Highest absolute return : MarketSci
Best risk adjusted return : MarketSci
So now things are getting interesting. By definition, if Automatic 7 does not rotate into a new asset class at the end of the month, then we can expect that the returns of Automatic 7 will be near the top of the pack. In April, we had a half – rotation, meaning that one of the two asset classes held by Automatic 7 was swapped (VNQ replaced VTI), while DBC remained in the portfolio. So while Automatic 7 did keep pace with GTAA (3.7% vs. 3.6%), it did not keep up with MarketSci, which benefited on the absolute return front from the out performance of Gold (+8.9%) and on the risk – adjusted return front from the diversifying effect of Treasuries (+1.8%). A7 does not consider owning gold outright, and at most, will own some gold via an investment in DBC, which functionally means that A7′s precious metal’s allocation never gets much higher than 5%. When gold is a stand out performer, A7 will not keep pace. I would like to include gold in A7′s investable universe, but I will only do so if and when I can find a method that makes sense and doesn’t impair backtested results.
I’m still stuck on the idea of benchmarking, and I think that perhaps the most appropriate benchmark for A7 is the VT ETF, which is Vanguard’s Total World Stock Index Fund, or maybe even a simple timing version of VT (i.e. long VT only when it closes above it’s 10 month SMA) given the fact that A7 is really a risk seeking program that relies on return and volatility timing to switch asset classes including “less risky” assets such as bonds and cash.
At the end of April, Automatic 7 still had the best absolute return since inception (5 months ; +20.6%), but storm clouds are brewing as DBC has dropped more than 10% in the first week of May, and it’s fall from the top of the A7 asset table comes at the very beginning of the month, which represents a worst case scenario if in fact the decline continues. A7′s only trading rule at the moment is time based (i.e. trade once, at the end of the month), so this is an area for research…at what point, if any, should A7 switch asset classes mid-month?
| System | April 2011 | |
| Automatic 7 (A7) | Sym (Wt) | VTI (50), DBC (50) |
| Return | 3.7% | |
| Volatility | 11.2% | |
| Ret/Vol | 33.0% | |
| Cumu $ | 1.206 | |
| GTAA (A7 Bench) | Return | 3.6% |
| Volatility | 7.8% | |
| Ret/Vol | 46.2% | |
| Cumu $ | 1.111 | |
| A7 Mom and Pop | Sym (Wt) | BND (25), STPZ (25) ,VTI (25), DBC (25) |
| Return | 2.6% | |
| Volatility | 5.4% | |
| Ret/Vol | 48.1% | |
| Cumu $ | 1.101 | |
| A7 M&P Benchmark | Sym (Wt) | BND (40), VTI (60) |
| Return | 2.3% | |
| Volatility | 5.6% | |
| Ret/Vol | 41.1% | |
| Cumu $ | 1.103 | |
| MarketSci TAA | Sym (Wt) | SPY(29),GLD(29),GSG(19),IEF(16),VNQ(7) |
| Return | 4.9% | |
| Volatility | 7.1% | |
| Ret/Vol | 69.0% | |
| Cumu $ | 1.070 | |
| Cumu $ = Cumulative Growth of $1 since A7 inception (11/30/2010) | ||
Returns and volatility calculated at ETFReplay.com (03/31/2011 – 4/29/2011), rounded to nearest .1%