UnThreaded | Threaded | Whole Thread (8) | Ignore Thread Prev Thread | Next Thread
Author: klouche Big red star, 1000 posts Old School Fool Add to my Favorite Fools Ignore this person (you won't see their posts anymore) Number: of 254112  
Subject: "Seven-Rules-Wall-Street-Crash-Tested" Date: 2/9/2013 10:48 PM
Post New | Post Reply | Reply Later | Create Poll Report this Post | Recommend it!
Recommendations: 30
We've been discussing the book "Seven-Rules-Wall-Street-Crash-Tested" in the "Sector Momentum 84 years" thread.

http://www.amazon.com/Seven-Rules-Wall-Street-Crash-Tested/d...

"Rule 3" the author proposes a twist on the "Sell in May and Go Away" timing method. He suggests that rather than buy the S&P500 index SPY, buy the sectors that do the best for the season. The author presents the Winter (Nov-Apr) and Summer (May-Oct) performance for each sector for 1991 through 2007.

I decided to do an out of sample forward test for 2008-2013 by choosing the three sectors best suited for Winter and for Summer. These choices were based on 1991-2007 and retained throughout the test. The Rydex equal weight ETF prices were used though heavier volume cap weighted ETFs are also available.

These are the sectors, ETFs and seasons used in the 2008-2013 test:

Sector	                Equal	Cap	Season
Consumer Discretionary RCD XLY Winter
Consumer Staples RHS XLP Summer
Energy RYE XLE
Financials RYF XLF Winter
Health Care RYH XLV Summer
Industrials RGI XLI Winter
Information Technology RYT XLK Summer
Materials RTM XLB
Utilities RYU XLU


Here are the results:


SPY Both Winter
CAGR 2.9% 8.9% 11.6%
GSDm 0.206 0.260 0.199
Sharpe(2.5) 0.02 0.25 0.46

MaxDD -49% -51% -35%
Mos to Brk Even 40 27 25
Overall Gain 15% 55% 77%


Both Winter and Summer investing tripled SPY returns. Pretty good for a largely passive method, out of sample. Winter ONLY return was even more impressive since most of the Summer 2008 loss was avoided.

This sector and timing combination may bear more research rather than SPY / RSP alone timing.

KL
Post New | Post Reply | Reply Later | Create Poll Report this Post | Recommend it!
Print the post Back To Top
Author: jakalant Big red star, 1000 posts Old School Fool Add to my Favorite Fools Ignore this person (you won't see their posts anymore) Number: 241515 of 254112
Subject: Re: "Seven-Rules-Wall-Street-Crash-Tested&q Date: 2/9/2013 11:20 PM
Post New | Post Reply | Reply Later | Create Poll Report this Post | Recommend it!
Recommendations: 0
So ? this is an RS 1991 through 2007? = 17 years ? , and that is how those sectors were selected. Is that right?

Print the post Back To Top
Author: klouche Big red star, 1000 posts Old School Fool Add to my Favorite Fools Ignore this person (you won't see their posts anymore) Number: 241516 of 254112
Subject: Re: "Seven-Rules-Wall-Street-Crash-Tested&q Date: 2/9/2013 11:55 PM
Post New | Post Reply | Reply Later | Create Poll Report this Post | Recommend it!
Recommendations: 0
So ? this is an RS 1991 through 2007? = 17 years ? , and that is how those sectors were selected. Is that right?

Yes, simply the average return for each sector for Summer and the average for Winter. Choose the top three and go forward.

I have no idea if the same three (Cons Staples, Health and Info Tech) would be chosen for Summer 2013.

Print the post Back To Top
Author: mungofitch Big gold star, 5000 posts Top Favorite Fools Top Recommended Fools Feste Award Winner! Old School Fool Add to my Favorite Fools Ignore this person (you won't see their posts anymore) Number: 241521 of 254112
Subject: Re: "Seven-Rules-Wall-Street-Crash-Tested&q Date: 2/10/2013 9:28 AM
Post New | Post Reply | Reply Later | Create Poll Report this Post | Recommend it!
Recommendations: 10
How does this sound as an experiment:

Using the 1926-2012 industry return data set from Fama/French.

Start in, say, 1935 or 1940 allowing a lookback to have a bit of priming,
stepping one year at a time to the present.

Use the Sy Harding seasonal autumn-buy and spring-sell signal dates.

On each autumnal buy date, find the N sectors that did best overall in the
winter periods up till then, and this winter hold only the ones (say, 10-15% of industries) that were best in hindsight up to the prior year.

Thus the sectors held in any given good/winter period are not subject to any
hindsight, they are always based only on the returns for all winters prior to that date.

I'd do the testing with the sector returns themselves, to find out
how well strategy is working. Only if it does would I concern myself
about what exact ETF or trading vehicle best corresponds to the
industries that the test recommends. Since you'd have only one
buy and one sell per year, simply buying all the main stocks making up
an industry would be pretty viable, e.g. top 10 equal weight.

The same test could be done at industry or sector level.

Jim

Print the post Back To Top
Author: klouche Big red star, 1000 posts Old School Fool Add to my Favorite Fools Ignore this person (you won't see their posts anymore) Number: 241533 of 254112
Subject: Re: "Seven-Rules-Wall-Street-Crash-Tested&q Date: 2/10/2013 5:21 PM
Post New | Post Reply | Reply Later | Create Poll Report this Post | Recommend it!
Recommendations: 8
More forward testing based on the rules.


Method 08-13 Performance

SPY 15.7%
RSP 31.8%
Blend of 9 EW Sectors 41.0%
Blend Winter 75.4%
Blend Summer -20.2%
Top 3 Sectors Based on Jan -5.7%
Best Sectors past 12 months -11.0%
Best Sectors past 12 months Winter 88.0%
Best Sectors past 12 months Summer -52.7%


All the above are total gains for the 50 month period.

Definition: the SP500 contains nine sectors and 130 industries.

Blend of the 9 equal weight sectors is surprisingly strong. This would be buying, holding and rebalancing each of the nine Rydex equal weight SP500 ETFs. An equal weight of equal weights. Presumably better than RSP since RSP industry member counts are disimilar.

The Winter/Summer disparity is strong during this period.

Top 3 based on January is applying a rule in the book out of sample. The "rule" is, as goes January so goes the year. Pick the best three industries based on January performance and hold February through next January. This is a fail.

Best Sector, past 12 months is classic relative strength. Pick the best sector based on 12 month performance, hold for a month. Not surprisingly, this failed for the period, though the author proposed it to be the most likely success.

No one could guess in January 2008 that sector relative strength would fail going forward, but calendar timed market exposure would succeed.

In any event, according to recent past, holding anything during the Summer is bad reward to risk deal. I would like to see our backtesters display results with these periods separated.

Post New | Post Reply | Reply Later | Create Poll Report this Post | Recommend it!
Print the post Back To Top
Author: TopLineFirst One star, 50 posts Add to my Favorite Fools Ignore this person (you won't see their posts anymore) Number: 241592 of 254112
Subject: Re: "Seven-Rules-Wall-Street-Crash-Tested&q Date: 2/15/2013 5:41 PM
Post New | Post Reply | Reply Later | Create Poll Report this Post | Recommend it!
Recommendations: 0
koluche:

The Guggenheim Russel 1000 Equal Weight ETF (EWRI) seems to follow the “Blend of 9 Sectors” approach. In case the following link fails, see post 241196.

http://boards.fool.com/equal-sector-weight-etfs-30500200.asp...

TLF

Print the post Back To Top
Author: rrjjgg Big red star, 1000 posts Old School Fool Add to my Favorite Fools Ignore this person (you won't see their posts anymore) Number: 241599 of 254112
Subject: Re: "Seven-Rules-Wall-Street-Crash-Tested&q Date: 2/16/2013 7:36 AM
Post New | Post Reply | Reply Later | Create Poll Report this Post | Recommend it!
Recommendations: 0
<In any event, according to recent past, holding anything during the Summer is bad reward to risk deal. I would like to see our backtesters display results with these periods separated.>

I think Gritton's backtester can be used for something like this, since it gives the average monthly return in a column. At least a fixed blend/screen can be tested. Many have much weaker results in the summer, usually August through October.

rrjjgg

Print the post Back To Top
Author: klouche Big red star, 1000 posts Old School Fool Add to my Favorite Fools Ignore this person (you won't see their posts anymore) Number: 241614 of 254112
Subject: Re: "Seven-Rules-Wall-Street-Crash-Tested&q Date: 2/16/2013 5:41 PM
Post New | Post Reply | Reply Later | Create Poll Report this Post | Recommend it!
Recommendations: 0
I think Gritton's backtester can be used for something like this, since it gives the average monthly return in a column. At least a fixed blend/screen can be tested. Many have much weaker results in the summer, usually August through October.

I downloaded a few large cap screens and applied 6 month out of market assumptions.


If screen held November through April only, 0.2% monthly return while in cash.

Screen Return Sharpe Improvement
Blitz -4% 34%
CAP_PLOW -24% 19%
CapRS -11% 41%
HBSP 26% 98%
Hi Yield -23% 1%
Keystone -30% -4%
PLOW26WK -12% 36%
RSCap -20% 24%

Average -12% 31%


Return went down slightly, Sharpe up substantially. Some screens go back to 1969 (CapRS and RSCap). Virtually all return decline is pre 2000. Most of the above imploded 2000-2012. RSCap would have lost you 66% all months 2000-2012 or lost 6.5% Winter months only.

Print the post Back To Top
UnThreaded | Threaded | Whole Thread (8) | Ignore Thread Prev Thread | Next Thread
Advertisement