巴黎银行-全球-投资策略-美国2s10s：巴甫洛夫式的回应-20190408-15页.pdf

|FOCUS 08/04/2019 1 Rates Buy 6m2y atmf-5/atmf+10/atmf-25 Payer Swaption ladder.Buy 10y TIPS breakeven.Credit Long US corporate credit with a preference for non-financial credit.Equities Buy BNP Paribas Equity Low Volatility Europe ETF(EVOE FP),ref.EUR125.6 Equities Long SX5E June 2019 ATM(3450)call for 1.2%.Alternatively,June 2019 3450-3550(100%-103%)SX5E call spread,for 0.8%(3.6x max return to premium)or a Dec-19 100%-110%SX5E call spread for 2%,contingent on a higher EUR 10-year CMS rate.Premiums indicative.TRADE IDEAS FOCUS|GLOBAL 08 April 2019 US 2s10s A Pavlovian Response KEY MESSAGES Its not about a recession.Its about disinflation.Disinflation without a recession may support risky assets.Sell rates volatility,buy inflation protection.Flatter US 2s10s does not signal tighter US credit conditions.Lower rates term premium may lead to higher P/E in SPX.Lower rates term premium may favour low vol factor in equities.In 1927,Pavlov conducted his famous experiment on classical conditioning.He looked at the association of an unconditional stimulus(UCS)a stimulus that produced an involuntary or unconditional response(UCR),with a neutral stimulus(NS),a stimulus that does not generate an UCR.In his experiment he gave meat powder(UCS)to dogs and observed they salivated(UCR).Independently,he rang a bell(NS)and the dogs did not salivate.When he repeatedly paired the meat powder(UCS)with a bell(NS),the dogs response changed so that they salivated at the sound of a bell even when there was no meat powder present.Pavlov was able to elicit a reflexive response to a previously neutral stimulus.Please refer to important information and MAR disclosures at the end of this report EQUITY&DERIVATIVES|G10 INTEREST RATES|CREDIT|QUANTITATIVE STRATEGY Kaushik Banerjee,Head of Global Macro Research|BNP Paribas London Branch Shahid Ladha,Head of Strategy for G10 Rates Americas|BNP Paribas Securities Corp.Edmund Shing,Global Head of Equity&Derivative Strategy|BNP Paribas London Branch Kris Gjini,Macro Quantitative Strategist|BNP Paribas London Branch Does Pavlovs experiment on classical conditioning hold true for the US 2s10s curve?Is a fall in global inflation expectations along with a yield curve flattening generating a conditioned response of an expected recession?Looking at the US cycle from 1988,we see that every US recession(shaded grey area)and a subsequent fall in inflation,has been preceded with a flattening of 2s10s(shaded yellow area Fig.1).What we observe today is a tepid outlook for growth but a fall in global inflation expectations.The implications are quite different for markets with soft growth and low inflation and those in recession and low inflation.In the first,risky assets perform well,supported by low defaults and interest rates,while and in the latter risky assets perform poorly as default risk trumps interest rate support.Fig.1:US 2s10s and CPI&GDP Sources:Macrobond,BNP Paribas Sources:Bloomberg,Macrobond,BNP Paribas Sources:Bloomberg,Macrobond,BNP Paribas|FOCUS 08/04/2019 2 The three factors which are the most relevant for markets today are PCA1 US 2y5y Inflation expectations,PCA2 US 10y real yields and PCA3 G10 Economic surprises.What we see is that the most important factor as measured by the Z-score of the sensitivity of 2s10s to the PCA factor is inflation expectations.(Fig.4).What does this all mean?1.US 2s10s is responding to inflation and is not a good indicator of recession expectations at the moment.If the 2s10s were to correlate back to data surprises,the yield curve would steepen based on our fair value estimates.2.Risky assets may continue to perform in a low inflation,below trend growth scenario in the USA.3.Can the Fed cut rates?Maybe,if they cared more about symmetric inflation targets.There is no information in the curve pointing to a recession.What state is the market pricing now?We analyzed the sensitivity of US 2s10s to inflation expectations and to BNPPs US Economic Strength Index(ESI).We use the US 2y5y inflation swaps as a proxy for inflation expectations.Our analysis shows 2y5y inflation swap is the better indicator of near term inflation expectations than 5y5y.We observed that in 2019 the US 2s10s has a much stronger correlation to US 2y5y inflation than to the ESI(Fig.2).If we consider the difference in covariances of US 2s10s to US 2y5y inflation and ESI,we get a relative importance chart of each factor(Fig.3).This validates our premise that US inflation expectations are driving US 2s10s more than economic data surprises.Macro factor analysis Another way of looking at the curve is through a macro factor analysis on the US 2s10s.We analyzed 75 assets on a rolling 12-month window and determined the first three principal components.We then correlated these components to a set of macro and market factors(BNPP MarFA model*).It is not about a recession.It is disinflation.EQUITY&DERIVATIVES|G10 INTEREST RATES|CREDIT|QUANTITATIVE STRATEGY Fig.2:Correlation of US 2s10s to inflation and data Fig.3:Drivers of US 2s10s Kaushik Banerjee,Head of Global Macro Research|BNP Paribas London Branch Shahid Ladha,Head of Strategy for G10 Rates Americas|BNP Paribas Securities Corp.Edmund Shing,Global Head of Equity&Derivative Strategy|BNP Paribas London Branch Kris Gjini,Macro Quantitative Strategist|BNP Paribas London Branch PCA 1 PCA 2 PCA 3 Market Variables Inflation Expectations(2y5y)US 10yr Real yields G10 Data Surprises Description Rsq 71%9%12%Rsq of US2s10s to each PCA Shock(1sd)+14bp+15bp+11 =1 standard deviation move in each variable Sensitivity(z-score)0.8 0.2 0.2 Z-score sensitivity of US2s10s to each shock Sensitivity(bp)8bp 3bp 3bp Basis point sensitivity of US2s10s to each shock :Fig.4:Sensitivity of US 2s10s to shocks in principal components Sources:Bloomberg,BNP Paribas *See Introducing MarFA:BNP Paribas Market Factor Analysis for further details.|FOCUS 08/04/2019 3 In terms of curve,we expect 2s10s to steepen because:The next rate move for the Fed is more likely a cut,although the Fed is on hold indefinitely.Fed is expected to add liquidity at the front-end by shifting reinvestments to shorter maturities.US Treasury is considering longer dated issuance to capitalize on domestic demand.The problem with 2s10s steepeners is the negative carry/roll of-2bp/month,in an environment where volatility is expected to stay suppressed.Steepeners expensive we prefer correlated trades Our analysis of previous cycles suggests 2s10s can remain flat after the last hike until delivery of the first rate cut(Fig.7).In the absence of visibility on the timing of any changes to Fed reinvestment or US Treasury issuance pattern,steepeners are an expensive option.We prefer correlated trades.We expect US inflation breakevens to continue behaving like the 2s10s curve benefiting from rate cuts and higher nominal rates.Crucially,TIPS offer strong positive carry over the next 3m:+11bp on long 10y TIPS breakevens(vs.-6bp on 2s10s swap curve steepeners).Trade:Long 10y TIPS breakeven.Entry:182bp.Target:200bp.Stop:170bp.Current:190bp.Carry:+11bp/3m.TR-9449 An inverted curve usually signals recession as the Fed typically takes nominal and real fed funds to a high and restrictive level.Recession usually occurs when real fed funds is higher than(prevailing and potential)real GDP growth,R*(Fig.5).It is different this time Fed has signaled the end of tightening at its lower bound of neutral given limited inflation pressure,lower household credit growth than previous expansions and increased focus on financial conditions(ie,asset prices)see Fig.5.With real fed funds rising more slowly and to a much lower level than historically,even adjusted for a new neutral rate,we do not expect a recession.Fed should continue suppressing rates volatility Rather,we have argued the Feds dovish shift from January will take rates volatility to new structural lows as the Fed has reduced the upside(no more hikes,lower dots)and downside(early dovish pivot)distribution of rates.Given our expectations for a Fed on hold,results of our volatility regression(below)and dynamics of the March rebound in vol,we find the upper left corner offers a good opportunity for being short volatility.Our rates vol rolling regression vs.nominal rate forward,2s10s swap curve,and G3 balance sheet finds 6m2y is most overvalued(Fig.6).Buying a 6m2y-5/+10/+25bp payer ladder:Sells vol in this sector,particularly high strikes,which we prefer.Takes advantage of the roll from fwd(2.35%)to spot 2.45%.Terminal breakeven at 2.75%currently.New trade one may consider:Buying 6m2y atmf 5bp/atmf+10bp/atmf+25bp payer ladder.Entry:6 cents(3bp running).Target:15bp.Stop:-3bp.Carry:+3.25bp/3m.(TR-9537)It is different this time:Sell rates volatility,buy inflation EQUITY&DERIVATIVES|G10 INTEREST RATES|CREDIT|QUANTITATIVE STRATEGY Fig.5:Real Fed funds is much less restrictive this time Fig.7:2s10s between Fed rate cycles Sources:Federal Reserve,Macrobond,BNP Paribas Sources:Marcobond,BNP Paribas Kaushik Banerjee,Head of Global Macro Research|BNP Paribas London Branch Shahid Ladha,Head of Strategy for G10 Rates Americas|BNP Paribas Securities Corp.Edmund Shing,Global Head of Equity&Derivative Strategy|BNP Paribas London Branch Kris Gjini,Macro Quantitative Strategist|BNP Paribas London Branch Fig.6:US rates vol expected to fall further Sources:Federal Reserve,ECB,BoJ,Macrobond,BNP Paribas|FOCUS 08/04/2019 4 However,that relationship has changed in this cycle and in fact a flatter curve has been associated with a rising NIM.We would attribute that to a more balanced asset-liability matching by banks,where the level of interest rates matter more than the shape of the curve itself.This reduces the impact of the yield-curve on credit conditions.As a result we remain positive on US corporate credit,with a preference for non-financials over financials This Credit section has been prepared by Viktor Hjort part of the Credit Trading Desk Analyst team who work closely with the Sales and Trading function.This is not a research report and has not been prepared by the BNP Paribas Research Department.Please see disclaimer.A flatter yield-curve is not a reason to turn bearish on credit markets.There are two reasons a flat yield curve has been a concern for credit markets.It has signalled an imminent recession,with which comes a credit bear market.It is associated with tighter credit conditions,with which comes rising default risks.Neither seems imminent this time.As discussed earlier in this note we dont think the recession signal is valid at the moment.We also do not see tighter credit conditions as imminent.A flatter yield curve has historically been associated with tighter credit conditions.Fig.8 shows BNPPs US Credit Conditions Index which reflects banking system liquidity,lending standards and debt capital market credit availability.Credit conditions lead default rates,and as the chart illustrates,the yield-curve often leads credit conditions.Flatter yield curve,stable credit conditions Recently,however,the two have deviated:the yield-curve is very flat,but credit conditions appear stable.This may be reflecting changes in how the US banking system matches assets and liabilities.Historically,banks would rely meaningfully on short-dated funding to lend further out the curve.This meant that periods where the yield curve steepened,net interest margins(NIM)would improve,boosting profits,and easing lending conditions(Fig.9).Flatter yield curve may not signal tighter credit conditions EQUITY&DERIVATIVES|G10 INTEREST RATES|CREDIT|QUANTITATIVE STRATEGY Fig.8:A flatter yield curve often means tighter credit conditions Fig.9:NIM vs yield curve Sources:Bloomberg,BNP Paribas-1-101122334-2-1-101121990199319961999200220052008201120142017BNPP Credit Conditions Index vs default rates RecessionCredit conditions indexYldCurveTighter credit conditions 3.03.54.04.55.0-1012319831988 19931998 2003 20082013 2018Sources:Bloomberg,BNP Paribas|FOCUS 08/04/2019 5 the average S&P 500 gain of 13%up to subsequent stock market peak,over 15%for the DAX index and over 9%for the UKs FTSE 100 index So if we accept that a flat or inverted yield curve points to lower expected long-term inflation and thus lower nominal GDP growth,then we can surmise that a corporate profit growth slowdown can also be in order as a result.But this,in itself,does not seem a sufficient condition to warrant an immediate equity market sell-off,judging by the evidence below.The complicating issue is this:if we decompose the equity market price level crudely into its two underlying drivers of earnings and valuation(price/earnings),then even when earnings grow more slowly or even decline,we still need to determine what happens to the index P/E in order to determine whether the index rises or falls as a result.Interestingly,at least since the 1990s there has been a leading relationship between the US 10-year term premium(a key driver of the 10-year Treasury yield and thus of the yield curve)on the one hand,and the S&P 500s P/E valuation on the other.The lower the US 10y term premium today,the higher the S&P 500 P/E is likely to be in 30 months.This relationship of equity valuation to term premium can act as at least partial offset to pressure on corporate earnings from lower expected inflation when the yield curve flattens/inverts.Is a negative yield curve necessarily bad for equities?No,in a single word.It is true that a negative US yield curve has often(but not always)predicted a subsequent US recession.However,even when true,the lag between the first occurrence of a negative yield curve and a recession can be long,typically around a year but often longer.Looking at past inversions of the US yield curve,the same would seem to be true of the lag between an inverted yield curve and subsequent equity bear markets.Since the late 1960s,on the seven occasions that the US 2y/10y yield spread has turned negative for at least a month,the S&P 500 index has peaked anywhere between zero and 26 months later,with an average of nine months from curve inversion to stock market peak.Taking the German DAX index,we see a similar pattern,with a lag of zero to 16 months between US yield curve inversion and the DAX index peak.Hence,since the early 1970s,yield curve inversions have been poor in timing stock market downturns.Should investors cut equity exposure when the yield curve inverts?Judging by Fig.10,the answer again seems to be usually no.While there is huge variation in the size of stock market advances post US yield curve inversion,note What does a negative yield curve mean for equities?EQUITY&DERIVATIVES|G10 INTEREST RATES|CREDIT|QUANTITATIVE STRATEGY Fig.10:Substantial lags from inversion to equity market peaks seen post early 1970s Fig.11:Exiting the stock market once the US yield curve inverts can be expensive Sources:Bloomberg,BNP Paribas Sources:Bloomberg,BNP Paribas Kaushik Banerjee,Head of Global Macro Research|BNP Paribas London Branch Shahid Ladha,Head of Strategy for G10 R