SELL signal still active. S2EW analysis back to 1831!!!

(Update Thu 8/12/2010 12:45AM):
I have not spent much time using my new RSI formula below the daily level. But, for grins, I looked at the 5min-60min charts tonight. Based purely on S2EW RSI, the move down from 1129 has not proven itself as an impulse yet. In fact, SPX currently looks like an ABC or 121 and it will only look like a full 5-wave impulse on the 5,10 and 15-min charts if SPX can trade down for most of the day Thursday without any large bounces. If a large bounce starts by early afternoon or SPX grinds and stalls without a new low, then the odds will probably favor the ABC or 121 count. Given a likely completed count into 1129, I would favor the 121 in that scenario with a possible test of 1107ish.

Using FIBBEWIE (see Terminology) and assuming the larger bear count, a wave 3-4-5 down from 1220 should target 810-860 in 3-5 months with wave 3 pausing at 850-950 in 5-10 weeks. Wave (1) of 3 should be in the 80-120 pt range with [1] of (1) of 3 being 30-50 pts. So far, 1129-1089=40 pts meaning the first wave (or 1-2-1) is about to end. I'd expect 1010-1040 to be tested twice over the next 2 weeks followed by a larger bounce through Labor Day before the real fun begins.

Let's see if my System and S2EW confirms that scenario day by day. The System stop has dropped to 1093 + 1pt/15min with final 1/4 profit-taking at 1080 if we get there. And another 20dSMA-test signal should come in the next 1-3 days. Good luck.

____________________________________________
(Update Wed 8/11/2010 1:00PM):
1/2 profit was taken at 1104 shortly after the open. Due to the 50dSMA, 1090 pivot and intraday congestion, I went back to the original plan of taking another 1/4 profit at a 2% gain at 1091. SPX is still at that price as I type. I will be aggressive finding a place to add back that last 1/4 position while more conservative to add the 1/2 position back. Officially, the System has now declared a neutral daily trend due to the 20dSMA test, but gaps beyond the 20dSMA are usually continued so I suspect the System will give another fully loaded SELL signal within the next couple days. But, SPX will need another large gap down or a close byond Friday's (tomorrow) high/low to set the next daily trend. Good luck.
____________________________________________
(Update Wed 8/11/2010 8:20AM):
It looks like SPX will test the 20dSMA this morning. In my experience, the market is speaking when it gaps strongly beyond a key s/r level like the 20dSMA (almost 1105 at the open) and 1107. Work the plan which is to sell half of our short at the 20dSMA and let the rest ride until resistance is broken or we get a new signal. I think there is a good chance SPX will continue down but, if so, I suspect the 20dSMA will be backtested from below, so there will likely be opportunities to reload our short in a better risk/reward scenario. Good luck.
____________________________________________________
Active SPX signal: SELL signal from 1113 on daily trend turn down on morning of Friday 8/6. 2nd chance opportunity occurred around 1121 on Monday when hourly stoch crossed down from above 50 while below resistance.
Hourly trend: Down with resistance at 1129.24
Daily trend: Down with resistance at 1129.24 and support at the 20dSMA
Last signal: Win. BUY signal at 1115 on Monday 8/2 gap up with no real drawdown. Profit was taken 1/2 at 1121, 1/4 at 1126 and 1/4 at 1121. Total gain .7%. 17 wins, 2 losses, 3 draws since 4/26.
Exit levels: Stop recommended at 1133 (& 15min rule). Profit-taking recommended 1/2 at the 20dSMA and let the rest ride until hourly resistance is broken per rules.

System Notes:
Our system stop was pushed to the limit on Monday and our profit-taking spot has risen as it is the 20dSMA.
 
Opinion:
Although another high above 1131 is still possible, the technicals are looking horrible to me. Due to all the price-volume at 1107-1130, I think our System stop at 1133 is smart. I don't think the market can get much past the 1140s-1150s but, even though a head fake is possible, it is likelier SPX would stay for at least a few days above 1130 if it can close there so it's safer to get out there as a lot of people will do. You can always get back in short later when a new signal occurs.
 
S2EW analysis back to 1831!!!
Tonight, I'm just going to give a summary of my findings thus far, and I'll fill in the details over the next couple weeks.
 
First, let me explain what I found regarding RSI-based wave analysis which I kinda grew into and never saw taught anywhere. See my S2EW link above for rule details which won't change much. Most people agree that wave 3s are typically the strongest impulse waves in terms of price momentum and often volume too. RSI is a solid price momentum indicator which you can study on many other sites including the Free Chart school at stockcharts.com. There are numerous RSI trading techniques, but I'm combining RSI in a unique way with the most fundamental EW rules to determine a count and hopefully better predict the future. I determined that the RSI period (14 is most typically used) needs to be adjusted mathematically based on the number of periods being studied. Many people naturally know this since they use RSI14 or RSI5 (like I did) on an hourly, daily or weekly chart depending on the number of trading days involved. However, most people blindly use only 1 or 2 RSI periods for all situations.
 
So, what RSI period should you use? If you think about things logically, in an ideal EW impulse, wave 1 would take about 1/5 of the time for the entire impulse. However, we all know wave 1s are often shorter so they are likeliest to take 1/5th to 1/10th of the entire impulse's time with some outliers of course. So, you will never capture the full momentum of wave 1s consistently unless you are using an RSI period at least 1/5th to 1/10th the length of the wave structure being studied. My whole theory is to identify impulse waves since that's half the puzzle and corrective waves are expected to show complex indications with an endless variety of patterns. If a wave structure does not show impulsive behavior, it must be a correciton by default. Through trial and error, I determined that 1/7th to 1/8th of a studied time period should be used as the RSI period. That applies to all time frames: hourly, daily, weekly, monthly etc. Of course, sometimes RSI meets or breaks one of my S2EW rules by a hair in which case I can see if other RSI periods in the 1/5th to 1/10th range enlighten me but preferably I can just look at a chart on one lower or higher timescale and adjust the RSI accordingly to see if the picture is clearer. For instance, if I were studying a 75-week period (similar to October 2007 to March 2009), I would use RSI 10 (75/7.5) on the weekly chart or RSI 47 (353 trading days /7.5) on the daily chart. However, I also found that a 15-20 RSI is ideal for study with RSI 5-30 being OK too, and I think that occurs due to the smoothing mechanism built into RSI. A tiny RSI period gets too volatile and a huge RSI smooths the volatility too much, and we are trying to measure waves. So, weekly RSI 10 is best for the previous bear market while daily RSI 47 (or RSI 35=353/10) would be a secondary choice.
 
Now, to truly find the wave count, you can't assume 2 pivots start and end a full wave structure. There are flats, triangles, diagonals and otehr formations which make that difficult and to some extent you introduce wave bias. So, one has to study various pivot combinations to see where impulse waves and corrective waves become obvious by S2EW rules, and each pivot combination requires different RSI periods. I built all that into a spreadsheet so I can quickly pick dates and crunch the RSI, charts etc. For instance, did the 1932 low lead to a completed wave structure in 1961, 1966, 1974 or ??? I studied them all, and I did this intensively for 1932-2010 and loosely back to 1831 using Dow data back to 1896 and other sources to get stock market prices and inflation back to 1831. I will post charts soon and I still reserve the right to change my mind with more study (I'm thinking of incorporating volume and possibly other indicators in my rulesa nd formulas), but here is my wave count.
 
1929 - wave 1 of some degree
1932 - wave 2 - by any measurement, the 1929 high and 1932 low were extremes in the data going back to 1831. The retracement from nearly 400 to 41 nearly matched the 40.62 price level that Charles Dow introduced the DJIA at in 1896 and it broke the 1903 low at 42 and, from what I could gather, retraced 90% of the entire stock market back to 1839 (or Curb Market as it was called back then). There is almost no way that can be a wave 4 especially since RSI made a higher high on the way to year 1974 and then blasted it to smithereens on the way to year 2000. Apparently, the Philadelphia stock market began in 1790 with NYSE and others after that, but data is limited and it's difficult to know where to start the wave count. I did notice some pretty strong evidence of a 40-year cycle. Significant stock market lows occurred in 1839, 1857, 1896, 1932 1974 and 2014??? The first 2 dates were separated by nearly 20 years (a half 40-year cycle) and the rest were all 38-42 years apart. So far, 3/9/2009 minus 12/9/1974 is only 34.3 years. Hmmm.
1966 - wave (1) of lower degree - I know many people claim 1932-2000/2007 is 5 waves with 1942 being the internal wave 2 low. Well, almost every chart I produced showed the 1960s RSI surpassing year 1929 RSI and 1974 RSI falling below 1942 RSI. A wave 4's RSI can sometimes fall a little below wave 2 RSI in diagonal triangle situations or extended 5ths, but neither of those things are on the radar for 1932-1974. When I adjusted my charts for inflation and percentage-based RSI (instead of point-based RSI to account for the decades in between waves), the 1974 low became even more notable with RSI dropping below the incredibly low level seen in 1932. The 1974 low retraced 45% of 1932-->1966 which is suitable for a wave 2 especially one that is just before the heart of wave 3 as I count it.
1974 - wave (2) - this wave took 8 years in a 3-3-5 flat with 8 more years to 1982 before the runup began to break the trading range.
2007 - wave (3) - by any measurement or chart, RSI into 1998-2000 was the highest in stock market history going back to 1831. Depending on the time frame, RSI reached 90-99 percent meaning it will be nearly impossible for the stock market to ever beat and 1990-2000 (especially 1994-2000) probably marked the sweetest bull market spot in US stock market trading for a century or more before and after.
2009 - wave A of (4) - The 40-year cycle typically leads to a lot more damage than 17 months. The corresponding wave 2 in 1966-1974 took 8 years. This wave 4 should take at least 4 years and possibly 8-12. RSI in 2009 broke below RSI in 1987 and 1982 meaning those wave structures are done. However, 2009 RSI did not drop below 1974 RSI or 1932 RSI, although it got close, which makes them perfect wave 2s. My RSI technique tells me 2007-2009 was 3 waves, not 5 like I was counting it before, although it does not eliminate the 5-wave count. The jury is still out and I found a nice channel for either count, but I must stick with my RSI formula which narrowly prefers it as 3 waves as Tony at OEW counts it. 
2010 - wave B of (4) - RSI tells me March 2009 to January 2010 was wave "a", Feb 2010 wave "b" and April 2010 wave "c" with nearly c=a*.38 Fib. I suspect alternation will occur with 1974, so we won't see a flat meaning SPX has gone plenty far at a near-perfect Fib 61.8% retrace of October 2007. We could see a zigzag or triangle or wedge. Most people say we can't have a zigzag since we're alternating with the Great Depression but we're actually alternating with 1974. Given the amount of study I've done on historical economic charts, demographics and cycles, I don't think a quick 1-2 year zigzag C would be sufficient time for a correction even though it is possible and even though I think the US economy has another large leg down coming. More technically, given my count, I don't expect RSI to drop below the 1974 RSI. That leaves a little room but not much, so momentum can't get too crazy on the downside below SPX 667. So, putting those things together, I think a wedge would be perfect.

Wave A of (4) set a lower low below the previous degree wave 4 from year 2002 in 17 months, Wave B retraced 61% in 13 months, Wave C will likely take 8-12 months (to Dec-Apr) probably in the form of a zigzag to a lower low with subsequent wave D and E (or XABC) to a sequence of lower lows into 2012 or later with economic malaise into 2016-2020. If wave C is an 8-12 month zigzag as I suspect and we're currently completing 1-2 of a of C from 1220, then 3-4-5 are likely to finish in Oct/Nov with the remaining "b-c" waves shorter but running into early 2011. An ideal wave C of (4) target would be SPX 658 (wave A*.618=909*.618) or slightly lower with wave D then retracing 50-62% back to 1000ish.
 
I do have some caveats and charts I want to talk about, but that will have to come later. After doing this, I'm actually cheered up more about my retirement future. There will be a new 20-30+ year bull market once the economy bottoms in the 2012-2016 time period with a tough start like 1932-1942 and 1974-1982. The next bull market should struggle once China and India hit their demographic spending peaks (currently estimated at 2035 and 2050 respectively). Maybe 2035ish and 2050ish will be wave 4s to match the wave 2s in 1929 and earlier. Once those waves are done, Japan/China/US/India/Europe with basically 90% of the earth's population and spending power will have all peaked over the 1990-2050 period and I suspect that's when the real shit will hit the fan until a new technology or world war or Africa emerges. Another 40 year bottom should occur around 2054 so you've got plenty of prep time for that one if you can buckle down and survive this one. Good luck.

Here's one sample chart so you can get an idea of what I was studying...