EFTA00292767.pdf

ATorus, LLC Backtest & Trading Assumptions Entry/Exit Time: Positions are closed over a five day period with an equal amount of shares/contracts sold per day. After being at zero exposure for one period and provided the system has a new price direction signal the new position is entered over an equivalent five day period. Given the targeted return profile (20+ days of realized volatility, which only occurs over longer term intervals of time), we find averaging in over these 5 period intervals increases the probability of a better entry price in contrast to putting on all the exposure in one period (i.e. full exposure trade time / median targeted return profile). Additionally, this greatly expands the liquidity pool, which further enhances the ability to scale the strategy. Price Data: All listed equity prices are VWAP prices for the day. Index and credit data is based on close prices. We view VWAP as the true price where a position of size can realistically be expected to be transacted at. An added benefit, by committing to a certain size VWAP order for the day, the fund can benefit in lower transaction costs or even receive a net credit. On the first trade date of a new signal we always wait until midday to determine whether to begin executing the order. Specially, once we have seen half a trading interval occur, we can calculate the probability of where VWAP will be end of day, which is the price that inputs into the system to determine a trade signal. Position Sizing: Positions are sized based upon the amount of dollar risk allocated to the portion of the book (global macro instruments or listed equities) than multiplied by the respective maximum portfolio target loss (in this case 10%) then multiplied by the current position's realized vol times a factor (7X) and divided by the total amount of positions in the portfolio, and finally multiplied by an initial leverage target (2X). This approach allows every position, at least at inception, irrespective of dollars allocated to have equivalent potential NAV impact to the fund (e.g. higher vol names receive lower dollar amounts of exposure, while lower vol names receive higher dollar amounts of exposure) Directional Bias: The directional risk is determined by the system on each individual position in isolation. Directional bias is not determined, nor are positions resized, based up assumed cross asset correlations or the overall portfolio's net exposure at the time. This is a differentiating component of the strategy, especially in comparison to 'risk-parity', 'systematic global macro', and 'multi strategy' funds. Position Screener. Security characteristics of the listed equity book are based upon a screener that begins in 1995. The starting minimum market capitalization is $10B w/ minimum averaged daily dollar volume over the preceding 3 months of 2.5% of the $10B. The $10B threshold appreciates at 7% per annum with the relevant 2.5% liquidity parameter being reset to the relevant minimum market cap. The screener set to these thresholds, is to eliminate selection bias to small cap equities that appreciated in an exponential fashion and generate exceptional outsized returns, in addition to displaying the depth of the liquidity pool allowing the strategy to scale markedly. The screen is run on a quarterly basis. The listed equities also must be listed in only the following countries: Australia, Hong Kong, Japan, South Korea, Brazil, Canada, United States, Austria, Belgium, Finland, France, Germany, Greece, Ireland, Italy, Luxembourg, Netherlands, Portugal, Norway, Spain, Sweden, Switzerland, United Kingdom. Once the new securities appear, we begin to trade them in accordance with the directional trade signals we receive from the system. To the extent the securities no longer meet the market capitalization & liquidity thresholds at the next quarter, we still allow them in the portfolio for the subsequent two years. In some cases in the event of a merger, bankruptcy, or significant decrease in liquidity the CIO will remove the securities. Although we run the screen quarterly, we do not rebalance the portfolio every quarter. Similar to utilizing cross asset correlations to determine position weightings, we do not employ a rebalancing strategy, but rather let our winners, given MTM gains, dictate our increase/decrease in our exposures. All positions continue on their present exposures until such time a new opposing directional trade signal is observed. At that time the position sizing is recalculated to reflect the change in fund NAV and the number of positions, as described above. 1 Private & Confidential EFTA00292767  ATorus, LLC Backtest & Trading Assumptions Size of Portfolio/No. of Positions: The global macro book includes all G20. Eurozone (original member states). and BRIG equity index futures and credit indices, where available. We've removed Saudi Arabia and replaced with New Zealand. The anticipated number of positions is 100. The listed equity book generally has between 300-500 individual positions at any one time. The number of positions in this case is dynamic to the expansion/contraction of the opportunity set output by the screener. Allocation btwn Global Macro Instruments & Listed Equities: We allocate risk exposures (defined as targeted max loss given realized vol). not notional exposures, equally between the Global Macro Instruments and the Listed Equities. Given that the Global Macro book has a smaller number of positions and is fixed the dollar risk for each Global Macro book position is materially larger, which is in comparison to the Listed Equities, which had 100s of securities and a non constant number of positions. Dividends: We add (when long) and subtract (when short) dividends to our cumulative realized gains/losses on the ex-date. FX Risk to USD for Foreign Currency Denominated Positions: All positions are priced at the current dollar exchange rate. We eliminate our FX risk by hedging via spot at the same time we enter into a foreign FX exposure associated with a position. 2 Private & Confidential EFTA00292768