Wednesday, February 9, 2011

Dow Jones Industrials: Up 8 Straight Days and Extended, Not So Bearish


The Dow Jones Industrial Average may be somewhat obsolete in today's trading world. These are big, boring companies that usually move too slow (except around earnings) to entice traders on a regular basis. But when it comes to a broad market gauge, all eyes turn to the Dow.  It has been on a tear lately, prompting some (me included) to call it 'overextended' and 'due for a pullback'. Let's look.

Conditions

RSI(2)>99 ---> Normally means short term overbought
Closed above Bollinger Band (50,2) for 2 consecutive days ---> Perceived as overbought
Up 8 straight days ---> Overbought according to most people

So we have 3 'overbought' events hitting the Dow on 2-9-11. This is an interesting confluence and rather rare, with the last occurrence in 1996. So I decided to test the entire the entire history of the Dow Jones to get enough samples [data from Premiumdata.net].  I realize that the market in 1904 vs 2011 is like comparing George Mikan to Dwight Howard. But bear with me, this is for fun.


Buy $DJI on the 3 conditions above; sell 'n' days later. $10,000 per trade; 1904 to present.
Exit 'n'
 days
Net Profit
# Trades
# of winners
% of Winners
Max. Trade % DD
Avg % P/L
W. Avg. Profit
L. Avg. Loss
Profit Factor
Payoff Ratio
1
         (323.67)
60
26
43.33
-1.69
-0.05
        51.05
         (48.56)
0.80
1.05
2
            646.49
49
26
53.06
-3.07
0.13
        88.58
         (72.03)
1.39
1.23
3
            953.53
48
26
54.17
-4.08
0.20
        98.51
         (73.08)
1.59
1.35
4
        1,526.35
47
32
68.09
-4.08
0.32
        97.47
       (106.19)
1.96
0.92
5
        1,885.20
46
29
63.04
-4.08
0.41
      110.72
         (77.97)
2.42
1.42



Doesn't look too bearish. Some weakness in the first couple days, but then it turns, gulp, bullish. So maybe we go down a bit tomorrow, but things look brighter after the first day.

Looking at an intermediate time frame of 3 months:


Buy $DJI on the 3 conditions above; sell 'n' weeks later. $10,000 per trade; 1904 to present.
Exit 'n'
 weeks
Net Profit
# Trades
# of winners
% of Winners
Max. Trade % DD
Avg % P/L
W. Avg. Profit
L. Avg. Loss
Profit Factor
Payoff Ratio
1
              1,885
46
29
63.04
-4.08
0.41
            111
               (78)
2.42
1.42
2
              3,756
46
30
65.22
-4.20
0.82
            200
             (140)
2.68
1.43
3
              4,429
45
29
64.44
-7.05
0.98
            244
             (166)
2.67
1.47
4
              5,743
45
34
75.56
-9.16
1.28
            281
             (346)
2.51
0.81
5
              5,850
45
30
66.67
-11.02
1.30
            349
             (309)
2.26
1.13
6
              5,914
44
27
61.36
-11.02
1.34
            379
             (254)
2.37
1.49
7
              6,891
44
27
61.36
-11.80
1.57
            419
             (260)
2.56
1.61
8
              9,342
43
32
74.42
-10.02
2.17
            404
             (327)
3.60
1.24
9
            12,867
43
33
76.74
-10.02
2.99
            515
             (413)
4.11
1.25
10
            15,110
42
32
76.19
-10.02
3.60
            602
             (415)
4.64
1.45
11
            16,903
40
29
72.5
-10.82
4.23
            704
             (319)
5.82
2.21
12
            16,370
39
30
76.92
-10.83
4.20
            650
             (347)
6.24
1.87


This looks bullish to me. In my own research, I've realized that when the indexes (DJI, SPX, COMP) go up for 7,8,9+ days in  a row, it is a sign of a very strong market. When we pierce the upper levels of a bollinger band (and other metrics), it may mean a slight pause (2-3 days) before we continue to trade higher.

I was also a little weary about the 100 year old data, so I tested the 8 straight up days and BB(50,2) from 1990. The results are nearly identical: slight potential for a  pullback in the first couple days, then up again.

Caveat: if the Dow was climaxing (up like 3+%) today, then I would re-evaluate these results. But it's not, so no need to concern with that.

Conclusion

The Dow is boring.
Perceived 'overboughtness' [multiple up days, upper BB piercings) is actually a bullish sign in an intermediate time frame.  There are decent odds for short term pullback, but bullish after a few days.

Cisco was down today in the after hours on earnings, so that may be the excuse to pull back tomorrow.  But it has a small weight on the Dow. No POMO tomorrow too (I think).

Remember, do your own research and consider all relevant factors.

Good trading out there.
----------------------------------------------------------------------------------------------------------------------
Random Facts

 # of times DJI has risen 'x' consecutive days:

 8 days - 102 occurrences
 9 days -   46 occurrences
10 days -  21 occurrences
11 days -   8 occurrences
12 days -   2 occurrences
13 days -   1 occurrence, 1/20/1987

2 comments:

Anonymous said...

hi Chris,

Was it hard to program amibroker with the formulas for stockbee breakout scan, thanks hector

Chris said...

Hector - It was not too difficult. He has them for TC2000, so I just converted them to Amibroker language. Took some time, but not a lot.

Post a Comment