| View previous topic :: View next topic |
| Author |
Equity Put Call Ratio |
nodoodahs Moderator

Joined: 06 May 2005 Posts: 2408
|
Posted: Thu Jul 14, 2005 5:25 am Post subject: Equity Put Call Ratio |
|
|
Does anyone have a source for Equity Put Call Ratio, on a weekly (or better yet daily) basis, going back before Oct 2003? Ideally I would love to have it going back to 1980 or prior ... volume is not important.
I have an analysis of the post 2003 data but would like to expand it. _________________ I haven’t seen a beatin’ like that since somebody stuck a banana in my pants and turned a monkey loose. |
|
| Back to top |
|
 |
| Author |
Equity Put Call Ratio Replies |
nodoodahs Moderator

Joined: 06 May 2005 Posts: 2408
|
Posted: Fri Jul 15, 2005 8:00 am Post subject: |
|
|
FYI the CBOE offered for free the total volume to 1995. The only "free" daily equity volume ratio I found went to 2003.
SR I know you probably want Henry's links for your own purposes, but for my purpose I don't think 1995-present is much better (if any) than 1997-present which Henry gave me. _________________ I haven’t seen a beatin’ like that since somebody stuck a banana in my pants and turned a monkey loose. |
|
| Back to top |
|
 |
SRmanager Junior Poster

Joined: 20 Jun 2005 Posts: 26
|
Posted: Fri Jul 15, 2005 7:20 am Post subject: |
|
|
$WPCVA
CBOE Put Call Volume Ratio – All Products
$WPVA / $WCVA
$WPCVE
CBOE Put Call Volume Ratio – Equity Products
$WPVE / $WCVE
$WPCVI
CBOE Put Call Volume Ratio – Index Products
$WPVI / $WCVI
I have tradestation and the only data series before 1997 was $WPCVA, which goes back to October 1995.
Henry, could you kindly let me know what datasource you used. |
|
| Back to top |
|
 |
nodoodahs Moderator

Joined: 06 May 2005 Posts: 2408
|
Posted: Fri Jul 15, 2005 5:20 am Post subject: |
|
|
Yes, CBOE Equity Put/Call Ratio for all equities, the volume ratio. Henry has provided data from 1997 to present and I am working on analyzing that. The above post is using 2003 to present data.
If you can find much further back than 1997 that's great, if not, I think 1997 to present may be significant enough to be credible.
I suspect the results may be different for different markets, i.e. the 95-00 market, the 00-03 market, and the 03-present market. _________________ I haven’t seen a beatin’ like that since somebody stuck a banana in my pants and turned a monkey loose. |
|
| Back to top |
|
 |
SRmanager Junior Poster

Joined: 20 Jun 2005 Posts: 26
|
Posted: Thu Jul 14, 2005 10:01 pm Post subject: |
|
|
Is this what you are looking for:
"CBOE Equity Put/Call Ratio"
Are you looking for the open interest or volume ratio? Any particular index? If you have the symbol, I can probably find it on bloomberg fairly quickly. Let me know.
Best |
|
| Back to top |
|
 |
nodoodahs Moderator

Joined: 06 May 2005 Posts: 2408
|
Posted: Thu Jul 14, 2005 6:15 pm Post subject: Results so far ... |
|
|
For the results so far ...
Data:
EPCR from 10/21/2003 through 7/13/2005 downloaded.
SP500 closing values from 5/20/2003 through 7/13/2005 downloaded.
Variables:
LAST6 = SP500 return over the last six months. For example, the first data point is the close for 11/17/2003 divided by the close for 5/20/2003.
20AVG = average EPCR for the last 20 trading days, inclusive. For example, the first data point is the simple average of EPCR for 10/21/2003 through 11/17/2003.
60NEXT = SP500 return over the next 60 trading sessions. For example, the first data point is the close for 2/13/2004 divided by the close for 11/17/2003.
120NEXT = SP500 return over the next 120 trading sessions. For example, the first data point is the close for 5/11/2004 divided by the close for 11/17/2003.
Additionally I calculate averages for each field and the times particular SP500 returns exceed those averages.
The last data point fully evaluated is 1/20/2005, since this is 120 trading days from 7/13/2005. This gives only 296 data points, not really a sufficient amount for credibility.
Analysis:
20AVG has the following R values for the variables: 60NEXT 0.6174, 120NEXT 0.6552, LAST6 -0.6640. Note this is consistent with the use of EPCR as a contrarian indicator, high EPCR is "bearish" and has good correlation to "bullish" results over the next 60 or 120 trading sessions. Similarly, low EPCR is "bullish" and has good correlation to "bearish" results for the next 60 or 120 trading sessions.
Next I separated 20AVG into quintiles from lowest to highest values.
In the case of 60NEXT, the results for the lowest two quintiles had a very low percentage of cases where average return was exceeded, but the highest three quintiles had a reversal (i.e. the middle quintile which should be "neutral" had the better chance of exceeding the average return than the highest quintiles which are "bearish" sentiment indicating a contrary bullish result - further, the most "bearish" sentiment quintile had basically neutral results!).
In the case of 120NEXT, there were no reversals and the EPCR was very predictive. EPCR seems to have (at least for the last 1 1/2 years) a very good contrarian value.
HOWEVER ...
The variable with the highest correlation to 20AVG (the average EPCR for the last 20 sessions) was ... TADA! It was LAST6!
So I separate LAST6 into quintiles from lowest return to highest return. The average 20AVG reading for the quintiles of return were 0.723, 0.693, 0.642, 0.632, and 0.580.
So based on the last 6 month's return for the SP500 we can take a fair guess at the "bullish" or "bearish" sentiment of the market, as measured by the EPCR.
Thoughts:
EPCR is a function of the previous market's returns - makes sense as conventional sentiment is defined by what's happened recently. EPCR has value as a contrarian indicator. EPCR seems to be more predictive for a six-month outlook than for a short-term outlook. Note Investopedia states it is "never exact and often a bit early" which is consistent with this limited analysis.
It is possible that the true measurement taking place in EPCR is not sentiment, but actually regression to the mean in SP500 returns, where outperforming quarters or halfyears are followed by underperforming quarters or halfyears and that we may have more success by folding in previous stock market returns than by folding in EPCR in our analyses.
It is also possible there is value added by using both EPCR and previous returns, but care must be taken in doing so, as much of the data in the two variables is redundant.
Having EPCR values for longer periods would allow a more rigorous analysis, which is why I am requesting a data source. I would be happy to email a file to anyone who requests via email or PM.
Since my market instability model currently uses the previous year's SP500 return as an input, I will not be adding EPCR until I get more data and do a multivariate analysis. _________________ I haven’t seen a beatin’ like that since somebody stuck a banana in my pants and turned a monkey loose. |
|
| Back to top |
|
 |
nodoodahs Moderator

Joined: 06 May 2005 Posts: 2408
|
Posted: Thu Jul 14, 2005 5:27 pm Post subject: Specifics |
|
|
Thanks SR!
Specifically - I have a file with EPCR on a daily basis, one column for the date and the other column for the EPCR, one record for every trading day from 10/21/2003 through yesterday. I would like to append that file, as far back in time as possible, with the specific Equity Put Call Ratio for every trading day.
MS Excel 97 or compatible format, comma delimited text or CSV formats. _________________ I haven’t seen a beatin’ like that since somebody stuck a banana in my pants and turned a monkey loose. |
|
| Back to top |
|
 |
SRmanager Junior Poster

Joined: 20 Jun 2005 Posts: 26
|
Posted: Thu Jul 14, 2005 10:46 am Post subject: put/call |
|
|
| I might be able to help. Could you be a little more specific? |
|
| Back to top |
|
|
Please log in to view without the ad banners |
 |
|
|
You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum
|
Powered by phpBB
|