EFTA Document EFTA01385936

27 March 2015 US Fixed Income Weekly No directional cue emerged after the FOMC. It is effectively a distributional modification, a swelling of the tails. The probabilities of rates going both higher and lower have increased, at the expense of the likelihood of staying in the range. The novelty of the last FOMC meeting is the mechanism whereby Fed consensus is converted into market dissensus. Fed language became a pure volatility effect. With tails probability higher, confidence regarding a particular directional position is undermined - it is now subject to quick revisions at the slightest market move. Thus any residual overweight in assets for which valuation has been distorted by stimulus is likely to come under scrutiny and possibly be corrected. That should free some maneuvering space for the Fed and make potential hikes less damaging. As far as an attempt to return vol to the markets, this is mission accomplished. Withdrawal of stimulus, if not carefully executed, has potential to expose the underlying negative convexity of the market created by the monetary policy itself. This can be seen from two complementary angles. Implicit belief in the Fed as a global market stabilizer has made both credit and equities behave like a positively convex asset. Whether the economy was improving or not, there was always stimulus as an alternative support in case economic data disappoints. As a consequence, both asset classes developed additional desirability due to embedded optionality. In that sense, unwind of stimulus is a withdrawal of "free" optimality of risk assets - and vice versa, delayed exit is an extension. Risk of stimulus unwind is all about the speed of events. The mechanics of this can be understood by visualizing its actual realization. In less liquid markets, like credit, this is especially easy to see. If unwind is too fast, the street would not have time to flatten its position in the interdealer market and therefore would be unable to extend liquidity further. This would cause additional spread widening with stop outs and likely panic selling, further undermining already-fragile liquidity. This is why volatility should not be allowed to increase too much too fast. Its return to the market prior to the final stage of Fed exit is essential, and that process has to be fine tuned. In the same way the long period of artificially low volatility led to positioning buildup in carry trade and risk assets, the longer volatility remains elevated the better chances would be for "bad positions" to clear. We are buyers of tail risk at the short end of the curve in the mid-run. In our view, risk assets are also at a bifurcation point - their future path depends on the way the economy and stimulus unwind cooperate with one other. We are buyers of hybrid S&P calls and puts conditioned on different rates responses to the Fed. Hates: Straddle/sitangle switches Vol and tail risk are likely to be concentrated at the short end of the curve. To a large extent, vol surface is pricing this in through elevated volatility risk premia. Figs 1 & 2 show the history of realized and implied volatilities for 3Y and 10Y tenors. This rise in vol risk premia in the upper left corner is a relatively recent phenomenon that started in mid-2014. Fed communication at this stage of policy unwind is largely transmitted through the front end of the curve in the sense that most of the play defined by the dots and economic data concerns the timing and magnitude of rate hikes. At the same time, market flows and foreign central banks' actions are likely to constrain rate movements at the back end of the curve. Deutsche Bank Securities Inc. Page 23 CONFIDENTIAL - PURSUANT TO FED. R. CRIM. P. 6(e) DB-SDNY-0087404 CONFIDENTIAL SDNY_GM_00233588 EFTA01385936