EFTA Document EFTA01362020

4 September 2015 US Fixed Income Weekly "dirty" relent, which is to keep October and December FOMC dates in play. In this case 2s55 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. fern I y-2y2y as a leveraged proxy for 2s5s spot 200 1 —Ctody-2y2y (IeIt axis) 180 160 •• 140 120 100 Correlation = 98 4% Beta = 1 74 80 Flattening carry is 31% better in Urn I y-2y2y than in 2s5s 125 r•••••• 24.54 wet 08(1144,1) r 115 r 106 Level (%) Dv01 Ratio 3M carry (bp) Beta .9S 2Y spot 0.83 1.983 (8.3) 51 spot 1.54 4.838 5.3 SS 2y-5y spot 0.71 2.44x (2.9) 1.03 IS 6mly 0.82 0.992 (15.1) 65 4 2Y 1.87 1.927 11.6 6m1y-2y2y 1.05 1.94x (3.5) 1.74 Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Sant Onsets Sep Swot Dna., Bent Beta- adjusted Improve carry (bp) meM (2.9) (2.0) 31% 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, 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. CONFIDENTIAL - PURSUANT TO FED. R. CRIM. P. 6(e) DB-SDNY-0051317 CONFIDENTIAL SDNY_GM_00197501 EFTA01362020