RSI2 Overbought + New High = Short Term sell.

Over the past month, by far, i have posted more buy signals than sell. Recently (over the past couple of days) i have been seeing more and more short term sell signals crop up. The first set of such signals was discussed Sunday night and remain on the board tonight. I of the belief that any decline that results from these signals may be buying opportunities, but none the less, the fact that they are showing up in clusters does indicate the market may need a pause. I have one more short term sell signal to add to the ones already open tonight.

S&P 500 RSI(2) closed at 97.35, just under the 98 i usually scan for. However, i decided to run a study to see what the odds look like when the RSI(2) closed above 97 and the S&P has made a new high. There have been 17 previous similar setups. In 15 cases (or 88%) the S&P was trading lower within 3 days.

(Note: column '3Day+' displays the number of days it took the S&P to close higher within the next 3 days. a value of 0 indicates the market did not close lower within 3 days)

Ticker	Date/Time	Close	3Day-	%Chg N1	%Chg N2	%Chg N3
SP500	4/12/1999	1,358.63	1.00	-0.65	-2.22	-2.63
SP500	7/1/1999	1,380.96	0.00	0.74	0.52	1.08
SP500	12/23/1999	1,458.34	1.00	-0.09	-0.05	0.35
SP500	12/31/1999	1,469.25	1.00	-0.95	-4.75	-4.57
SP500	3/23/2000	1,527.35	2.00	0.01	-0.23	-1.28
SP500	9/2/2003	1,021.99	3.00	0.42	0.59	-0.06
SP500	12/1/2003	1,070.12	1.00	-0.33	-0.50	-0.04
SP500	12/31/2003	1,111.92	1.00	-0.31	0.93	1.06
SP500	1/8/2004	1,131.92	1.00	-0.89	-0.41	-0.95
SP500	11/12/2004	1,184.17	1.00	-0.03	-0.74	-0.19
SP500	7/14/2005	1,226.50	2.00	0.12	-0.44	0.23
SP500	11/21/2005	1,254.85	0.00	0.51	0.86	1.07
SP500	1/9/2006	1,290.15	1.00	-0.04	0.31	-0.32
SP500	3/16/2006	1,305.33	2.00	0.15	-0.02	-0.62
SP500	10/24/2006	1,377.38	3.00	0.35	0.85	-0.00
SP500	11/17/2006	1,401.20	1.00	-0.05	0.11	0.35
SP500	1/16/2007	1,431.90	1.00	-0.09	-0.39	-0.10
SP500	10/12/2009	1,076.19	0.00	0.00	0.00	0.00

This suggests a close below 1076 at some point over the next 3 days.

RSI(2) closes above 98

Another short term sell was triggered today. S&P RSI(2) closed above 98. In the last decade, this has occurred 26 times. In 22 cases (or 84.6% of the time), the market closed lower than the trigger date within the next 3 days.

Ticker	Date/Time	Close	3Day-	%Chg N1	%Chg N2	%Chg N3
SPX	4/12/1999	1,358.63	1.00	-0.65	-2.22	-2.63
SPX	7/2/1999	1,391.22	1.00	-0.22	0.33	0.23
SPX	12/23/1999	1,458.34	1.00	-0.09	-0.05	0.35
SPX	12/31/1999	1,469.25	1.00	-0.95	-4.75	-4.57
SPX	7/14/2000	1,509.98	2.00	0.03	-1.08	-1.86
SPX	8/8/2000	1,482.80	1.00	-0.67	-1.52	-0.74
SPX	4/19/2001	1,253.69	1.00	-0.85	-2.34	-3.53
SPX	5/21/2001	1,312.83	1.00	-0.26	-1.81	-1.50
SPX	3/19/2003	874.02	3.00	0.19	2.49	-1.12
SPX	6/4/2003	986.24	3.00	0.40	0.15	-1.05
SPX	9/2/2003	1,021.99	3.00	0.42	0.59	-0.06
SPX	4/5/2004	1,150.57	1.00	-0.21	-0.87	-0.98
SPX	5/25/2004	1,113.05	0.00	0.17	0.74	0.69
SPX	11/3/2004	1,143.20	0.00	1.62	2.01	1.90
SPX	6/17/2005	1,216.96	1.00	-0.07	-0.28	-0.25
SPX	7/15/2005	1,227.92	1.00	-0.55	0.12	0.59
SPX	9/30/2005	1,228.81	1.00	-0.17	-1.17	-2.64
SPX	11/22/2005	1,261.23	3.00	0.35	0.56	-0.30
SPX	3/17/2006	1,307.25	1.00	-0.17	-0.77	-0.17
SPX	10/25/2006	1,382.22	2.00	0.50	-0.35	-0.31
SPX	4/10/2007	1,448.39	1.00	-0.66	-0.04	0.31
SPX	1/2/2009	931.80	1.00	-0.47	0.31	-2.70
SPX	5/4/2009	907.24	1.00	-0.38	1.35	0.02
SPX	7/16/2009	940.74	1.00	-0.04	1.10	1.47
SPX	7/20/2009	951.13	0.00	0.36	0.31	2.65
SPX	7/23/2009	976.29	0.00	0.30	0.60	0.34

So currently we have two short term sell signals on the board:

• Seven consecutive higher highs on S&P argues for a lower close than 1049.34 by Thursday (not very likely at this point)
• RSI(2) closes above 98 argues for a lower close than 1068.76 by Friday.

RSI(2) and VIX

Interesting day. We had two high probability buys triggered today to add the slew of buys from yesterday. Taken together these two signals, the S&P 500 RSI(2) closing below 2 and the VIX closing above it's bollinger bands for the second consecutive day, shows some very favorable odds. This has only occurred 8 times. While not statistically significant, each time marked a short and intermediate term bottom. The table below show that in every case, the market was up within 3 days. Notable as well, the market was up 20 trading days later in 7 out of the 8 cases.

(Note: column '3Day+' displays the number of days it took the S&P to close higher within the next 3 days. a value of 0 indicates the market did not close lower within 3 days)

Ticker	Date/Time	Close	3Day+	%Chg N1	%Chg N2	%Chg N3	%Chg N20
SPX	8/6/1990	334.43	1.00	0.12	1.17	1.65	-3.39
SPX	1/8/1991	314.90	3.00	-1.08	-0.12	0.10	11.55
SPX	9/7/2001	1,085.78	1.00	0.62	-4.33	-4.88	1.07
SPX	4/29/2002	1,065.45	1.00	1.08	1.97	1.79	0.85
SPX	7/23/2002	797.70	1.00	5.73	5.14	6.91	17.52
SPX	3/11/2004	1,106.78	1.00	1.25	-0.21	0.35	2.94
SPX	8/6/2004	1,063.97	1.00	0.12	1.42	1.11	4.67
SPX	6/13/2006	1,223.69	1.00	0.52	2.65	2.28	2.85
SPX	9/2/2009	994.75	0.00	0.00	0.00	0.00	0.00

When we relax the RSI(2) criteria to look for cases where RSI(2) closed below 10 while VIX closed above it's upper BB(20,2) bands for two consecutive days, we get 36 previous occurrences. In these cases, the market was up 88.9% of the times 3 days later. This is about the same odds had we used the RSI(2) alone. However, where we see better odds by combining these two non-related triggers is looking out 20 days. There, we see that the S&P has historically been up 81% of the time 20 trading days later vs 73% odds with the RSI(2) alone.

Ticker	Date/Time	Close	3Day+	%Chg N1	%Chg N2	%Chg N3	%Chg N20
SPX	8/6/1990	334.43	1.00	0.12	1.17	1.65	-3.39
SPX	1/8/1991	314.90	3.00	-1.08	-0.12	0.10	11.55
SPX	6/18/1992	400.96	1.00	0.68	0.61	0.77	3.66
SPX	10/2/1992	410.47	0.00	-0.71	-0.80	-1.52	2.00
SPX	1/5/1993	434.34	1.00	0.04	-0.83	-1.22	1.89
SPX	2/17/1993	433.30	2.00	-0.32	0.21	0.45	3.46
SPX	9/8/1993	456.65	1.00	0.19	1.11	1.18	0.90
SPX	9/21/1993	452.95	1.00	0.72	1.06	1.03	2.93
SPX	11/4/1993	457.49	1.00	0.45	0.59	0.62	1.62
SPX	3/31/1994	445.77	2.00	-1.54	0.57	0.51	1.63
SPX	6/21/1994	451.34	1.00	0.39	-0.38	-1.89	0.06
SPX	1/10/1996	598.48	1.00	0.70	0.56	0.22	8.60
SPX	7/16/1996	628.37	1.00	0.91	2.42	1.65	5.07
SPX	12/16/1996	720.98	1.00	0.70	1.46	3.44	6.41
SPX	7/24/1998	1,140.80	1.00	0.57	-0.93	-1.37	-5.22
SPX	8/28/1998	1,027.14	0.00	-6.80	-3.20	-3.57	2.10
SPX	9/24/1999	1,277.36	1.00	0.47	0.38	-0.70	1.90
SPX	10/10/2000	1,387.02	0.00	-1.62	-4.13	-0.93	3.23
SPX	2/21/2001	1,255.27	3.00	-0.20	-0.75	0.99	-10.61
SPX	9/7/2001	1,085.78	1.00	0.62	-4.33	-4.88	1.07
SPX	4/29/2002	1,065.45	1.00	1.08	1.97	1.79	0.85
SPX	6/4/2002	1,040.69	1.00	0.88	-1.11	-1.26	-8.90
SPX	7/16/2002	900.94	1.00	0.57	-2.15	-5.90	-1.86
SPX	7/23/2002	797.70	1.00	5.73	5.14	6.91	17.52
SPX	1/27/2003	847.48	1.00	1.31	1.99	-0.34	-1.05
SPX	11/18/2003	1,034.15	1.00	0.80	-0.05	0.11	4.09
SPX	3/11/2004	1,106.78	1.00	1.25	-0.21	0.35	2.94
SPX	5/10/2004	1,087.12	1.00	0.77	0.93	0.86	5.06
SPX	8/6/2004	1,063.97	1.00	0.12	1.42	1.11	4.67
SPX	10/6/2005	1,191.49	1.00	0.37	-0.35	-0.56	2.39
SPX	5/18/2006	1,261.81	1.00	0.41	0.02	-0.41	-0.81
SPX	6/13/2006	1,223.69	1.00	0.52	2.65	2.28	2.85
SPX	3/5/2007	1,374.12	1.00	1.55	1.30	2.02	3.67
SPX	6/7/2007	1,490.72	1.00	1.14	1.23	0.15	2.66
SPX	7/27/2007	1,458.95	1.00	1.03	-0.25	0.47	1.40
SPX	10/7/2008	996.23	0.00	-1.13	-8.66	-9.74	0.96
SPX	9/2/2009	994.75	0.00	0.00	0.00	0.00	0.00

Pretty good odds when compared to the the odds of the S&P closing higher on any given day under the same exit conditions over the same data sample period - those odds are 69.6% for a higher close WITHIN 3 days and 59.92% for a higher close 20 days later.

These studies combined with yesterday's suggest the market has a high probability of closing higher within the next few days and additionally suggests that the market may remain on firm ground looking further out, 1 month later.

If the market continues lower over the next few days despite these occurrences, bears will have made their mark quite strongly.

NASDAQ: Perfect Bullish record looking out 4 weeks basedon RSI signal

Persistently OB RSI is bullish..
When the Nasdaq's RSI(2) has closed above 98 for 7 consecutive days, the Nasdaq has historically always finished higher 4 weeks later. There were 20 prior occurences, all of which closed higher 20 trading days later. In 18 of 20 cases (or 90%) of the time, the Nasdaq was trading higher 10 tradings later. And finally, even for the very short term, the odds are favorable: In 17 of 20 cases the market was trading higher with in the next 3 days.

Streaks..
Nasdaq has closed up for 10 consecutive days. Once more this persistant buying appears to present rather favorable odds for the intermediate term. In 25 of 30 prior instances, the Nasdaq was trading higher 20 days later. Similarly, in 23 of those 30 cases, the market was trading higher 10 days later.


Once more these studies suggest strength begets strength and that we should consider any retrace or dips as buying opportunities.

DOW: Interim term buy based on RSI(2) and higher highs in overbought conditions

Running similar studies as i did on the Nasdaq, but this time on the DOW, reveals the following.

The DOW is up 8 straight days. The short term is not as bullish as what the nasdaq revealed. For instance from 1965 to to 1995, the dow went on to close higher within 3 days in 16 of 18 occurrences, which ofcourse are excellent odds. But if we focus on recent history, say the last 2 decades, we see that in only 8 out of 13 occurrences did the dow close higher within 3 days. Looking out 10 days, or two weeks, still has favorable odds however. In the last two decades, the dow closed higher 11 out of 13 occurrences.

The RSI is also telling a similar story. When the RSI(2) has closed above 98 for 5 consecutive days on the DOW (as it did yesteray), the DOW has closed up 17 out 22 (or 77%) of the time. 20 days out, or 4 weeks later, the win ratio is still good at 16/22.

These types of data seem to suggest that we should be nibbling long on any short term pull back, rather than...well, think too logically and move to cash? not saying that is bad thing - but based on what this data suggests, there are still some rather favorable trades here on the long side.

Evening Summary: RSI(2) sell, Nasdaq100 Streaking...

RSI(2) again...and again...
On July 14th i pointed an RSI(2) sell which argued that the S&P should trade lower within 3 days. That signal is off the board tonight and has failed to produce positive results. This signal used the RSI 90 as a trigger value.
Last night i pointed out once more another RSI(2) sell signal, this time based on consecutive readings above 90. That signal has a 3 day timeframe, so it has until monday to play out. For it to succeed the S&P has to trade below 932.68 tomorrow.

Tonight we have yet another RSI sell, this time using RSI 98 as trigger value. In the past decade, when the RSI(2) has closed above 98, it has generally been a good signal to sell into. Out of 23 occurrences, the market was trading lower 21 times within the next 3 days.

Side Note: I think it is important to keep in mind that in periods of high speculation, Overbought and Oversold conditions can persist for long periods of time. Further more, when the market makes higher highs in overbought, it is usually BULLISH looking further out. Now i would not use a fast reading such as the RSI(2) to guage this type of persistance, but i would use an RSI(5) or higher. Here we are beginning to see the market make higher highs in OB.

Nasdaq 100 Streaks:
The nasdaq 100 closed up for 7 consecutive days while also closing above it's BB(20,2) upper band. Since 1990 there were 18 such occurrences. In 16 cases, the Nasdaq 100 was trading lower within 3 days (or 89% of the time). In fact, in all but one of the case, the Nasdaq 100 closed lower within 2 days. This study implies a likely lower close by monday on the nasdaq.

Evening Summary: Breadth and RSI

More lopsided Breadth...
We ended the day with 9:1 A/D ratio and a whopping 28:1 u/d volume ratio on the NYSE.

This appears to be a slight negative short-term, BUT bullish looking out a 2 to 4 weeks. Only four other times since 1950 have we seen similar readings. While in the short term the market closed lower at some point with in the next three days, in every case the market closed higher 10 days out. Obviously not statistically significant, but given the fact that in every case we were lower within three days and higher by two weeks - it does i think offer a possible trade to consider.

Looking at just occurrences where we have had a 9:1 or greater A/D ratio, we see that again, in the majority of the cases, the market was trading up 2 to 4 weeks later. There were 13 prior instances, and in 11 cases the market was trading higher two weeks later.

RSI(2) again...
Last night an RSI(2) overbought signal was triggered. Obviously, today that signal has been triggered again. Looking back in the past decade for all the times we have had two consecutive RSI(2) sell signals: there were 95 occurrences. In 77 of these cases (or 81%) of the time, the S&P was trading lower 3 days later.