Chart 16: SAN, BNP, ING, ISP and DBK average 6M ATMf implied vol is Chart 17: Despite the recent drop in realized correlation, implied trading historically low (13" percentile since 2008) correlation is priced near the high end of the reaiised range for the basket of EU banks Basket of SAN, BNP, ING, ISP & DBK 3 100% 90% 80% 80% 70% 60% 60% 40% 13th percentile 50% 40% 20% 8 3 2 = 2 2 = 2 2 30% L ob a & a ale a ale a ale N MO Fw ORF DOD DODO T nN OTF OO OO © © © © © © © © © © Oo GC OQ OO CG ee, & he Se Se Se Me SE = => = FSF > FF FF FEF ELD <s<© ¢ ¢ ¢ € © & © & Ce G&G GC FG Cc _E oO oO oO oO oO oO oO oO oO oO oO oO oO oO oO oO 7 wa 7 7 wa a) 7 oe) 7 a) oe) 7 a) 7 7 wa 6M avg realised vol ———=6M avg implied vol = Last IV (16-Jun-17) 3M realised correl ———=6M realised correl —-Dec17 implied correl Source: BofA Merrill Lynch Global Research. Data: 2-Jan-08 to 16-Jun-17 Source: BofA Merrill Lynch Global Research. Data: 2-Jan-08 to 16-Jun-17 - Attractive risk-reward profile at current pricing: As highlighted in Exhibit 1, historically the trade held to expiry, at current pricing, would have generated an average P&L of 8.4% when positive and -1.8% when negative. The risk-reward looks even more attractive in extreme market outcomes as the max P&L of the trade which is greater than 75% compares to the max loss of only 5.7%. The trade also provides an effective way to gain long exposure to EU equities with limited risks, as evident from the call-like payoff in Exhibit 1 (vs the ESTX50). It is worth noting that, by construction, the maximum loss of the trade is 6.8% with the most likely loss limited to the upfront premium of 1.8%. We also note that the trade payoff profile is superior to a SX5E Dec17 ATM call when sized such that: (i) the call premium is the same as the upfront premium for the dispersion trade (=1.8%, blue line), as well as (ii) when the call premium is the same as the theoretical maximum loss for the trade (=6.8%, orange line). Exhibit 1: Hypothetical back-te