Thank you for the response Jessica, I appreciate the time and I'm very grateful for any feedback you can provide. If you'll bear with me, I'd like to share a couple thoughts on the subject and outline the reasons why I believe a CTA allocation is appropriate at this time. I apologize if this response is somewhat lengthy, but the information is all relevant.
Before February, conditions meant systematic strategy equity positioning was highly concentrated on the long side. When the short-vol trade blew up (starting with higher vol due to rising treasury yields/higher inflation concerns and ending with the VIX ETP/F microstructure), the increased volatility meant forced deleveraging from systematic strategies including CTAs and risk parity (since these strategies incorporate volatility as an input). Now, this forced deleveraging campaign is largely complete. From here, if market conditions allow for new highs in equity indexes, these strategies should be supportive as the positions liquidated in February are partially re-established. Alternatively, if markets experience more turbulence, that would also be good since my algorithm tends to significantly outperform during times of stress. For example, throughout my 2006-2016 backtests, my algorithm outperforms during almost every single stress period (as ID'd in the full tear sheet backtest analytics).
Another reason to allocate to a CTA is direct commodity exposure. Each of my algorithms are backtested using 22 futures and only 4 are equity indexes. (Backtesting issues arose on some contracts; I'm in the process of backtesting with a much higher number of contracts) Current conditions favor commodity outperformance as inflation is increasing and we are in a late-stage business cycle. Additionally, while commodity exposure at this time is desirable, the logic will capture downside trends if a shock occurs. (Performance in 2008 was excellent by any measure)
Finally, another reason to allocate to a CTA is simply diversification. Although you surely eliminate a lot of market risk through market neutral/absolute return strategies, I'm sure you can appreciate the value in exposure that is independent of equities altogether, particularly during periods of extreme stress. The fact that the algorithm possesses exposure to currencies, rates, and a wide-range of commodities (from different sub-groups, with little correlation between them) is a huge advantage over equity-only strategies. I know there are several ETFs based on commodity futures, but they will have significantly less liquidity and trading costs will be much higher (e.g., with futures, a 2$ commission can provide $100,000+ notional exposure).
I'll just finish by saying my algorithms implement futures due to the benefits of this asset class, NOT in order to take highly leveraged positions. In fact, the leverage ratios of my backtests are fairly low, typically 1-3. Even so, risk exposure is roughly equal regardless of underlying notional value since the position sizing formula incorporates a measure of volatility. As such, the leverage ratio is less important (e.g., compare 2Y treasuries and crude oil). Also, as I stated in my first post, I can calibrate parameters to achieve very low volatility while still providing superior returns. Below is an image of backtest results from 2006-2016 where volatility is 0.12 yet still outperforms SPY by over 19%.
I would absolutely appreciate any feedback you can provide and I'm also interested in gaining more insight into Quantopian's investment objectives. Since I only provided images, could you indicate what else I can send to receive feedback without compromising the algorithm logic? My personal email is [email protected]. Please feel free to send me an email anytime. I appreciate your time and again apologize for the lengthy message.
Thank you
Backtest 4 - Low Volatility 2006-2016
