EFTA01405764.pdf

Deutsche Bank Markets Research United States Economics Rates Credit US Fixed Income Weekly IIMarkets are fixated on the potential for Fed normalization to start earlier than currently priced and whether China's recent FX adjustment is the beginning or the end. IIAt a superficial level there appears to be conflicting influences on rates. The Fed and China may undermine risk asset performance but the consensus is that if risk assets find support, fewer FX reserves are likely to pressure rates higher. IIOn the contrary, we think the most important thing is that both the Fed and China's FX (ongoing?) unwind represent a tightening of global liquidity that clearly is negative for risk assets and clearly, at least for the last decade, has been positive for real rates and the curve. 5y5y is well correlated with changes in global liquidity and based on recent trends should be closer to 2 percent. IIThis reinforces our view that the Fed is in danger of committing policy error. Not because one and done is a non issue but because the market will initially struggle to price "done" after "one". And the Fed's communication skills hardly lend themselves to over achievement. More likely in our view, is that one in September will lead to a December pricing and additional hikes in 2016, suggesting 2s could easily trade to 1 14 percent. This may well be an overshoot but it could imply another leg lower for risk assets and a sharp reflattening of the yield curve. Decline in liquidity implies a lower 5y5y 10 15 20 25 30 -10 -5 0 5 20001 Source: Fed and Deutsche Bank 20061 20121 Fed plus fx reserves yoy 5y5y rhs 0.0 1.0 EFTA01405764 2.0 3.0 4.0 5.0 6.0 7.0 Date 4 September 2015 Dominic Konstam Sfl Aleksandar Kocic Joseph LaVorgna Economist Alex Li Research Anal st tuart par s Research Anal st Daniel So rid teven eng, Aditya Bhave Economist Table of Content US Overview US Credit Strategy Chart Pack Page 06 Page 23 Page 28 Deutsche Bank Securities Inc. DISCLOSURES AND ANALYST CERTIFICATIONS ARE LOCATED IN APPENDIX 1. MCI (P) EFTA01405765 124/04/2015. EFTA01405766 US Fixed Income Weekly 4 September 2015 Page 2 Deutsche Bank Securities Inc. 2015 Outlook Recommendations Trade Detail Rationale Option Buy lxl, lyly receiver spreads with strikes ATMF and ATMS The post-Fed sell-off has left the spot/forward spread near multi-year post-crisis highs. Swaps RV Pay 3y1y versus 2y1y Option Option Option Option Option Option Source: Deutsche Bank Sell 1X2 payer spreads at the short end: Sell $100mn 6M3Y ATMF vs. buy $200mn 34.5bp OTM payers at zero net cost Sell $100mn 6M10Y straddles vs. buy $300mn 6M3Y straddles for a net premium of 175K Quiet flatteners: sell $1bn 6M 5s/lOs 9.5bp OTM curve cap vs. buy$1bn 6M 5s/lOs atmf/9.5 curve floor spread at zero cost Quiet bulls: Sell $100mn 1Y10Y 50bp OTM payers vs. buy $100mn 1Y10Y ATMF/33 receiver spreads costless Buy $100mn 1Y30Y receivers, struck at spot, at 1270c 6M dual digital: 2s> F+10bp & lOs < F-10bp offer 11.5% This curve segment might be expected to steepen if, for example, higher inflation produces greater pricing power, or if the long-absent cyclical increase in productivity finally materializes. The repricing of Fed hikes could begin in Q2 with the short end rebounding sharply after initial rally. With expectations of Fed hikes, volatility should move to the front end of the curve, while the back end movements remains Potential for considerable bear flattening should EFTA01405767 the market reprice the Fed hikes. This captures the risk of bullish flattening of the curve where growth is unable to take off either due to fundamental weakness or in response to a policy mistake of premature hikes. Bull/flatteners at the back end. This is a leveraged expression of a policy-mistake trade where premature hikes cause a rally at the back end. Risks Maximum total loss is the premium outlay Opened 12/19/14 Entry 29c Current P/L Curve flattens 12/19/14 Vulnerable to rally below the breakevens, with potentially unlimited downside. Unilateral spike in backend vol. Curve steepens. Sell-off beyond 3.10%. Loss equal to the options premium Loss equal to the options premium +40 bp 12/19/14 12/19/14 12/19/14 12/19/14 12/19/14 12/19/14 EFTA01405768 US Fixed Income Weekly 4 September 2015 Deutsche Bank Securities Inc. Page 3 2015 Outlook Recommendations Trade Detail Rationale Treasury RV Inflation Swaps Inflation Inflation Agencies Agencies Sell rich bond futures against cheap off-the-run bonds The classic bond futures look rich in the long end Risks Further outperformance of the 6.25s of 5/2030 in the long end Buy 2yr2yr forward breakevens The 2yr2yr inflation appears attractive on a longterm history Buy long end inflation Buy 5yr5yr forward breakevens as a hedge to high rates Buy 3ncly and 5nc6m callables vs. matched-maturity bullets 2-year vs. 5-year agency spread curve flattener The long end inflation market looks undervalued on a long-term perspective, with the 30-year TIPS breakevens trading below 2.00%. The 5yr5yr forward breakevens have dropped to their multi-year lows. With the Fed moving closer to its first rate hike in a low-inflation, moderate-growth environment, there are few themes as sure as the flattening of the curve, likely going beyond the forwards. On the bullet agency curve, spreads are relatively tight to the level of rates volatility, and they risk widening 5-10bp from current levels on our model incorporating forward vols and the projected level of outstanding debt. US Credit US High Yield: Sell covered puts on HY CDX With CCC energy bonds trading at 60 cents on the dollar, and oil just $10 away from matching the EFTA01405769 most severe percentage drop in oil prices over 1997-8, our sense is that we may be reaching the latter stages of a pronounced move lower in a commodities-driven decline in HY credit valuations Source: Deutsche Bank Further decline in medium-term inflation expectations Inflation markets further underperform. Decline in energy prices and a stronger dollar Higher implied vol cheapens callables relative to bullets Increased GSE risk widens intermediate spreads Widening of credit spreads beyond the breakeven point as well as a rally in credit beyond the breakeven, with potentially unlimited downside in either scenario Opened 12/19/14 Entry Current +5 bp +21 bp (Closed on 2/25) 12/19/14 1.95% 1.60% P/L +1,249k -1,367k 12/19/14 12/19/14 12/19/14 1.92% 2.18% 1.71% 1.97% -3,400k -648k 12/19/14 12/19/14 EFTA01405770 US Fixed Income Weekly 4 September 2015 Page 4 Deutsche Bank Securities Inc. Other Current Recommendations Trade Detail Rationale Treasury RV Treasury RV Short lOs versus 5s and 30s Sell rich bond futures against cheap off-the-run bonds Inflation 10s/30s breakeven curve steepener Inflation Long front end TIPS breakevens Inflation Inflation Inflation Real yield curve steepeners, either 10s-30s or 5s-30s. Long 10yr inflation swaps versus 10yr TIPS breakevens Long 1/2029 breakevens vs 10yr breakevens Inflation Long 30yr TIPS breakevens Inflation Swaps Inflation Swaps Agencies Muni Option Source: Deutsche Bank Long lyrlyr inflation swaps Long 2yr2yr inflation swaps Buy long-dated GSE debt: Buy $100mm FNMA 6.625 11/30s vs. T 5.325 2/31s Receive $100m 3y3y SIFMA ratio at 78.2%. (Sorid) 1X2 1Y 5Y5Y ATMF/41 receiver spreads costless lOs look rich on the curve against 5s and 30s Sell the rich classic bond futures versus off-the-run bonds in the 2026 to 2028 sector Long end TIPS offer good value Front end TIPS look cheap to our inflation forecast EFTA01405771 Possibly delayed first Fed rate hike is likely to help intermediate sector outperform in real yields, steepening the real yield curve. The spread between 10yr inflation swaps and TIPS breakevens is too tight 10yr TIPS to 1/2029 breakeven curve is too flat The long end inflation market looks undervalued; 30yr TIPS breakevens near multi-year lows We like lyrlyr forward inflation swaps. Front end breakevens look attractive. We like being long 2yr2yr or 2yr3yr forward breakevens to take advantage of cheap 5s, while avoiding negative carry in front end TIPS Legislative momentum of JohnsonCrapo on GSE reform is credit bullish for long-dated GSE debt. Attractive roll down profile Long-end rallies on premature or fast rate hikes (policy mistake) Risks lOs richen further Classic bond futures richen 30yr underperforms relative to 10yr Energy prices drop Opened 5/8/15 11/26/14 6/26/2015 4/10/2015 Long end outperforms 1/20/2015 TIPS outperform inflation swaps 1/2029 breakeven cheapen further Long term inflation expectations decline Inflation expectations decline Medium term inflation expectations decline Reform bill stalls in Congress or language on government EFTA01405772 guarantee modified. Further ratio curve steepening Rally below the breakevens; unlimited downside 1/20/2015 10/3/14 12/12/14 3/3/15 12/12/14 Entry +9 bp +21 bp 0.13% 1.23% 5s/[email protected]% 10s/[email protected]% +21 bp +2 bp 1.91% 1.84% 1.77% Current +8 bp +20 bp 0.30% -1.45% 5s/[email protected]% P/L -6k -106k +1,042k -1,563k 105/[email protected]% +3,464k +17 bp +6 bp 1.71% 1.22% 1.68% -249k +502k -2,107k -662k -868k 3/14/14 4/25/13 9/26/14 +48 bp +62 bp 78.2% EFTA01405773 Oct 72.0% —18.44 -953k +941k -311k EFTA01405774 US Fixed Income Weekly 4 September 2015 Deutsche Bank Securities Inc. Page 5 Other Current Recommendations Cont'd Trade Detail Rationale Option Option Swaps Rv Swaps Rv Swaps Rv Cross Market Cross Market Buy $100mn 2Y2Y ATMF receivers vs. sell $22 7mn 2Y10Y ATMF receivers for the net takeout of $55K Payer spreads: Sell $500mn 2Y2Y 92bp OTM payers vs. buy $50mn 2Y30Y 25bp OTM payers at zero net cost Receive $1,023.4mm 2yly rate versus pay $1,002.7mm lyly rate Receive $1,023.4mm 2yly rate versus pay $431.2mm lyly rate and $597mm 3yly rate Forward fly: Pay fixed on $298.6 mm 10y5y versus receive fixed on $72.9 mm 5y5y and $257.6 mm 15y5y Buy $10m each of SPNTAB 2.95% 3/16; SPABOL 2.625% 5/16; DNBNOR 2.90% 3/16 on ASW. (Sorid) US-Europe spread tightener: Receive fixed in $244 mm USD 5y5y rate vs. pay fixed on €165.8mm EUR 5y5y rate Trend growth and low inflation limit the rise of long rates Vol differential is favorable for initiating a positive carry bear steepening trade Positive carry look at repricing Fed Further rally via Fed delay benefits 2yly rate 5y rate, lOy forward is historically rich versus 5y rate, 5y forward and 5y rate, 15yfoward Risk-on retightening of covered bonds in stable rates regime EFTA01405775 US recovery disappoints Risks Recessionary mode with bull flattening of forwards The curve bear flattens The curve bear steepens 2yly underperformance Further 10y5y outperformance Bank credit underperforms; Eurozone credit crunch; Widening in a rate sell-off Spread widens Opened Entry Current -6 bp 10/3/13 1/2/14 5/20/14 5/20/14 4/29/14 7/25/13 1/24/14 +2 bp +95 bp -10 bp +22 bp +25 bp +37 bp +31 bp +127 bp -99 bp -0 bp +95 bp -17 bp +21 bp +30 bp +25 bp +31 bp +136 bp P/L -925k -25k +2,305k +405k -416k -930k -10k P/L as of 09/03/2015 prices. We started tracking the performance of our trade recommendations on June 18, 2010. This table shows our current open recommendations; a table of our closed positions is in the back of this publication. Both tables will be a EFTA01405776 regular feature in the Weekly. Performance numbers are based on trader end -of-day marks, and do not include bid/offer spreads or transaction costs. We consider the relevant benchmark for our trades to be a zero position, given the leveraged or generally market neutral aspects of these trades. Historical performance is not a guarantee of future performance Source: Deutsche Bank EFTA01405777 4 September 2015 US Fixed Income Weekly United States Rates Gov. Bonds & Swaps Rates Volatility US Overview IIMarkets are fixated on the potential for Fed normalization to start earlier than currently priced and whether China's recent FX adjustment is the beginning or the end. IIAt a superficial level there appears to be conflicting influences on rates. The Fed and China may undermine risk asset performance but the consensus is that if risk assets find support, fewer FX reserves are likely to pressure rates higher. IIOn the contrary, we think the most important thing is that both the Fed and China's FX (ongoing?) unwind represent a tightening of global liquidity that clearly is negative for risk assets and clearly, at least for the last decade, has been positive for real rates and the curve. 5y5y is well correlated with changes in global liquidity and based on recent trends should be closer to 2 percent. IIThis reinforces our view that the Fed is in danger of committing policy error. Not because one and done is a non issue but because the market will initially struggle to price "done" after "one". And the Fed's communication skills hardly lend themselves to over achievement. More likely in our view, is that one in September will lead to a December pricing and additional hikes in 2016, suggesting 2s could easily trade to 1 14 percent. This may well be an overshoot but it could imply another leg lower for risk assets and a sharp reflattening of the yield curve. IIWe think risk/reward has shifted toward paying spreads in the front end. Financing is challenging with term GC trading high relative to LIBOR, but we think rolling the position overnight should allow investors to average in financing better than LIBOR, providing some backstop against tightening if significant additional intervention-related selling does not materialize. IIWe like being long front end breakevens in forwards, e.g., one-year breakevens implied by short maturity TIPS, such as the 7/2016s and the 7/2017s. One can also hedge out energy prices in that trade to create a synthetic exposure to core CPI. A simpler version of the implied front end forward breakevens is to be long front end breakevens outright. They have lagged oil prices. II5-year inflation basis has recovered, while 30-year inflation basis has done less well, and remains in the low end of the long term trading range. Investors should consider inflation basis steepeners by being long 30-year inflation basis against 5-year inflation basis. EFTA01405778 The case for more liquidity Investors are rightly concerned about the impact of both a possible early start to Fed normalization and the probably yet-to-be-resolved Chinese FX adjustment. There is a reasonable consensus that both encourage further downside to risk assets. There is more uncertainty around bond yields. Potential FX intervention might imply selling of Treasuries, especially the front end where most reserves are held. But if higher short rates from either those sales or Fed tightening, undermine equities, bond yields might actually fall. Page 6 Deutsche Bank Securities Inc. Dominic Konstam Research Analyst (+1) 212 250-9753 [email protected] Aleksandar Kocic Research Analyst (+1) 212 250-0376 [email protected] Alex Li Research Analyst (+1) 212 250-5483 [email protected] Stuart Sparks Research Analyst (+1) 212 250-0332 [email protected] Daniel Sorid Research Analyst (+1) 212 250-1407 [email protected] Steven Zeng, CFA Research Analyst (+1) 212 250-9373 [email protected] Aditya Bhave Economist (+1) 212 250-0584 [email protected] EFTA01405779 4 September 2015 US Fixed Income Weekly The right framework to view potential Fed tightening as well as China's FX adjustment is in the context of global liquidity and that relationship with financial assets. Liquidity in the broadest sense tends to support growth momentum, particularly when it is in excess of current nominal growth. Positive changes in liquidity should therefore be equity bullish and bond price negative. Central bank liquidity is a large part of broad liquidity and, subject to bank multipliers, the same holds true. Both Fed tightening and China's FX adjustment imply a tightening of liquidity conditions that, all else equal, implies a loss in output momentum. Typically this should be associated with lower yields. This runs counter to a common perception that forex intervention that leads to Treasury sales pushes up yields. To the extent that it does, we suspect this is a short lived temporary affair and will easily be dominated by the more sinister implications of dwindling global liquidity. We note that the recent weakness in global nominal growth that we highlighted last week is highly consistent with weaker global liquidity and that the weakening in liquidity is not new news but has been ongoing since late last year. Not only has it been driven by falling FX reserves but also by the slowing of the Fed's balance sheet. To the extent that other central banks have tried to expand liquidity, in terms of historic relationships to financial assets, FX reserves and the Fed's balance sheet are more important. We think this reflects the role of the dollar as the reserve currency in the global financial system. Let's start from some basics. Global liquidity can be thought of as the sum of all central banks' balance sheets (liabilities side) expressed in dollar terms. We then have the case of completely flexible exchange rates versus one of fixed exchange rates. In the event that one central bank, say the Fed, is expanding its balance sheet, they will add to global liquidity directly. If exchange rates are flexible this will also mean the dollar tends to weaken so that the value of other central banks' liabilities in the global system goes up in dollar terms. Dollar weakness thus might contribute to a higher dollar price for dollar denominated global commodities, as an example. If exchange rates are pegged then to achieve that peg other central banks will need to expand their own balance sheets and take on dollar FX reserves on the asset side. Global liquidity is therefore increased initially by the Fed but, secondly, by further liability expansion, by the other central banks. Depending on the sensitivity of exchange rates to relative balance sheet adjustments, it is not an a priori case that the same balance sheet expansion by the Fed leads to greater or less global liquidity expansion under either exchange rate regime. Hence the mere existence of a massive build up in FX reserves shouldn't be viewed as a massive expansion of global liquidity per se — although as we shall show later, the empirical EFTA01405780 observation is that this is a more powerful force for the "impact" of changes in global liquidity on financial assets. The chart below shows the RMB vs. the ratio of PBOC to Fed balance sheets, using prevailing exchange rates at the time as the conversion factor. The initial post crisis period sees the Fed balance sheet expand relatively while the exchange rate is unchanged. There is then a phase of RMB appreciation and relative stability in the balance sheet ratio and then the PBOC balance sheet expands with continued RMB appreciation. Deutsche Bank Securities Inc. Page 7 EFTA01405781 4 September 2015 US Fixed Income Weekly RMB vs. ratio of Fed to PBOC balance sheet 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1 20084 Source: Bloomberg and Deutsche Bank The table below highlights these three periods in terms of the actual notional impact on global liquidity via the combined effects of revaluing the PBOC balance sheet as well as the changes in the underlying domestic liquidity. Under a relatively stable currency the PBOC expanded its balance sheet aggressively in the first phase, presumably in part being obliged to accelerated FX reserve accumulation; the Fed was more or less in between expanding their balance sheet. The second phase saw the more dramatic currency appreciation with a strong Fed expansion but also strong PBOC liquidity expansion. The third phase saw even stronger Fed balance sheet expansion but weaker PBOC expansion and more modest RMB appreciation. The last two phases combined saw global central bank liquidity expand by notionally similar amounts i.e. $1500 billion. More than double the first phase when the currency was more stable and the Fed was quieter. However note that as expected, the reserve accumulation was almost the same in each period, around 500-600+ billion. So even though the Fed wasn't expanding the balance sheet much, the hangover of the previous expansion and capital flows in general required a more aggressive intervention by PBOC to acquire reserves and maintain the a stable currency. So a notionally less aggressive expansion in global central bank liquidity under a stable exchange rate regime was disproportionately more skewed to reserve accumulation. Changes in central bank balance sheet liquidity chg Fed BS chg 2010q2-2008q4 2012q1-2010q3 2013q4-2012q4 4.1% 25.8% EFTA01405782 39.3% $ bn 90 581 1126 Source: Haver Analytics and Deutsche Bank The next issue is given changes to liquidity how does it impact asset prices. We can think of the three components of liquidity: the Fed's balance sheet, the accumulation of FX reserves by other central banks; and the residual of other central banks' liquidity expansion after the accumulation of FX reserves. As the chart below shows in terms of growth the explosion of the Fed stands out during the crisis but there have been strong expansions in other central banks' liquidity excluding reserve increases. FX reserve accumulation has been quite weak since 2012 and is now negative. In absolute terms liquidity is strongest in FX reserves and other central banks ex reserves by a factor of three times for the Fed's balance sheet. Page 8 Deutsche Bank Securities Inc. chg Ch BS RMB change RMB bn 17.3% 14.2% 7.7% 3584 3534 2274 start 6.84 6.77 6.24 20114 20144 ratio of balance sheets RMB/$ rhs 6.0 6.1 6.2 6.3 6.4 6.5 6.6 6.7 6.8 6.9 RMB % chg RMBUSD change Total Reserve chg EFTA01405783 finish bn 6.82 6.31 6.09 0.23% 7.26% 2.51% 532 827 492 622 1408 1618 bn 508 657 510 EFTA01405784 4 September 2015 US Fixed Income Weekly Sources of central bank liquidity — change yoy Sources of central bank liquidity — $ billion 100 120 140 160 180 20 40 60 80 -20 0 20031 Source: Bloomberg and Deutsche Bank 20081 20131 FX reserves Fed Other CBks ex Reserves 2000 4000 6000 8000 10000 12000 14000 16000 0 20031 Source: Bloomberg and Deutsche Bank Let's start with risk assets, proxied by global equity prices. It would appear at first glance that the correlation is negative in that when central bank liquidity is expanding, equities are falling and vice versa. Of course this likely suggests a policy response in that central banks are typically "late" so that they react once equities are falling and then equities tend to recover. If we shift liquidity forward 6 quarters we can see that the market "leads" anticipated" additional liquidity by something similar. This is very worrying now in that it suggests that equity price appreciation could decelerate easily to -20 or even 40 percent based on near zero central bank liquidity, assuming similar multipliers to the post crisis period. From q2 levels that implies an MSCI level of around 1350 for EFTA01405785 2015q4 (reference q2 @ 1735), the end August level was 1645 i.e. still another 10-15 percent decline. World equities yoy vs. central bank liquidity yoy 10 20 30 40 50 -50 -40 -30 -20 -10 0 20041 Source: Bloomberg and Deutsche Bank 20101 WORLD EQUITIES YOY Fed plus fx reserves plus other cbs (ex fx) yoy rhs 10 15 20 25 30 35 40 0 5 World equities yoy vs. components of liquidity yoy -50 -30 -10 10 30 50 20031 Source: Bloomberg and Deutsche Bank Interestingly, the components of liquidity themselves behave a little differently with FX reserves and Fed balance sheet being more in line recently than other central bank liquidity. This reflects the ECB and BoJ tardier reactions to balance sheet expansion in the post crisis period. If we only consider the FX and Fed components of liquidity there appears to be a tighter and more contemporaneous relationship with equity prices. The suggestion is at one Deutsche Bank Securities Inc. Page 9 20081 20131 FX reserves Fed EFTA01405786 Other CBks ex Reserves FX reserves Fed Other CBks ex Reserves world equities yoy 20081 20131 EFTA01405787 4 September 2015 US Fixed Income Weekly level still the same, absent Fed and FX reserve expansion, equity prices look more likely to decelerate and quite sharply. The tie out, presumably with the "leading" indicator of other central bank action is that other central banks have been instrumental in supporting equities in the past. The largest of course being the ECB and BoJ. If the Fed isn't going doing its job, it is good to know someone is willing to do the job for them, albeit there is a "lag" before they appreciate the extent of someone else's policy "failure". And just to ram home the point — this differential relationship is entirely consistent with the idea that FX reserves are accumulate don the back of Fed balance sheet expansion and so if the Fed's balance sheet is not expanding then it is a double whammy that FX reserves are also not expanding and as we shall see below are contracting! World equities yoy lead by 6 qtrs vs. central bank liquidity yoy 10 20 30 40 50 -50 -40 -30 -20 -10 0 20041 Source: Bloomberg and Deutsche Bank 20101 WORLD EQUITIES YOY Fed plus fx reserves plus other cbs (ex fx) yoy rhs 10 15 20 25 30 35 40 0 5 World equities yoy vs. Fed plus FX reserves change yoy 10 20 30 40 EFTA01405788 50 -50 -40 -30 -20 -10 0 20001 Source: Bloomberg and Deutsche Bank So now let's be a little more specific on the Fed balance sheet and FX reserves now. The next chart shows both are decelerating sharply. The Fed's balance sheet is almost flat on the year and reserves are down around 5 percent and counting. The two as we have demonstrated are clearly connected. In the reverse scenario (as opposed to the above, when we demonstrated the connection when the Fed was expanding its balance sheet), tighter Fed policy forces other central banks to spend reserves to defend their currency peg and in principle shrink their balance sheets. This is the example recently with the adjustment in China's FX regime to accommodate more market based fixings. The ensuing unwind of the China carry trade has solicited what appears to have been significant FX intervention, judging by the move in front end swap spreads and dealer inventory of shorter dated Treasuries. The main point however is that it is not a change in FX regime per se that drives the loss of liquidity but that that change emanates from a tighter Fed balance sheet. Hence in the event that the Fed raises rates and we start to worry about balance sheet unwind this becomes a much more significant issue going forward. The Fed's balance sheet for example could easily be negative 5 percent this time next year, depending on how they manage the SOMA portfolio and would be associated with further FX reserve loss unless countries, including China allowed for a much weaker currency. This would be a great concern for global (central bank liquidity) So one counter is that FX reserve loss can be offset by other central banks' liquidity injection. At one level this is tempting but flawed; at another level it is more plausible. The first level is that FX reserve loss typically is "sterilized". The shock to a country's financial system from the sudden loss of liquidity Page 10 Deutsche Bank Securities Inc. 20061 20121 WORLD EQUITIES YOY Fed plus fx reserves yoy 10 15 20 25 30 EFTA01405789 -5 0 5 EFTA01405790 4 September 2015 US Fixed Income Weekly needs to be offset and recently in the case of China the PBOC has acted to reinstate domestic liquidity and also has cut reserve requirements. However as we demonstrated above this component of liquidity seems to have a lagged impact on say (equity) financial assets relative to either the Fed or FX reserves themselves This is actually quite intuitive. The liquidation of FX reserve holdings reflects forced redemptions of domestic currency holdings. Simply forcing currency back into the system to satisfy those redemptions shouldn't be associated with restoring asset prices to where they were before. Ultimately in a fiat money system asset prices reflect "outside" i.e. central bank money and the extent to which it multiplied through the banking system. The loss of reserves represents not just a direct loss of outside money but also a reduction in the multiplier. There should be no expectation that the multiplier is quickly restored through offsetting central bank operations. PBOC injection of funds vs. CNY 200 400 600 -600 -400 -200 0 Jan-14 Jul-14 Source: Bloomberg Finance LP and Deutsche Bank We now move on to interest rates. If equities have a negative correlation with liquidity, it is not surprising to find that interest rates have a positive correlation at least since the crisis. Again in line with the above analysis regarding equities, the correlation in contemporaneous time is better if we focus on Fed and FX reserves. However even then we notice the correlation is a little loose at times. This raises an obvious issue in terms of how one thinks about nominal yields in terms of additional central bank liquidity and FX reserve accumulation. On the one hand the more Fed may help lower real yields but raise inflation expectations; more FX reserve accumulation may be just lower nominal yields and if anything real yields to the extent that by accommodating Fed monetary policy expansion the US "exports" inflation risk. Running across everything is the problem that equities are generally stronger (weaker) of liquidity is expanding (falling). Jan-15 Jul-15 10-day total of net injection of funds by PBoC (Bn Yuan) USDCNY (rhs) 6.00 6.05 6.10 6.15 EFTA01405791 6.20 6.25 6.30 6.35 6.40 6.45 Deutsche Bank Securities Inc. Page 11 EFTA01405792 4 September 2015 US Fixed Income Weekly 10 yr yield vs. FX/Fed defined liquidity 10 yr yield vs. broader defined liquidity 10 15 20 25 30 -10 -5 0 5 20001 Fed plus fx reserves yoy 20061 Source: Bloomberg Finance LP and Deutsche Bank lOy rhs 20121 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 10 15 20 25 30 35 40 -10 -5 0 5 20001 Source: Fed and Deutsche Bank Breaking down the breakeven and real yield components verifies that central bank liquidity has been more associated with real yields then breakevens, however the relationship is perverse! Real yields have tended to fall when balance sheet expansion is slowing while breakevens have generally been more sticky. This suggests that risk assets drive (real) yields and that breakevens anticipate a (delayed) liquidity injection. This is corroborated by also considering the curve. Like real yields 5slOs is well correlated (positively) with real yields. Note that prior to the crisis the relationship looked more "normal" in that EFTA01405793 expanding liquidity drive yields lower and vice versa. So something has changed since the crisis—this we think is very important and again, will revisit below. Liquidity vs. 10 yr real yields 10 20 30 40 50 -10 0 20001 FX + Fed balance sheet yoy 10 yr real rhs 20061 Source: Bloomberg Finance LP and Deutsche Bank 20121 -0.8 -0.3 0.3 0.8 1.3 1.8 2.3 2.8 3.3 Liquidity vs. 10 yr breakevens 10 20 30 40 50 -10 0 20001 Source: Fed and Deutsche Bank The relationship between 5slOs and lOs in real terms screams 5y5y! And indeed we overlay 5y5y to liquidity there is a very tight, almost scary, relationship. The relationship even predates the crisis. Tighter liquidity essentially forces the 5y5y nominal rate lower reflecting some combination of a flatter curve and higher yields with a steeper curve and lower yields. Fundamentally we think this ultimately speaks to a lower terminal policy rate so that it doesn't really matter whether the term structure is trying to shift higher or lower but the curve will more than compensate so that if the trend is towards less central bank liquidity, the terminal rate is falling. Page 12 Deutsche Bank Securities Inc. FX + Fed balance sheet yoy EFTA01405794 10 yr bei rhs 20061 20121 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 Fed plus fx reserves plus other cbk (ex fx) yoy lOy rhs 20061 20121 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 EFTA01405795 4 September 2015 US Fixed Income Weekly Right now the decline is central bank liquidity suggest 5y5y should be closer to 2 percent or below not 3 percent to above. And this is before the Fed has tightened and China has potentially "finished" its adjustment. Liquidity vs. 5slOs 10 15 20 25 30 -10 -5 0 5 20001 20061 Source: Bloomberg Finance LP and Deutsche Bank 20121 Fed plus fx reserves yoy 5s10s rhs -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 Liquidity vs. 5y5y 10 15 20 25 30 -10 -5 0 5 20001 Source: Fed and Deutsche Bank And of course the breakdown in 5y5y between real and inflation reinforces the story that it is the real rate not inflation expectations that drive this result. And this is again consistent with the risk asset concern that it is the lack of liquidity that undermines risk assets that in turn drives real yields lower, despite EFTA01405796 keeping breakevens relatively inflated. One conclusion is that if investors believe that liquidity is likely to continue to fall one should not sell real yields but buy them and be more worried about risk assets than anything else. This flies in the face of recent concerns that China's potential liquidation of Treasuries for FX intervention is a Treasury negative and should drive real yields higher. It is possible that if risk assets do very well then maybe the correlation with interest rates is broken. But like all these relationships for us, it is easier to work with the correlations that currently persist rather than to predict random breaks. And the potential breaks should be more cheaply hedged rather than making for a core portfolio allocation. I.e. cheap SPX calls based on rates lower. More generally the simple point is that falling reserves should be the least of worries for rates — as they have so far proven to be since late 2014 and instead, rates need to focus more on risk assets. Liquidity vs. 5y5y real 10 15 20 25 30 -10 -5 0 5 20001 Fed plus fx reserves yoy 5y5y real rhs 20061 Source: Bloomberg Finance LP and Deutsche Bank 20121 -0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 Liquidity vs. 5y5y BEI 10 15 20 25 30 -10 -5 EFTA01405797 0 5 20001 Source: Fed and Deutsche Bank Fed plus fx reserves yoy 5y5y bei rhs 20061 20121 0.0 0.5 1.0 1.5 2.0 2.5 3.0 20061 20121 Fed plus fx reserves yoy 5y5y rhs 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 Deutsche Bank Securities Inc. Page 13 EFTA01405798 4 September 2015 US Fixed Income Weekly Even without considering the empirical relationships, it is also clear that FX intervention is very much a short term affair. As the chart below shows the recent jump in dealer positions in less than three years is consistent with the Treasury data for 2014 that shows the preponderance of foreign official Treasury holdings is held in the sub 3 year sector. Very little is held in longer dated maturities so any FX intervention is anyway more likely to flatten the yield curve than steepen it. About 56% foreign official holdings of Treasuries are under three years in maturity 0% 10% 20% 30% 40% 50% 60% 0-3y Source: Treasury and Deutsche Bank Distribution of Maturities in Treasuries Held by Foreign Investors Official institutions Private investors Dealer positions in Treasuries maturing in 3 years or less 10,000 20,000 30,000 -20,000 -10,000 0 3-5y 5-10y 10y+ Jan-14 Source: Fed and Deutsche Bank The relationship between central bank liquidity and the byproduct of FX reserve accumulation is clearly central to risk asset performance and therefore interest rates. The simplistic error is to assume that all assets are treated equally. They are not — or at least have not been especially since the crisis. If liquidity weakens and risk assets trade badly, rates are most likely to rally not sell off. It doesn't matter how many Treasury bills are redeemed or USD cash is liquidated from foreign central bank assets, US rates are more likely to fall than rise especially further out the curve. In some ways this really shouldn't be that hard to appreciate. After all central bank liquidity EFTA01405799 drives broader measures of liquidity that also drives, with a lag, economic activity. The indicators of excess liquidity (see below) are but derivatives of central bank liquidity and the bank or "inside" money multipliers. If liquidity is tightening relatively to nominal growth, real growth will tend to slowdown later. Right now the message is not good for the OECD, excess liquidity indicators point to real growth losing momentum. The IMF seems to get the picture. China is probably getting the picture but faces the conundrum of how to manage the carry trade unwind with minimal disruption. The grass is definitely though greener if the currency is weaker and they hang onto most of their reserves. Ironically the excess liquidity indicator has recently improved for China although this is as much to do with decelerating nominal growth. Jul-14 Jan-15 Jul-15 Dealer Net Outright Position: Govt Coupon Securities,Due 3Yr or Less(EOP,Mil$) Page 14 Deutsche Bank Securities Inc. $ millions EFTA01405800 4 September 2015 US Fixed Income Weekly Excess liquidity indicator vs. output: OECD Excess liquidity indicator vs. output: China -30% -20% -10% 0% 10% 20% 30% Global excess liquidity yoy +12m lead OECD output momentum OECD -20% -10% 0% 10% 20% 30% 40% 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 Source: Bloomberg and Deutsche Bank China excess liquidity yoy +12m lead China output momentum China 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 Source: Bloomberg and Deutsche Bank The more sinister undercurrent is that as the relationship between negative rates has tightened with weaker liquidity since the crisis, there is a sense that policy is being priced to "fail" rather than succeed. Real rates fall when central banks back away from stimulus presumably because they "think" they have done enough and the (global) economy is on a healing trajectory. This could be viewed as a damning indictment of policy and is not unrelated to other structural factors that make policy less effective than it would be otherwise -including the self evident break in bank multipliers due to new regulations and capital requirements. Of course our definition of "failure" may also be a little zealous. After all why should equities always rise in value? Why should debt holders be expected to afford their debt burden? There are plenty of alternative viable equilibria with SPX half its value, longevity liabilities in default and debt deflation in abundance. In those equilibria traditional QE ceases to work and the only road back to what we think is the current desired equilibrium is via true helicopter money via fiscal stimulus where there are no independent central banks. One step at a time... EFTA01405801 6mly-2y2y as a carry-efficient flattener We recommend a 6mly-2y2y flattener as an optimal carry proxy for USD 2s5s. The 6mly-2y2y can be thought of a leveraged version of the 2s5s spot: it has a 98% correlation and a beta of 1.74 with the latter over the last 12 months. Because of the 1.74x leverage, the beta-adjusted 3m carry is -2.0bp instead of -2.9bp for 2s5s, a 31% improvement. This flattener takes advantage of a recent 2.3 standard deviation decline in the (negative) roll for the 6mly paying leg, which compares to a 1.7 standard deviation decline in the 2y spot. The positive roll for receiving the 2y2y is also more attractive; it had just a 0.5 standard deviation reduction compared to a 0.7 standard deviation reduction in the 5y spot. Historically, 2s5s flatteners have performed well going into a tightening cycle, with nearly 70 percent of trades put on within three months of the liftoff beating their ex-ante forwards and thus being profitable. The market clearly thinks Friday's mixed jobs report was not enough to take a September liftoff completely off the table. DEC15 Fed funds future sold off 1.5bp after payrolls, and the implied probability of hiking in September rose slightly from Thursday to 34% at the time of writing. The 2s5s slope also flattened 2.5bp to 71.5bp, but still remains 7bp+ above its 2015 lows. A policy error by the Fed (i.e. hiking more than once this year in spite of declining global liquidity and falling inflation) can easily flatten 2s5s to 50bp or below. The risk to this trade is if the Fed relents in September but we think they will more likely than not do a Deutsche Bank Securities Inc. Page 15 EFTA01405802 4 September 2015 US Fixed Income Weekly "dirty" relent, which is to keep October and December FOMC dates in play. In this case 2s5s could steepen slightly but such a move would be short lived and limited in magnitude, if not for a hyped expectation of an October liftoff it would be because China's FX intervention flows continue to exert a flattening pressure on the curve, which we discussed earlier in this note. 6mly-2y2y as a leveraged proxy for 2s5s spot 6mly-2y2y (left axis) 100 120 140 160 180 200 Correlation = 98.4% Beta = 1.74 80 Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Source: Deutsche Bank Source: Deutsche Bank Flattening carry is 31% better in 6mly-2y2y than in 2s5s 2s5s spot (right axis) 105 115 125 55 65 75 85 95 2y spot 5y spot 6mly 2y2y BetaLevel (%) 0.83 1.54 2y-5y spot 0.71 0.82 1.87 6mly-2y2y 1.05 Dv01 / Ratio 1.983 4.838 2.44x 0.992 EFTA01405803 1.927 1.94x 3M carry (bp) (8.3) 5.3 (2.9) (15.1) 11.6 (3.5) 1.74 (2.0) 31% 1.00 (2.9) Beta adjusted carry (bp) Improve ment Risk/reward shifting towards paying front end spreads Front end spread tightening has been considerable given concerns about possible intervention-related selling, and has reached levels we think offer value. At the time of writing the most recent Chinese reserves data have not been released, and markets will naturally be looking for concrete evidence that intervention-related sales have indeed been material. While this may introduce event risk into paid positions in spreads, we think risk reward should be biased toward spread re-widening from current levels. There are three primary supporting arguments. The first is that China will be increasingly defensive of its reserves, and is more likely to devalue in a larger increment to discourage new speculation against the RMB and trap speculative capital. A large enough increment should significantly reduce further speculation on the margin and hence reduce the need to liquidate Treasury positions to sell dollars and buy domestic currency. The second is that there remains some possibility that if the Fed does indeed raise rates (which we think would increase the probability of further devaluation in a lumpy increment) that IOER will have to be set higher than the top of the desired band for overnight effective funds in order to create adequate incentive for banks to do the "arb" whereby they absorb cash balances in the overnight market and then deposit them at the Fed. Third, both a devaluation and the likely risk-off market environment that would accompany it should bias spreads wider. If, as remains our central expectation, the Fed does not raise rates, then we would expect speculative pressure against the RMB to decrease somewhat, slowing reserve loss and Treasury liquidation. So even though diminished financial stress might work against spreads in this scenario, EFTA01405804 intervention related selling could well decline. Financing is obviously critical with front end spreads, and this trade is complicated somewhat by high term repo rates relative to LIBOR. The September 2y note, given current levels, is likely to a reopening of the Page 16 Deutsche Bank Securities Inc. EFTA01405805 4 September 2015 US Fixed Income Weekly September 2017 5y note, which means the large issue is unlikely to trade tighter than GC. We note that as usual this September 2y should be the CTD issue into the December TU contract. In fact 3m GC has traded at levels above LIBOR as financing markets price defensively for the possibility of a September rate hike. This would effectively mean borrowing to fund the position at higher rates than offered by the Treasury asset itself. In this case investors are likely better served by rolling on open rather than locking in term financing. September month/quarter end could see elevated overnight GC levels, which would argue for exiting the trade at or shortly following the FOMC meeting. Naturally the trade is exposed to further spread tightening, and in theory potential losses are unlimited. However, more pragmatically, financing spreads offer some support against dramatic spread tightening. Dealer positions in Treasuries maturing in 2 years or less -10.0 -5.0 0.0 5.0 10.0 15.0 20.0 25.0 30.0 Primary dealer inventory <2 y Source: Federal Reserve and Deutsche Bank Did dealer positions tell much about intervention flows in the past? The concentration of foreign official holdings of Treasuries in the front end of yield curve suggests foreign reserve losses lead to yield curve flatteners to the extent that central banks sell their Treasury holdings. Treasury's TIC data shows that about 56% foreign official holdings of Treasuries mature within three years as of June 2014, up from about 48% as of June 2010. We note that primary dealer positions in short dated coupon Treasuries and TIPS have increased rapidly over the past few weeks. For example, dealer positions in Treasuries maturing in three years and less jumped to $18.6 billion on August 26; they were as low as -$11 billion in early July. Dealer positions in short dated TIPS set a record high on August 19. Deutsche Bank Securities Inc. Page 17 $ billion EFTA01405806 4 September 2015 US Fixed Income Weekly There has been an increased concentration in short dated Treasury holdings by foreign official institutions Dealer positions in TIPS maturing in less than or equal to 2 years 1,000 2,000 3,000 4,000 5,000 6,000 7,000 -1,000 0 Jan-14 Source: Treasury and Deutsche Bank Source: Fed and Deutsche Bank How much did intervention-related flows affect dealer positions in the past? To answer that question, we analyzed Japan's foreign exchange operations in US dollars and dealer positions in short dated coupon Treasuries from 1991 to the present. The most recent operations occurred in 2010 and 2011, when Japan bought US dollars and sold yen. The operations that sold US dollars and bought yen were less frequent and have not occurred since 1998. It was evident that dollar buying foreign exchange operations coincided with a decline in dealer positions in short dated Treasuries, but the effects were not overwhelming. On a related note, there has been an uptick in PBoC's OMO net injections of funds recently, in the order of CNY215 billion in the second half of August, which came along with the CNY depreciation. Last time when the net injections in this order of magnitude occurred was late February. Opportunities abound in inflation markets Volatility in inflation markets has continued along with commodities and equities, creating opportunities for active traders. We like being long front end breakevens in forwards, e.g., one-year breakevens implied by short maturity TIPS, such as the 7/2016s and the 7/2017s, currently trades around 1.3%. One can also hedge out energy prices in that trade to create a synthetic exposure to core CPI. For example, one can use gasoline RBOB futures Dec16 and