Market Talk / Nov. 26 - Dec. 2

[Griffin]

Ayla,

This looks like a viable strategy, although I know I personally would have a very tough time staying out of the market for weeks at a time. Did you back test the return this system would have generated?

 

[Birchtree]

We need 1.77 points and we'll have a new all-time high on the NYSE Composite. I just love the smell of manure when the bull is running - great pin action today. Anybody know where the top is - how about you Techy, I can use your help. I think it waits until 2010. Snort. I'm waiting on the ratio adjusted NYAD line to go to new all-time highs - then we go Primary wave 3 for the Elliott folks.

 

[ayla]

Ayla,

This looks like a viable strategy, although I know I personally would have a very tough time staying out of the market for weeks at a time. Did you back test the return this system would have generated?

Apologies if I'm not familiar with the terminology. Not sure what you mean by "back test". The Ifund is real data so doesn't need to be "tested" right? The VIX data is calculated because I have to fit the dates for the holidays. I guess you could say I back tested the VIX data because I checked the first and last and numerous points in between and they jive. This was what I failed to do thoroughly when I posted incorrectlyl the first chart.

Then the averages are just that, averages that are calculated via the Excel "Average" formula. I also have looked at the Yahoo chart for VIX, 20, 50 and 200 and added EFA and went back two years and the results were very close, not as "clear" as my data but still pretty much the same given that the EFA values aren't the same as the Ifund so the scale on the Yahoo graph wasn't the same.

Is that what you mean by "back tested"? I hope to use it with other stocks and funds. Just haven't done that yet. Still looking at this one a bit more to see if there is more it can reveal - such as providing information about the best opportunity to get back on the train after the bottom has been reached... or information about which are fake crossovers, i.e. ones you can ignore so you can stay on the train until the real significant ones happen.

Still learning.

p.s. I downloaded the historical data for the VIX from Yahoo so that is basically "real" data. Then calculated for the holidays (weren't many here). For a holiday when there was VIX data but no TSP Ifund data, the VIX change is calculated based on two days and replaces the VIX real data for the day after the holiday. The VIX datapoint for the holiday is removed. Clear as mud?

 

[SkyPilot]

Backtesting
The process of testing a trading strategy on prior time periods. Instead of applying a strategy for the time period forward, which could take years, a trader can do a simulation of his or her trading strategy on relevant past data in order to gauge the its effectiveness.

Most technical-analysis strategies are tested with this approach.

Notes:
When you backtest a theory, the results achieved are highly dependent on the movements of the tested period. Backtesting a theory assumes that what happens in the past will happen in the future, and this assumption can cause potential risks for the strategy.

For example, say you want to test a strategy based on the notion that Internet IPOs outperform the overall market. If you were to test this strategy during the dotcom boom years in the late 90s, the strategy would outperform the market significantly. However, trying the same strategy after the bubble burst would result in dismal returns. As you'll frequently hear: "past performance does not necessarily guarantee future returns".

See also: Autoregressive, Data Mining, Forecasting, Technical AnalysisSources=Sources | 131072

Attribution()

Powered by Investopedia.com. Copyright © 1999-2005 - All rights reserved. Owned and Operated by Investopedia Inc.

 

[ayla]

Ayla,

This looks like a viable strategy, although I know I personally would have a very tough time staying out of the market for weeks at a time?

After how much I got burned last May by staying in too long, then getting out, then getting back in too soon..lots of whipsawing...I won't have any trouble staying out -- though there is always the F fund to consider.

FYI - Just a superficial view but I'm thinking that when the 50 day VIX average crosses back over the 20 day, may be a very good time to get back in...

 

[ayla]

Backtesting
The process of testing a trading strategy on prior time periods. Instead of applying a strategy for the time period forward, which could take years, a trader can do a simulation of his or her trading strategy on relevant past data in order to gauge the its effectiveness...
As you'll frequently hear: "past performance does not necessarily guarantee future returns". ..

Thanks for the clarification and yes, I totally understand about "past peformance" not necessarily indicating future performance but all I got is past data for now.

And the good news is that this data covered three examples of a downturn over a period of 2 years (though all were not as serious as one in May) and this crossover theory held up pretty well, IMO. Though don't know yet what to do about the crossovers that are "fakes".

Thanks again.

 

[Griffin]

To back test it, determine the dates that generated either a buy or sell signal (when the 20/50 crossover occurs) then get a blank version of TSPtalks returns spreadsheet and fill it in based on the data. Compare the return your system generated to a buy and hold strategy for the same period. You will end up with a result such as "over the twelve month period from A to B, the system return X% compared to a buy and hold of Y%. That will give you some idea of how effective it is.

Then you can start getting tricky with it. See how it performs in a bull market compared to a bear market. Like sky pilot said, you might find that it works very well in a bull market but not in a bear market.

Since reading your posts, I've been starting to monitor the VIX. I'm always looking for a more reliable, faster way to make my decisions. I think your on to something here. I hope you keep at it.

 

[fabijo]

Anybody know where the top is - how about you Techy, I can use your help. I think it waits until 2010.

Birch, I was just looking over and over again at different historical charts for the S&P 500. It's so easy to draw support trend lines to find bottoms, but to find long term resistance seems near impossible. I'm attaching a chart of the S&P 500 with just a couple of lines I've drawn. One line served as resistance from the 1960's until 1995. Once it broke that line, it now is serving as a support line.

My attempt at coming up with some sort of resistance long term brought me to draw a line from the bottom at the beginning of year 1975, through the top of 1987. That line continuing just happened to also be the top for year 2000. If that is the new resistance line, then we still have a long way to go till we hit it. If we shot straight up from here, we'd have an S&P over 2100. The longer it takes us to hit resistance, the higher the resistance number will be.

There are still very large parts of the world needing plenty of development, which means there is plenty of room for economic growth. The world's largest corporations have been meeting together and strategizing on how to profit by making the world a better place. They've created the World Economic Forum and have been meeting in Geneva since 1971. Check out their website at http://www.weforum.org . After reading some of their papers, I believe they are pushing even more for world growth. Look at the Bill and Melinda Gates Foundation. They have billions of dollars to offer as grants. They will help millions of people climb out of poverty. They are educating people on how to start small businesses and to tap into the world economy. The good thing for the world's largest corporations is that they can look good by donating to organizations like the Gates Foundation AND they do not have to do much work. The work of getting people out of poverty is done by the people in it. Once they get themselves out and have a successful business, a large company can purchase the business. Now the large corporation can just keep increasing its market into the developing nations.

Any thoughts?

 

[Spaf]

Daily Yak

The Kingdom of TSP
Daily Edition
November 29, 2006 Closing​

Yak, Doodles, Tea Leaves & The Tin Box

Kingdom Yak:
Pro-Yak....................................Where is the entry point?

Con-Yak...................................Well, throw some $ at the extended socks!

Jester-Yak................................No conclusion yet.

Doodles:
Socks [$SPX] Closed at..............1399.48 up +12.78
Stops......................................Alert: 1393 (reclaimed). Trail: 1380
Trend (MACD-Hist)....................increasing at -1.881.
Overbought/sold (S-STO)...........[80] 42.63 [20] cooling!

Lube (NYM) Closed at.................62.46, up +1.47
Oil Markers...............................<70= ok, 70-75= worry, >75= panic.

Tea Leaves:
Yakndoodles.............................Red.

Tin Box:
TSP........................................Safe; capital preservation.

 

[Birchtree]

If the S&P 500 moves back to trend, it will be above 2600 by the end of 2008. The SPX is currently not at a new high because of its' technology components - but that has started to change in the last 30 days or so. Both financial and energy stocks will lift the SPX not to mention their large degree of export orientation. Large cap growth could be one of the better performing areas of the market for the next few years. Patience is virtuous.

 

[Fivetears]

I'm outside all day working on the house and such, and I come in and see a page and a half of IFT's. Y'all just bought those Lilly Pads the other day. Don't they float?? :nuts:

 

[Birchtree]

Looks like you are holding the RPMs pretty steady - are you accumulating with your DCA also? You know you're gonna take heat.

 

[Fivetears]

Nothing like the Birch DCA system .
I just thought it was funny to a page & a half of IFT's on that bump in the road.
Threw everyone outta the truck bed. :nuts:

 

[ayla]

To back test it, determine the dates that generated either a buy or sell signal (when the 20/50 crossover occurs) then get a blank version of TSPtalks returns spreadsheet and fill it in based on the data. Compare the return your system generated to a buy and hold strategy for the same period. You will end up with a result such as "over the twelve month period from A to B, the system return X% compared to a buy and hold of Y%. That will give you some idea of how effective it is.

Apologies for taking up so many posts here. I will provide any further results/comments in my own "account talk". Guess this stuff is more "technical analysis" than "Market talk". But I'm very happy to report that I did some "backtesting" and got some VERY interesting results IMO.

Turns out that if you use criteria for buying I fund when VIX is below 200 day average and criteria for selling I fund when VIX is above 200 day average, the results are MOST impressive for time periods reported below.

Not sure what intervals are bull or bear. I used these because of data set I already had. I will probably do further backtesting with more data later.

I think the real reason the 200 day average criteria worked the best because it allowed for almost no loss during the major downturn we had starting in May. If there had been no major downturn, results would probably not be so dramatic. My conclusions at this point are to use some hybrid strategy of the 20-80 average criteria plus the 200 day average criteria along with eyeballing other criteria (as always).

Thanks for suggesting the backtest - it has really helped in my evaluation. I think results would probably be very similar if using S Fund purchases (or even C Fund) instead of I fund. Need to check though.. always more to look at...

Acronyms I've used in results below:
B&H = 100 % Buy and Hold I fund, no selling whatsoever during specified time
B&S20-50 = Buying I Fund when VIX 50 day average is above
VIX 20 day average and selling when VIX 20 day
average is above VIX 50 day average
B&S200 = Buying I Fund when VIX is below 200 day average and
selling I Fund when VIX is above 200 day average

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

Results:

for Feb 15, 2005 to Nov 28, 2006
(approx 21 months)
-----------------------------------------
B&H Return = 31.98%
B&S20-50 Return = 33.2%
(note: I think this result mainly shows that "strict enforcement" using the 20-80 averages doesn't give an advantage. If there is a way to use it as a "backup indicator" or "confirmation indicator", I think it is more helpful.)


for Sept 20, 2005 to Nov 28, 2006
(approx 14 months)
---------------------------------------
B&H Return = 24.98%
B&S20-50 Return = 23.2%
B&S200 Return = 39.4%
(note: ditto)


for Jan 3. 2006 to Nov 28, 2006
(11 months - i.e. since January of this year))
-----------------------------------------
B&H Return = 19.3%
B&S20-50 Return = 19.15%
B&S200 Return = 32.26%
(note: here is where avoidance of the downturn in May really would have paid off and this is really what I'm looking for.)

 

[fabijo]

Ayla -

Very nice work. Those would be awesome returns any year.

 

[Pilgrim]

Inflation again!!! I suppose this will sidetrack what was to be a good day.

ECONOMIC REPORT
Consumer spending perks up in October
Core inflation rises 0.2%, a tenth higher than expected


Last Update: 8:32 AM ET Nov 30, 2006


By Rex Nutting, MarketWatch



WASHINGTON (MarketWatch) - Real consumer spending in the United States rose 0.4% in October following two lukewarm months, the Commerce Department reported Thursday.
Core inflation rose slightly faster than expected at 0.2%, keeping the year-over-year gain in the core personal consumption expenditure price index at 2.4%, well above the Federal Reserve's implied cap of 2%.
Economists had been expecting core prices to rise just 0.1%, according to a survey conducted by MarketWatch.

http://www.marketwatch.com/News/Sto...BC65-74AA-4BCD-8E64-3A5E1B8ED731}&siteid=mktw

 

[mlk_man]

Bernanke and three other Fed guys speak tomorrow. Should be interesting......

 

[Birchtree]

The Chicago PMI comes in at below 50. Is there a correlation between the PMI and ISM? Regarding the ISM, at no time in history of the data has the index hit 51.2 in a downtrend without breaking below 50. That level is important, because it is the dividing line between an expansion and a contraction in individual activity. Experience indicates that once the index hits 51.2, it typically slides below 50 in one to three months. Once it does that, history shows that the Fed cuts rates within two months. In other words, if past is a good guide, the Fed will be eading in March 2007 at the latest. The rate cut is back on the table. Snort.

 

[mlk_man]

YEah, I'm hoping this morning was just a knee jerk and Bernanke will comfort people again tomorrow just like he did Tuesday.

 

[Birchtree]

Natural gas is exploding - I like it.