James,
Out of the 17 ETF strategies available as designer models with two years of data, 13/17 (76%) underperformed the benchmark and 4/17 (24%) outperformed.
Why not launch a designer model and see what happens over the next year or two?
Chipper,
I am not very familiar with designer model but my ETF strategy requires daily rebalancing (esp the leveraged and inverse ETF system) to work. I don’t believe that works for designer model that weekly rebalance based on my understanding.
Furthermore, there is a capacity to how much the Leveraged and Inverse ETF system can take since the daily trading volume of both ETFs is only around USD 50 mio.
Regards
James
Okay James. At 84% a year you should be the richest man on earth soon. 
Chipper,
Out of sample result for the past year is about half of the long term backtested returns. (only about 40-50%).
But I am happy with the performance of the system so far.
Regards
James
“Only” 40% - 50%. Good luck keeping up those numbers.
Chipper,
One more point that was missing in my previous reply,
As my investment portfolio grows, the Leveraged/Inverse Strategy sub system can no longer be fully invested. (ie capacity constraint - not exceeding 5% of average daily volume)
Regards
James
So it appears that 40% is not that uncommon on p123.
I am inspired to reexamine ETF strategies. Thanks for the motivation!
James/@ustonapc,
I agree with Walter – thanks for inspiring me as well. The best I have been able to do using ETFs in back-testing is CAGR 16.51%, during 01/02/2002 … 12/31/2022, with QQQ risk-on, GLD risk-off, and SPY Golden cross/Death cross / UI trend timing signal.
Cary
I’m still following Smartly.com, but I agree that most of the models trail S&P. That said, I don’t think that S&P is the best index to compare to. The index has given a performance the last 10 years that seems exceptional. https://www.youtube.com/watch?v=RR7e1Y-HJxQ
Here are the results for 2015–2023 for the tree strategies that I follow, compared to the SPY VTI and VT.
Its not that bad, and then with a low total drawdown, low correlation to my stock portfolio, low volatility, and good ulcher-ratio performance, its a good addition to my stock portfolio:
| SPY | VTI | VT | Bold A | o3 | Dual M | |
|---|---|---|---|---|---|---|
| 2023 | 13.03 | 12.39 | 9.65 | 1.80% | -0.20% | 5.00% |
| 2022 | -18.17 | -19.51 | -18.01 | 2.30% | -7.20% | -24.10% |
| 2021 | 28.74 | 25.67 | 18.27 | 5.20% | 18.70% | 19.90% |
| 2020 | 18.37 | 21.03 | 16.61 | 26.60% | 29.60% | 28.40% |
| 2019 | 31.22 | 30.67 | 26.82 | 16.00% | 8.40% | 12.40% |
| 2018 | -4.56 | -5.21 | -9.76 | 13.00% | 2.40% | 4.30% |
| 2017 | 21.7 | 21.21 | 24.49 | 11.90% | 15.60% | 23.00% |
| 2016 | 12 | 12.83 | 8.51 | 10.10% | 9.10% | 0.50% |
| 2015 | 1.25 | 0.36 | -1.86 | -2.10% | 0.10% | 0.80% |
| 103.58 | 99.44 | 74.72 | 84.80% | 76.50% | 70.20% |
James,
thank you for posting this.
It is very impressive, to say the least.
How do you determine your “risk on / risk off” periods?
Werner
Werner,
As I have posted in another thread, this is part of what I use to determine risk-on/risk-off periods (daily rebalancing) for the medium/long term.
(sMA(10,0,##CORPBBBOAS)<(sMA(500,0,##CORPBBBOAS)+sMA(500,500,##CORPBBBOAS))/2 OR Close(0,GetSeries(“qqq”))>1.15* ema(200,0,GetSeries(“qqq”)) or Close(0,GetSeries(“iyt”))>1.05* HighValbar(200,0,GetSeries(“iyt”)) or Close(0,GetSeries(“dia”))>1.05*ema(200,0,GetSeries(“dia”))) &RSI(7,0,$SPALLINT)<85 & Close(0,$VIX)/sMA(10,0,$VIX)<1.2 & eMA(5,0,##TEDSPREAD)/eMA(20,0,##TEDSPREAD)<1.1
In addtition, I also use short term momentum rules in my sub-systems.
Regards
James
James,
thank you for your reply.