Higher market and lower volume

[nitro]

Quote from Pabst:



My unbacktested twenty year empirical observation. Big volume in a down market is bearish because people puke, indiscriminate of "value". Low volume in an up market is bullish because the public likes to buy on reactions that often don't come. When you finally see big volume on this rally, IMO that's a sign to get short.

I agree. In fact, I have seen evidence time again the last month of momentum dip buyers come in a drive stocks/spoos back up.

The buyers have gotten WAY smarter as to when to apply the squeeze, knowing full well that the recent psychology is on their side.

nitro

 

[DHOHHI]

Quote from Pabst:

Never sell a quiet market. Buyers keep waiting for a concession in prices that doesn't come and volume shows there isn't a lot for sale.

But one can also make the case there aren't a lot of buyers (at the current price) if volume is down. In an up market I see the low volume as suggesting the upward momentum is about to end. Conversely, in a down trending market the light volume would suggest we're at (or near) a bottom.

 

[WinSum]

I'm glad someone else noticed it too. Higher prices on Low volume. Normally, I wouldn't put too much credibility on this type of upward move because low volume allows games to be play. However, I also see a 3rd Dimension that clouds my view. The breadth of stocks that went up on low volume is wide. This confuses me. Sure, a couple of people could pump up a few selected stocks on low volume but they wouldn't try to pump up so many stocks at once. They would need to commit a lot of capital. This would indicate that there is still some leg left to support these higher prices for a while.

Anyone rememeber a time period when these 3 event occurred ? It might illuminate how to interpret this.
1. Low Volume
2. Higher Prices
3. Wide Breadth of Stocks that participated in moving higer

 

[Girlpower]

Aren't reversals usually also characterised with volume spikes? Wouldn't that indicate there would be more to come until a build in volume and a spike (which ties in with your 3rd dimension)?

Natalie

 

[patefern]

Quote from Girlpower:

Aren't reversals usually also characterised with volume spikes? Wouldn't that indicate there would be more to come until a build in volume and a spike (which ties in with your 3rd dimension)?

Natalie

Natalie
That's the same question upside down, ie; a volume spike off the lows(maybe a gap)to start a new trend? (the index's)

Until there is a "blowoff top" with volume shouldn't" there be more to come? Will it be a rollover top? That's what's bad about interruption and also good.

When those who bought lower, and those who bought higher, feel threatened, that's when there will be a change. Maybe technical, maybe fundamental, but probably with volume, just have to wait and see.

 

[spreadem]

It looks like its time to start selling those price spikes.

 

[man]

I am testing the difference between average environment and the one following a day with two characteristics

1. volume below 75% of recent average volume
2. the close was above 75% of the daily range


data: sp future
start: 04.01.1984
end: 13.05.2003

CONDITION
EMAVolDist
upper
-25%
lower
-500%

C/DR
upper
110%
lower
75%

RESULTS
dailyGain
all
0.02%
condition
0.13%

rangeGain1 2 days
all
0.10%
condition
0.26%

rangeGain3 5 days
all
0.21%
condition
0.41%

rangeGain4 12 days
all
0.45%
condition
0.51%


Interpretation
the day following such a day shows significant higher than average dailyGain. the effect lasts for some days and fades out after about ten days.

Adding information on where the close was in recent range does not add anything.


peace

 

[Foz]

Quote from man:

Interpretation
the day following such a day shows significant higher than average dailyGain. the effect lasts for some days and fades out after about ten days.

Great study! However, when you say the following day shows a "significantly higher average dailiyGain", the word "significant" should have a specific statistical meaning. Did you or can you do a Z-test? We'll need the size of your sample and the standard deviation for both the sample and the overall market. We'll assume a normal distribution is a decent enough distribution for the one day change in the S&P.

I hope you see the reason for a statistical test. Your results might be due to a couple big outliers in a small sample that were just flukes. A statistical test can get at whether your results truly are significant.

 

[man]

ttest on dailyGain: 0.00

STANDARD DEVIATIONS
dailyGain
all
1.11%
condition
0.90%

rangeGain1
all
2.05%
condition
1.98%

rangeGain3
all
2.71%
condition
2.68%

rangeGain4
all
3.80%
condition
3.15%


Usually I put things like this in, but I learned that people often get confused by too many numbers within one post ...

BTW, the median tells the same story as the mean.


peace