StockFetcher Forums · Filter Exchange · PORTFOLIO SELECTION AND MANAGEMENT USING RISK/REWARD RATIOS<< 1 ... 13 14 15 16 17 ... 65 >>Post Follow-up
gmoney
8 posts
msg #94419
Ignore gmoney
modified
6/29/2010 11:42:04 AM

That was easy once I found the backtest screen that you use at ETFReplay.com. It is simple, indeed. Thanks! And nice work on your adaptation of the Connors filter.

wkloss
230 posts
msg #94421
Ignore wkloss
6/29/2010 1:43:26 PM

Kevin,

Re:Your post on 6-27

What was the time period that your annualized returns cover?

Bill

Kevin_in_GA
4,546 posts
msg #94424
Ignore Kevin_in_GA
6/29/2010 2:35:47 PM

From whenever each ETF started trading (not all started in 2003) until the close on the day of the post.

cwn6161
40 posts
msg #94429
Ignore cwn6161
6/29/2010 4:38:55 PM

With the month's end tomorrow, I'm going to start using the 6 ETF method in my Roth. Its small enough that extra volatility doesn't really bother me.

wkloss
230 posts
msg #94450
Ignore wkloss
6/30/2010 2:28:10 PM

Kevin,

I just read every post on this and must be brain dead. There is something I'm not getting. Monthly rebalancing? If you are just in 1 ETF at a time and are fully invested, what are you rebalancing from/to? Is "rebalancing" the term used to describe running the filter at the end of each month? (I associate rebalancing with the concept of having multiple holdings and trimming the size of the best or better ones and applying the cash to the lesser performing ones.)

The yield on this strategy makes me wonder if it should be considered for non-retirement accounts or a retirement account that allows leverage using options, perhaps LEAPS. Have you given this any consideration? If you consider the small amount of time your strategy requires, it would take an excellent yield to justify the many labor hours required to trade a system that trades often. I think that is an aspect of "yield" that is often overlooked.

Bill


cwn6161
40 posts
msg #94451
Ignore cwn6161
modified
6/30/2010 3:20:14 PM

Rebalancing is simply running the filter at the end of the month and if the ETF has changed, you move your funds from the previous month's ETF to the new one signalled by the filter. Even if you are 100% invested into an ETF, you technically are "rebalancing" that amount to another ETF. But yeah, it's that simple.



Kevin I've got a question for you. You say you use this system in your 401k - do you have any limitations in terms of excessive trading? I have Fidelity, and they limit 401k trading to 30 day round trips. This system could still work well for me, if I:

1) rebalance quarterly or

2) rebalance every 30 days as close to the end of the month as I can, with sometimes holding an ETF for an extra month to avoid trading in the middle of the month.

gmoney
8 posts
msg #94452
Ignore gmoney
modified
6/30/2010 4:47:10 PM

Kevin and Fellow Members,

I signed up and am studying the www.ETFReplay.com, website. I have constructed an interesting ETF portfolio I'd like to share with all, along with some questions to you and/or the members for responses & comments.

1. My paper beta-test "Monthly ETF Trade B" portfolio consists of: EEM, EUO, FXE, IWM, RWM, SH, SPY, UDN, UUP
2. The model, per my setup shown below, selects the best ETF of those contained within this portfolio for a Buy & Hold for the period specific in the Update Schedule (i.e., one month, unless it remains in the top buy sort in subsequent months).
3. I ran this portfolio through the ETFReplay 'Portfolio Relative Strength" backtest with the following settings:

Buy Top: 1 ETF
Benchmark: SPY
Update Schedule: Monthly
....................Time Period......Weight
ReturnA..... 10-Days............. 50%
ReturnB..... 20-Days............. 30%
Volatility..... 20-Days............. 20%
Backtest Start: .....2003

This backtest provided an astounding Total Return% of 867.2% vs 36.5% total return% from SPY (benchmark) from 2003 through 29 June 2010. And this portfolio and setup significantly exceeded the performance of SPY each and every year from 2003 through the present, although at a somewhat higher volatility of 27.0% vs 21.5% for SPY benchmark. I've tested many permutations of this portfolio, such as buy the top 3 ETFs, or Buy top 1 ETF and use 100% weighting on ReturnA with 0% on ReturnB and 0% on Volatility, etc.. with lower but outstanding results and lower volatility percentages than SPY.

As in your filter, Kevin, the ETFReplay backtest model selects the top ETF each month to buy and hold for 30 days with these settings. However changing/replacing ANY of the portfolio's ETFs, Time Periods or Weights impacts the Total Return% in minor or major ways, in some cases, changing the selection of the top ETF. I also discovered some significant differences between total return% and the actual percentage gains or losses shown on ETF price charts for any given month.

I then took your long version filter and loaded the symlist with the beta-test ETF group shown above. Running your filter, I get different "Buy Top" selections using your model vs the selections produced by ETFReplay. I am concerned due, perhaps, to the ETFReplay model's sensitivity to changes, duplication of results produced by the ETFReplay model in your filter model is challenging.

Questions: Have you seen these same issues? What could be the reasons why? Have I missed something?

Question: What can be done to more closely correlate the ETFReplay backtest results and rank order to align the results of both models?

The benefit of ETFReplay, in my view, is to be able to experiment with different groups of ETFs to test, optimize and then be able to validate the "actual" performance of each ETF group's past performance, and implement it to maximize the probability of achieving similar future performance. (caveat noted: although impressive, past performance is no guarantee of future results...).

In summary, I would really, really like to use ETFReplay to develop/test for optimum portfolio ETF candidate combinations and then implement and manage the ETF groups in my trading accounts using Kevin's excellent filter to provide me with rebalancing updates once per month.







Kevin_in_GA
4,546 posts
msg #94456
Ignore Kevin_in_GA
6/30/2010 9:56:05 PM

I think I have explained this before - ETFReplay uses a ranking system - 1st, 2nd, and 3rd - for each of the criteria you use (highest returnA, highest returnB, lowest volatility). They them multiply each ranking times your weighting factor, sum them, and take the lowest overall score.

My filter simply uses 100% weighting on a 3 month return. No volatility input, although I do calculate it in the more complex filter that I posted.

You are right in sying that slightly different timings (last day of one month versus the first day of the next month) result in different ETFs being selected.

I think the way around that is to use weekly data, and just calculate the RS for the past 13 weeks. To be even more balanced, one should look back at the returns if you used week 1, week 2, week 3, or week 4 as your rebalancing week - take the average of those, and it will give you a more accurate prediction of what return should be expected for a given RS time frame.

alf44
2,025 posts
msg #94457
Ignore alf44
modified
6/30/2010 10:39:20 PM

...this is without question the "nerdiest" ... thread ... EVER created on SF !!!

Reading this BS is much better than AMBIEN !!!

Jeeez !!!


ZZZZZZZZZZZZZZZZZZZ !!!


gmoney
8 posts
msg #94459
Ignore gmoney
6/30/2010 11:32:47 PM

Kevin,

Thanks Kevin. I now see how your prior ranking explanation and current post address my questions and provides me with simple adjustments to begin trading this portfolio.

End of month data posted after the market close today shows an 878.3% total return% since 2003.

Simply a new SF member trying to learn and get it right. And thanks to Kevin's filter and helpful responses, was able to share a very attractive ETF portfolio approach with other "interested" SF members.




StockFetcher Forums · Filter Exchange · PORTFOLIO SELECTION AND MANAGEMENT USING RISK/REWARD RATIOS<< 1 ... 13 14 15 16 17 ... 65 >>Post Follow-up

*** Disclaimer *** StockFetcher.com does not endorse or suggest any of the securities which are returned in any of the searches or filters. They are provided purely for informational and research purposes. StockFetcher.com does not recommend particular securities. StockFetcher.com, Vestyl Software, L.L.C. and involved content providers shall not be liable for any errors or delays in the content, or for any actions taken based on the content.


Copyright 2016 - Vestyl Software L.L.C.Terms of Service | License | Questions or comments? Contact Us
EOD Data sources: DDFPlus & CSI Data Quotes delayed during active market hours. Delay times are at least 15 mins for NASDAQ, 20 mins for NYSE and Amex. Delayed intraday data provided by DDFPlus