I am working on a low draw down screen, low correlated ETF's that has good momentum. The idea is to have it similar to Dalio all weather, Harry Browne screens and Permanent Screens.Instead of static list of ETF's it would be dynamic and what is working lately.Right now QQQ and GLD are very strong and low correlated.Add TLT and the 3 seem uncorrelated and give good results with low draw down.I am familiar with all asset allocation systems and testing on internet.Anyone doing this or any suggestions?
wan123:I am familiar with all asset allocation systems and testing on internet.I've had to buy gifts for people who seemingly have everything, which is hard enough, but this is the first time I've ever been asked for input from someone claiming to have read everything that the internet has to say on a topic.Robbie Geary
wan123:I am familiar with all asset allocation systems and testing on internet.I've had to buy gifts for people who seemingly have everything, which is hard enough, but this is the first time I've ever been asked for input from someone claiming to have read everything that the internet has to say on a topic.Robbie Geary------------------------------------------------------------------------------Thanks for your comment. I agree with what you say.Sorry, I did not mean I have everything.I meant I know about ETF replay, Allocate Smartly, Portfolio Vizualizer, no where near everything. My intention was to precede someone suggesting those test sites. I guess I stated it incorrectly. We all make mistakes. I know much less, than most who post on this board, and am amazed at the knowledge, effort and sharing on this board.
I appreciate it when someone corrects me, especially if is not my wife.
I am working on a low draw down screen, low correlated ETF's that has good momentum. The idea is to have it similar to Dalio all weather, Harry Browne screens and Permanent Screens.Instead of static list of ETF's it would be dynamic and what is working lately.Right now QQQ and GLD are very strong and low correlated.I’ve been using this ETF screen for the stable portion of my investments. It is a slight modification of Keller and Butler’s Generalized Protective Momentum which combines momentum and correlation in the decision process. And it has been in QQQ, TLT and GLD recently.Up almost 15% this year. The standard version has excellent historical performance with synthetic ETFs at Allocate Smartly. Best write up:https://indexswingtrader.blogspot.com/2016/06/deciphering-co...Allocate Smartly write up and backtest results:https://allocatesmartly.com/keuning-kellers-generalized-prot...GTR1 backtest of slightly modified version from 20071015:http://gtr1.net/2013/?~Protective%20Asset%20Allocation%20Fuz...Results from 20071015 with 0.15% friction. GPM SP500wDiv CAGR: 8.05 8.42 TR: 169.83 181.33SAWR(20; 0.95): 6.71 6.24 GSD(20): 9.77 19.73 DIGSD(20; 0%): 10.17 23.86 LDD(20; 0%): 5.68 13.88 LDDD3: 5.35 13.10 MDD: -15.63 -54.38 UI(20): 4.27 14.64 Sharpe(20): 0.81 0.52 Beta(20): 0.13 1.00 TI(20): 61.51 9.09 AT: 4.44 0.00RAM
GPM SP500wDiv CAGR: 8.05 8.42 TR: 169.83 181.33SAWR(20; 0.95): 6.71 6.24 GSD(20): 9.77 19.73 DIGSD(20; 0%): 10.17 23.86 LDD(20; 0%): 5.68 13.88 LDDD3: 5.35 13.10 MDD: -15.63 -54.38 UI(20): 4.27 14.64 Sharpe(20): 0.81 0.52 Beta(20): 0.13 1.00 TI(20): 61.51 9.09 AT: 4.44 0.00
Similar to / uses some components of the old Riding the Wave approach!
Hi Wan,I went to the Portfolio Visualizer site where you can test different strategies on equities. From 1985 to present, your combination of equal parts QQQ, TLT, and GLD produced the following results:CAGR__11.6%WORST DRAWDOWN__-16%WORST YEAR__-1.7%SHARPE__1.1SORTINO RATIO__2.1CORRELATION WITH MARKET__44% Except for 1/4 of my portfolio which is invested in SAAS stocks, I am a very risk averse investor and I consider your suggested strategy to be excellent when compared to the S&P 500. S&P 500CAGR__8.7%WORST DRAWDOWN__-51%WORST YEAR__-37%SHARPE__0.56SORTINO RATIO__80.0CORRELATION WITH MARKET__100% My only small nit with your suggestion is could I stomach the 16% worst continuous drawdown? Note: I considered that you would rebalance annually.For the Portfolio Visualizer results, please see the following linkhttps://www.portfoliovisualizer.com/backtest-portfolio?s=y&a...Elliot
Elliot, Read the banner above your results. You'll see your portfolio begins in 2005, not1985. You can extend it to 2000, the earliest QQQ can be used in the free PV site, by changing TLTto VUSTX and GLD to RGLD (not completely equivalent to gold, but related to gold and trading at that time). Robbie just posted a number of extended indexes, and one might be a version of QQQ so thatmight allow you to further extend your test at gtr1.net/2013.If you believe, as I do,that bonds will begin to drop as interest rates rise sometime in the next ten years or so, and inflation will eventually increase as congress blows up the dollar, you might consider50-50 QQQ and RGLD or FNV. To avoid worrying about the drawdowns, which are larger than you might want to see, don't look at it often, but fund it as often as you can! Seems like a joke, but this 'set it and forget it, ignorance is bliss' thing fortunately happened to me. When my pension fund was started I didn't know much about stocks or bonds, so I got equal amounts of a spy like fund and a bond & mortgage fund. At that time, long ago, I think those and a straight annuity were the only choices I had. I mostly ignored it except when my annual report was sent to me. Ended up fine. Probably better than if I had played with it and with much less stress.rrjjgg
This is an model from Portfolio Visualizerhttps://tinyurl.com/y5zway8e1993 - 2020CAGR 12.45 VS SPY 9.93Sortino Ratio 2.04
Hi Charlie, I’d like your model but I am not sure how to implement it myself. For this model, How do I compute rank weight orders?For performance periods what does multiple periods mean?For allocation weights, how do I compute minimum variance for each etf?I don’t understand how do you find out what is the 20 day Volatility for each etf.I did check portfolio visualizers site but could not find answers to these questions. The only thing about these definitions that I could find was some not usable verbiage when I chose the “dot” next to the criteria. Is there a website where I can find all these measures for each of the ETFs?Thanks for your help. Elliot
ElliotI'm sorry I do not have answer for your questions.The answers are beyond the scope of my knowledge or capabilities to explain.I'm sure that you can find some answers within these search results:https://duckduckgo.com/?q=adaptive+allocation+minimum+varian...I'm not trying to give you a smart butt response here. But my problem is that I have read enough about it to believe it has value but cannot explain it nearly well enough for someone to implement it. I'm in the position of trying to explain in detail what enables an airplane with a weight over 1,000,000 pounds to lift off from the ground, travel in a stable mode for 2,000 miles by an efficient route, and safely land within 100 ft. of a predetermined spot on earth.The result of the model is not exactly fantastic. Certainly some years are disappointing compared to SPY. however almost any diversified asset model will have under performed for the last 10 years as SPY has been the outperforming asset class. Something such as the model could allow an investor to believe that using a 4% or 5% withdrawal rate would not lead to ruin while in retirement.
Elliot,This may be a good place to start. Not particularly the best explanation but a good start. Sometimes it works better to tart in the middle and work backwards to get a full understanding.https://indexswingtrader.blogspot.com/2017/11/matrix-iterati...I'm at a point where I accept that it works to somewhat stabilize returns.
Charley, This may be a good place to start. . . .Indexswingtrader.blogspot . . .Indexswingtrader is affiliated with allocatesmartly. They have extended the Adaptive Asset Allocation back further, tested and compared to other algorithm’s using same extended ETF/Funds plus compared alternate day of the month trading results. Their blog spot on AAA.https://allocatesmartly.com/adam-butler-gestaltu-adaptive-as...The alternate day and safe annual withdrawal results are not quite as good as the eom results but still significantly outperform any of the static allocations. The specifics are behind a allocatesmartly subscription paywall.RAM
Thanks Charlie for your ideas and how I can learn more about adaptive allocation. I’d like to mess around with portfolio visualizer but I am limited to the metrics that I know so I don’t use some like I mentioned in the post. Elliot
Elliot,This article from Seeking Alpha may have an interesting model for you. The author has made many nice contributions there. The article and comments are about the asset rotation strategy and avoiding massive drawdowns. It is not really an attempt to be a gunslinger but achieve improved acceptable returns. Comments are very good. The PV web application is about $275 a year, IIRC, and seems to be much easier that building Excel models and using d/l suspect data. Click on the author's name for links to his other articles and model suggestions. At least it will generate some interesting thoughts.https://seekingalpha.com/article/4355636-vanguard-sector-etf...Vanguard Sector ETFs Portfolio: Enhanced Momentum Strategies
Best Of |
Favorites & Replies |
Start a New Board |
My Fool |