to decline further,but at a slower pace than last year,which could suggest a limited rewidening in the xccy basis(left-hand chart above).But a slower reduction in Fed balance sheet,while the ECB ended QE,rather suggests a slight tightening ahead:the annual change in Fed balance sheet versus the ECBs would rise from-14%to-5%,supporting a 4bp rise in xccy basis(right-hand chart above).R2 30%since 2010 R2 50-60%since 2010 8 Credit:Dovish FOMC-last minute relief for USD credit US credit to outperform EUR Source:BNP Paribas,Bloomberg Barclays Aggregate Xccy versus Italy CDS:lost sensitivity Source:BNP Paribas Xccy versus US/EUR credit spread differential Source:BNP Paribas,Markit Less sensitive to risk-off:a view supported by the latest turmoil in Italy.Idiosyncratic versus systemic risk MMF reform and other financial regulation US credit performance relative to EUR:main driver in normal circumstances.Our view:We expect US credit outperformance to support a tighter EUR/USD xccy basis.Asset markets sensed a problem with quantitative tightening,with credit weakening substantially in late Q4 2018 as a result.We expect the end of QT to be bullish for credit.Our credit strategy team expects US credit to outperform EUR in 2019 as the Fed shifts to a more dovish tone regarding both rates and the balance sheet,while the ECB appears to be in wait and see mode.9 02 Cross-border issuance 10 Summary:Relative funding costs are back to neutral The US Tax Cuts and Jobs Act 2017 led to a significant drop in reverse-Yankee issuance last year.This was one of the key factors behind our call for a tightening(rise)in the xccy basis in 2018.Reverse-Yankee issuance halved from 2016-2017 highs,in sharp contrast to Yankee issuance,which reached record highs in the UK and overall remained around historical highs across jurisdictions.Impact from the tax reform is likely to fade,but we would still expect less reverse-Yankee issuance than in the past.So far in 2019:Reverse-Yankee issuance bounced back in January,while Yankee-issuance declined slightly from January 2018.USD-issuance fell in a number of jurisdictions,particularly in the UK where it fell to a new low.However,Yankee-issuance out of the eurozone remained heavy,setting a second record high after 2017,USD10bn above its past ten year average.That said,we suspect a multi-tranche in USD was not swapped,which would better align with developments in xccy bases seen at the start of the year.If we are correct,Yankee-issuance outside the eurozone would actually be slightly below average.Overall,eurozone/US net foreign supply may have declined slightly from 2018,remaining above average,while versus all non-eurozone entities it may have risen USD5bn from 2018(and from the past ten-year average).But,were we to remove that same multi-tranche deal in USD,net supply would actually have fallen from previous years.This would be more consistent with the exceptional widening in xccy basis that took place in January this year.Relative funding costs have normalised from extreme levels.Levels look much more neutral now for both US and eurozone entities.USD issuance Foreign currency*issuance by US entities 11 Cross-border issuance:Annual stats and focus on Q1 Sources:Bond Radar,BNP Paribas(*G4:EUR,GBP,CHF,CAD)Sources:Bond Radar,BNP Paribas Foreign currency*issuance by jurisdiction USD-supply by eurozone vs EUR-supply by non-eurozone:net supply in Q1 Sources:Bond Radar,BNP Paribas(*USD,EUR,GBP,CHF,CAD)Sources:Bond Radar,BNP Paribas(average over 2005-2018)Trends persist relative to total syndicated supply Jan 2019:Foreign currency issuance drops in the UK 12 Cross-border issuance:January in focus Sources:Bond Radar,BNP Paribas(*USD,EUR,CAD,GBP,CHF)Sources:Bond Radar,BNP Paribas(*USD,EUR,CAD,GBP,CHF)EUR-issuance by US bounces back Yankee issuance fell(vs previous years)in Jan 2019 ex.eurozone Sources:Bond Radar,BNP Paribas(average over 2005-2018 NB:a multi-tranche in USD has been excluded Sources:Bond Radar,BNP Paribas 13 EUR/USD:Seasonality analysis Bi-weekly seasonality:5y EUR/USD basis vs x-border supply Source:BNP Paribas Monthly seasonality analysis Source:BNP Paribas Net supply:USD supply by eurozone entities-EUR supply by non-eurozone entities.Patterns previously seen in Q1:Start of January:the 5y basis has previously tightened as the average net supply has surged after the December lull:often,the first two weeks are among the heaviest of the year(if not the heaviest)in terms of USD-issuance by eurozone entities,unlike EUR-issuance by non-eurozone entities,which tends to spike in May.Middle of February:the basis has historically widened as average net supply reaches a trough,owing to a slowdown in USD,while EUR issuance has picked up.End of March:the basis has previously tightened as the average net supply has reversed(USD supply rises while EUR supply falls).2-weeks starting from2014201520162017Avg chgeAvg 0Occurrence 001-Jan4.2-8.9-5.30.42.03.979%25-Mar2.11.4-0.83.52.02.971%15-Jul0.31.2-2.3-0.10.81.671%23-Sep1.32.94.12.10.82.286%16-Dec3.31.72.34.60.61.479%2-weeks starting from2014201520162017Avg chgeAvg 0Occurrence 011-Feb0.8-0.8-3.81.8-1.4-2.571%21-Oct-1.6-4.3-1.0-1.8-0.4-1.571%04-Nov-3.01.0-0.6-1.2-3.1-4.279%2004-2017 tightening patterns(5y EUR/USD basis,in bp):2004-2017 widening patterns(5y EUR/USD basis,in bp):Source:BNP Paribas 14 EUR/USD:Fails to tighten at the start of the year,wider in late Jan Cross-border issuance on a weekly basis Sources:Bond Radar,BNP Paribas Cross-border issuance on a bi-weekly basis Sources:Bond Radar,BNP Paribas N.B:a multi-tranche in USD has been excluded from January data 15 Relative funding costs:The current position Case studies 3 and 4-EIB and KfW USD credit spreads widened sharply between mid-November and early January,more than EUR spreads,and even more so after accounting for the xccy basis which tightened slightly.As a result,EUR bonds swapped into USD looked potentially attractive towards the end of December/start of January,relative to USD issuance costs.USD credit then tightened significantly into early February:EUR bonds swapped into USD have now lost their appeal relative to domestic USD funding costs,in our view.European SSAs:issuing in USD at the beginning of the year looked unattractive for European SSAs.Towards the end of January,relative levels became more neutral.Sources:Bloomberg,BNP Paribas Case study 1:US entities in 3-5y sector Sources:Bloomberg,BNP Paribas Case study 2:US entities in 7-10y sector Sources:Bloomberg,BNP Paribas Sources:Bloomberg,BNP Paribas N.B:the charts show the cost savings(positive number)when issuing in a foreign currency,swapped to domestic,compared to issuing in domestic 16 A look at credit indices:Same conclusion Hedged US/EUR:back to neutral Sources:BNP Paribas,Bloomberg US and EUR credit weaken into 2019 Sources:BNP Paribas,Bloomberg USD/EUR:back to QE range lows as EUR underperform again Sources:BNP Paribas,Bloomberg EUR and USD credit spreads:back to past 3 years highs EUR:ECB QE tapering in 2018 USD:Fed B/S normalization Global:Trade war,economy slowdown US/EUR credit spread differential:back to the QE range lows EUR credit underperformed USD throughout 2018 and further into 2019.USD-funding terms look historically attractive relative to EUR.Hedged US/EUR spread:back to neutral levels On aggregate,issuing in a foreign currency swapped to domestic looks equivalent to issuing directly in domestic currency.The spread had temporarily spiked early January,making the cost of Yankee issuance briefly surges,relative to domestic.-50-2502550751002550751001251501752002252502013201420152016201720182019USD credit spread(bp)EUR credit spread(bp)US/EUR spd diff(rhs,bp)-120-100-80-60-40-200204060802010201120122013201420152016201720182019US/EUR spd diff(rhs,bp)-40-30-20-10-10 20 30 40201420152016201720182019Hedged spread(bp)17 03 Cross-border flows 18 ECB data:Cross-border flows on a continued normalisation Annual debt flows Sources:ECB,BNP Paribas Equity&debt flows:6-months rolling net flows Sources:ECB,BNP Paribas Debt securities:6-months rolling net flows Equity:6-months rolling net flows Sources:ECB,BNP Paribas Sources:ECB,BNP Paribas bp 19 Global diversification 2-year sovereign bonds Sources:Bloomberg,BNP Paribas 5-year sovereign bonds Sources:Bloomberg,BNP P