No. of Recommendations: 29
Here is an update on posts that I have combined.

Results first:

End of period	Sector Switch	2 ratio	        SPY ETF
Jun-13	       $673,386 	 $270,384 	 $114,974 
Mar-09	       $409,828 	 $155,009 	 $57,217 
Jan-00	       $100,000 	 $100,000 	 $100,000 


Period  	Sector Switch	2 ratio	        SPY ETF			
 Mar 2009 to current 	64%	  74%	        101%
 2000 to Mar 2009 	310%	  55%	        -43%
 2000 to current 	573%	  170%	         15%

1) the sectors used were the  9 ETF sectors (e.g., XLE, XLU...)

2) weights given:
 
3 month return = 1x
6 month return = 2x
12 mo. return = 1x

3) the ETF with the best returns using the weights in step 2
is chosen.

4) be in the market only when the 2 ratio system says to be in or out
(post # 241906).

From the above results, I have been buying the ETF that does the
best in step 3 and buying the SPY ETF when the 2 ratio says to do so.
Seems like a good combination.

Current picks:
XLV - Health Care (best weighted returns from steps 1-3 above.
and SPY. (2 ratio returns says to be in the market, step 4 above)
Print the post Back To Top
No. of Recommendations: 4
Very interesting. Love the idea of limiting to 9 ETF'S
and developing a timing system.

Thank you for sharing your ideas and hard work back-testing.

Would love for you to keep us abreast using this system monthly.

Chuck
Print the post Back To Top
No. of Recommendations: 1
Thanks for sharing.

Is your plan to go long on 2 etfs (SPY + best sector ETF using criteria) when 2 ratio is bullish, go cash when bear?

It would be interesting to test if the returns are improved with equal-weight sector ETFs.
Print the post Back To Top
No. of Recommendations: 2
Very interesting. Love the idea of limiting to 9 ETF'S and developing a timing system.

I agree. The 10 sector S&P can be tested back using Fama-French data to 1926. Each year, the sector components are evaluated for re-assignment. Very good results using these in a backtest, since they are equal weighted. Plus Guggenheim (formerly Rydex) runs equal weighted ETFs for 9 of these i.e. RHS, RYE, RYH, RYU and others.

In a limited backtest these ETFs actually beat the Fidelity sector fund equivalent slightly.
Print the post Back To Top
No. of Recommendations: 0
Very interesting. Love the idea of limiting to 9 ETF'S and developing a timing system.

I agree. The 10 sector S&P can be tested back using Fama-French data to 1926. Each year, the sector components are evaluated for re-assignment. Very good results using these in a backtest, since they are equal weighted. Plus Guggenheim (formerly Rydex) runs equal weighted ETFs for 9 of these i.e. RHS, RYE, RYH, RYU and others.

In a limited backtest these ETFs actually beat the Fidelity sector fund equivalent slightly.

----------------------------------------------------------------
I find this interesting because I opened an account with Fidelity and transferred money there.

From this post and earlier ones it seem you can do as well with ETFS and not have the trading restrictions, so I may just trade ETFS and not Fidelity Sector Funds.
Print the post Back To Top
No. of Recommendations: 14
I used etfscreen.com to compare the performance of 9 select sector ETFs (XLB,XLE,XLF,XLI,XLK,XLP,XLU,XLV,XLY) and 9 equal weight ETFs (RCD,RGI,RHS,RTM,RYE,RYF,RYH,RYT,RYU). I used tpoto's 2 ratio on SPY for timing market entry/exit- formula used: 1/(1-[Rtn-6mo]/100)>.97 OR 1/(1-[Rtn-1yr]/100)>.97. The best sector from each portfolio is bought based on the criteria- formula used: [Rtn-3mo]+2*[Rtn-6mo]+[Rtn-1yr]. Purchases are made every 5 days. Since I don't have a paid account, I could only compare the last 5 years.



Select sector ETF:
http://www.etfscreen.com/screener.php?sbt=409ad
CAGR: 9.8
SD: 22.2
UI: 14.5
Time invested: 83.2%


Equal weight sector ETF:
http://www.etfscreen.com/screener.php?sbt=409ac
CAGR: 12.2
SD: 19.6
UI: 8.0
Time invested: 83.2%

It appears that select sector etfs have better return. However, it is interesting to note that with more etfs selected each week, equal weight sector ETFs get the upper hand.

For select sector, selecting 3 ETFs (instead of just 1) gives CAGR of 13.7.
For equal weight sector, selection of 5 ETFs (instead of just 1) gives CAGR of 14.8.
Print the post Back To Top
No. of Recommendations: 1
I used etfscreen.com to compare the performance of 9 select sector ETFs and 9 equal weight ETFs
It appears that select sector etfs have better return.


Good job. Except maybe you have the conclusions EqWt/CapWt inverted.

From the chart it looks like most if not all the advantage is due to the timing rule. During the 3 plus years invested the trend is parallel to the index. I prefer Asset rotation to sector rotation.

gdm
Print the post Back To Top
No. of Recommendations: 2
DreadPotato wrote: "Good job. Except maybe you have the conclusions EqWt/CapWt inverted."

=====================



Thank you for pointing out the error. Here is the correction:



Select-sector ETF:
http://www.etfscreen.com/screener.php?sbt=409ac
CAGR: 12.2
SD: 19.6
UI: 8.0
Time invested: 83.2%


Equal-weight sector ETF:
http://www.etfscreen.com/screener.php?sbt=409ad
CAGR: 9.8
SD: 22.2
UI: 14.5
Time invested: 83.2%
Print the post Back To Top
No. of Recommendations: 6
johnnyhabesha-

Thank you for testing out the sectors in etfscreen.com

I want to point out that you referenced my 2 ratio
system, but what you presented
I used tpoto's 2 ratio on SPY for timing market entry/exit- formula used: 1/(1-[Rtn-6mo]/100)>.97 OR 1/(1-[Rtn-1yr]/100)>.97.

is not what I had developed
(the second part of the OR is incorrect)

You might want to reference my original post.
http://boards.fool.com/2-ratios-30565283.aspx?sort=whole

What you presented does work fine for the past few years, but
going back to 1951 for ^gspc, there is a big difference.
(I mention this, in case someone thinks that your formula
is what I use.)
Print the post Back To Top
No. of Recommendations: 2
Comparing Fidelity Sector to Rydex-Guggenheim


Dec-2006 through March-2013
CAGR Blend of Funds Compared
Fidelity Sector 7.3% FDFAX FDENX FSPHX FSUTX
Rydex 8.5% RHS RYE RYH RYU
SPY 3.7%


So active management is not exceeding buy and hold equal weight blend.
Print the post Back To Top
No. of Recommendations: 0
tpoto wrote: "...is not what I had developed (the second part of the OR is incorrect)"

I stand corrected. I apologize for the blunder. This happened in my attempt to convert your equation to etfscreen.com. Despite the error, I think the signals for the two systems should be almost similar for the last 5 years.
Print the post Back To Top
No. of Recommendations: 1
Just noodling around in the etfscreen site:

http://www.etfscreen.com/screener.php?sbt=40a7f

one can run the CAGR up to 15.0 starting in March 2008.

Not that I would do this- 5 XL* sector ETFs, 5 day hold and a
slightly different weighting scheme.

(This is different that what I presented using my excel
solver, one ETF, monthly hold,and a different bear catcher
formula- but it seems to work for the past 5 years.
Too short to get excited about, though...).
Print the post Back To Top
No. of Recommendations: 0
This is different that what I presented using my excel
solver


NEW: DEFINE _UV1 = [Rtn-6mo]+[Rtn-1yr]

Previous: DEFINE _UV1 = [Rtn-3mo]+2*[Rtn-6mo]+[Rtn-1yr]

The final sort looks better (IMO), with less tuning.

gdm
Print the post Back To Top
No. of Recommendations: 0
I'm having trouble developing a weighted formula that gives the same ranking. Using:

Rank Sym 3 Mo 6 Mo 12 Mo Weighted

10 XLB 2.85% 2.37% 14.37% 0.22
7 XLE 3.38% 9.86% 22.70% 0.46
2 XLF 10.25% 17.55% 38.66% 0.84
5 XLI 6.12% 12.22% 25.51% 0.56
9 XLK 3.76% 4.67% 10.07% 0.23
6 XLP 1.85% 13.62% 18.61% 0.48
8 ITB 0.09% 4.62% 32.69% 0.42
11 XLU -2.69% 7.71% 6.64% 0.19
4 XLV 4.22% 19.57% 29.88% 0.73
1 XTN 8.01% 22.11% 35.48% 0.88
3 XLY 8.90% 18.51% 33.15% 0.79
12 SHV 0.01% 0.00% 0.04% 0.00



Are you just numerically adding the 3mo+(2x6mo)+12mo return?
Print the post Back To Top