EFTA01385923

27 March 2015 US Fixed Income Weekly The Front End The front end is all about the timing and speed of lift off. Speed will probably be more important than even timing in that the market twill easily absorb a one and done Fed. Speed can also backfire in that if risk assets behave poorly for any given speed, the Fe won't be making their terminal rate in one swoop. An intra tightening period of stable Funds would allow for a sharp re-rally in short rates at some stage (Note 5s typically converge to Funds at the end of a tightening cycle, so any sense that the fed would initiate a pause for several quarters would allow substantial spread convergence in our view). We remain of the view that this Fed does not want to commit the type 2 error and will be hard pressed to even begin raising rats in 2015. The bar is low for them to delay in that if the payroll report is all they have to go on, a soft patch in jobs could easily prompt ongoing delay. Using private final demand the long run output gap has barely improved since the crisis. One has to go the other extreme, to measure the gap from 2009 to argue that the gap has in fact closed. Either way the link with inflation has become tenuous. [Output gap since 1960 15 10 S I-I • -I •-• ...... ei in CO •-• 'e P. O in 1.O O1 P4 cri 1O ID 1.O CO CO CO CO O1 g m O W. CP O1 O1 O1 O1 O1 O1 O1 output gap since 1960 00 N. 0 en g 8 g 8 0 0 Sane 1W Woonberg fewer IP inze Doteschr 'Output gap on trend since 2009 2.5 — 2 1.5 1.5 1 0.5 0 -0.5 0.5 • I. CPI yoy -1 -output gap since 2009q1 1.5 P.444 c? <1. 4* 1) evt* 41. ebb. ;I* gist 61. oS 6S, t o% t% t% % 1 IF A 1, 1, 1, 1, 1, 1, 1§) 1? Seran• aff, 813Crntrerg leave (Pro , Gnu,* arst In terms of wage inflation below we update our wage model to incorporate the Fed's new NAIRU estimates. If unemployment does not continue to fall, there is still no wage acceleration. If it does then with the lower NAIRU estimates wage acceleration is still delayed until late 2015 - all consistent with a Fed struggling to raise rates. Unemployment Model In order to understand the relationship between wage acceleration and unemployment, we first project the year-over-year change in the growth rate of the average hourly earnings of production and nonsupervisory workers (ARE) on two variables. The first is a dummy variable that equals one if the unemployment rate is less than the CBO's estimate of the NAIRU, and equals zero otherwise, and the second is the year-over-year change in the unemployment rate. The thought experiment is that wage inflation should be affected by whether there is slack in the labor market and by the trajectory of job growth. As shown in the chart below, the above-discussed variables are able to explain a fair amount of the variation in wage acceleration: the R- squared of the regression is 33°k. Page 10 Deutsche Sank Securities Inc. CONFIDENTIAL - PURSUANT TO FED. R. CRIM. P. 6(e) DB-SDNY-0087391 CONFIDENTIAL SDNY_GM_00233575 EFTA01385923