From: Lesley Groff Sent: Thursday, February 6, 2014 5:04 PM To: Jeffrey Epstein Subject: Fwd: ATorus - Daily Portfolio Report - 2/S Attachments: Atorus_BacktestNAV_020514.pdt Untitled attachment 00446.htm Begin forwarded message: From: Michael Fowler =1 Subject: =/b>ATorus - Daily Portfolio Report - =15 Date: February 6, 2014 12:02:25 PM EST To: Lesley Groff Lesley, =nbsp; Please see attache Daily Portfolio Report for =15. Daily =ommentary: We added two =dditional metrics in the report. Firstly, we broke down the number of =ndividual names long and short by category, instead of just net and =ross exposure by category. In times of potential broad market =nflection points, you will see our number of individual shorts increase =rior to our net exposures changing materially. This is a result of the =ay we position size and then leverage the winners, rather than =argeting a specific net exposure level. We let the market's trajectory =ictate our exposures, when we are experiencing positive MTM =ains. Secondly, we displayed our P/L per position in what =e call "vol days." To us the notion of looking at the MTM gains =djusted for the realized volatility (recalculated daily), determines =hat part of the gains are significant and what is random oscillations. =eparating the signal from the noise, if you will. Our key insight in =ow the distribution of the exponents of how volatility can scale =eriod/period being nearly constant, allows us to determine the =robability of actually monetizing the current gains. For example, given =hat the 99th percentile of the distribution of the exponents is 3, any =ain of less than 6 vol days is, in our mind, statistical noise. In =ther words, if the entry and exit days were both 3X vol day moves, the =ensitivity to entry and exit days (2 periods @ 3X vol days) would =epresent a large amount of the total return. We can not forecast the =recise amount of volatility at the next day, but we do know with =ertainty the