EFTA01451144.pdf

9 January 2014 EX Blueprint Thin end of the wedge Theme In: Trend no bitter end In 2007 we replicated a RBA study that claimed non- ....-....-....-....- commercial "profit seekers" made money on their IMM Figure 1: Absolute P&L (Smil) of Profit-Seekers and positions at the expense of commercial "liquidity Liquidity-Seekers Using G5 IMM Data (1993-2013) seekers". 3 The striking conclusion is that currency speculators generally know the correct direction of 05 19 USD mm (19912013 except euro 1999-2013) currencies and investors can profit from knowing their allon-Cornmercial/Profrt-seekers positions (although these signals are less useful in ,c, 12 practice since IMM data is lagged three days). CommerclaUlletikety-seekers ca cr- 6 In a sense, money is "left on the table" by commercial foreign exchange users, and to a lesser extent by foreign bond and equity investors, which can be earned 3 0 ce to by non-commercial actors that provide liquidity to o currency markets. These profit seekers collect FX risk premia in the same manner Keynes first identified when describing risk transfer in commodity markets.' -12 919495909191910010200‘00007080910111213 &vv., oetac* ant &warn France 1LP Profit Seekers Made Money in Every Year since 2003 By our calculations, non-commercial positions made money in every year since data was first released in Figure 2: Speculators had the correct EUR positions in 1993, mainly at the expense of commercial users 2010-I I and a large JPY short in 2013 (dealers also made money).5 In recent years profits have come from timing big EUR/USD moves (2010-11) 12 n USD mm (1993-2013. except euro 1999-2013) and catching last year's USD/JPY rise. These P&L numbers more closely resemble pre-crisis profits than t M_ 8 the outsize gain in 2008 (reflecting higher FX volatility) and the nearly flat 2009 period (probably due to an unexpected GBP rebound). t. 0 et 4 IS P&L Eroded As Speculators Accumulate Positions? Our analysis rests on the crucial assumption that profit o -8 seekers accumulate positions over the course of the .4)-12 • Non.commerciali Profit-seekers week at the average price. By contrast, live trading • Commerciali Liquidity-seekers • -16 metrics such as the Parker Index of currency manager AUD GBP CAD JPY CHF BA returns show a loss since 2011 as currency volatility has overwhelmed macro trends (with the exception of Soso Dostcheao* Stones, Finn* UP USD/JPY in 2013) even as the correlation between weekly IMM P&L and Parker returns remains positive. Figure 3: The Parker Index of currency manager returns It is possible that infra-week volatility causes profit has lagged (MM P&L in recent years despite continued seekers (especially momentum traders) to "buy high and sell low" relative to WVAP and that IMM P&L is positive correlation between them eroded when currencies trade in a choppy range. Average IMM Cum P&L (5100m, Ihs) Fortunately for investors FX volatility continues to fall 45 1 Parker Index ben-2003 o 0, Ihst ttAM P&L V. Parker (1im corr.rhs) 100% r and promising trends (JPY, CAD) have emerged. 40 1 at , 90% 35 I Jr ...„ AA/ - 1- 80% Daniel Mellon, Now Yor; +7 (2721250 7639 !.., • 70% 30 1 1 25 20 I 50% 15 1 3-r • 40% iv ic 10 4 F 30% ;,..,./.--...-.. 3 See Kearns and Manners 120041 "The profitability of Speculators in 5 • ......: I 20% .0° Currency Futures Markets-. Reserve Bank of Australia, and Biel r ..-- -i. 10% Kafeez (2007), "Currency Markets: Is Money Left On the Table?" .0 , S. L ma ' Keynes, J.M. (1930). "Treatise on Money London: Macmillan, 03 04 OS 06 07 08 09 10 11 12 13 We calculate weekly P&L by determining the notional value of Son Deux.* Bett frorntrep Rona* UP positions on Tuesday and assuming longs and shorts are accumulated (or squared) at the average puce over the course of the week Page 14 Deutsche Bank AG/London CONFIDENTIAL - PURSUANT TO FED. R. CRIM. P. 6(e) DB-SDNY-0 107387 CONFIDENTIAL SDNY_GM_00253571 EFTA01451144