Market Indicators - Investor Sentiment

Top 331. May 18, 2000 7:13 AM

» Rande - Well, hopefully the two links provided in response to Demo's re

Well, hopefully the two links provided in response to Demo's request for where to get more data will be helpful (I'll bet Statman would be happy to share his raw data, which evidently goes back decades). As for whether the article was overwhelming or not, guess that's an individual thing (disclaimer did say "long" and "academic") -- either way, everyone needs to follow their own approach. My goal is to provide as much information as possible so that whatever the approach may be, it's an informed one -- even if the occassional newsletter writer toe gets stepped on. ;)

-- posted by Rande

Top 332. May 18, 2000 8:25 AM

» Rande - No problemo.

No problemo. You know, sometimes the academics are wrong too. Gene's posts remind us of where the II numbers were in July and Oct of 1998 -- pretty compelling. Still, it pays to dig. I especially liked the reference in the piece to "gambler's fallacy" with the example of the coin toss and the basketball player. Why is that if a coin comes up heads three times in a row most people would expect it to come up tails next time when the odds are ALWAYS 50-50? And why is that if a basketball player makes three shots in a row most people would expect the next one to fall in too, even if the evidence shows a random sequence? Crap table is another example. Each throw of the dice is an independent event and the odds are the same every time. But I KNOW the table can be hot and the table can be cold. Strange stuff. For investors, knowing the difference between emotion, wishful thinking, etc. and reality is critical. Sometimes, the answers are not black and white as we would hope. So, as long as the markets are made up of people, we continue to struggle along. S'okay....so long as we have FUN doing it. I'm with you on that score!

-- posted by Rande

Top 333. May 18, 2000 11:16 AM

» Demogremlin - I'm still here

Scare me away??

Are you kidding?

I love the article you posted Rande. The reason I am so interested in getting past data, is because I want to see what the empirical merits are for the indicator.

I tend to think that the market IS largely efficient. I don't think it is entirely efficient, or I wouldn't be here. But I think the burden of proof is properly placed on the person who claims to have found an inefficiency. I want to try to find the proof if it exists.

While Statman's study seems discouraging, I'd like to check out some additional things. He considered how the idicator worked for predicting tops. Well, how did it do with bottoms? How does it do at extremes? What are the odds of a 10% pull back when the ratio is over 70? A 5% pullback?

Make sure you check out my post to the Market timing model discussion. Let me know what you think should go into a good timing model. And comment if you wish on the consideration I brought up there.

This is fun. I just have to make sure I get my other work done, too.

Panspar

-- posted by Demogremlin

Top 334. May 18, 2000 5:10 PM

» Rande - Panspar,

Panspar,

Good enough, and interesting point about the extremes in the II indicator. Statman, et al, addressed this point in both papers as they reference the index creator's claim that is the extreme readings of the BSI that supposedly have predictive value. This goes along with Brinker's assertion that only a "reading over 70" is meaningful. For anyone who takes this stuff to heart, the Statman piece is must reading in my opinion. I do believe he and his cohorts are being as objective as possible in examining the data. In any event, it's fascinating stuff.

-- posted by Rande

Top 335. May 19, 2000 7:33 AM

» MadFrank - Sentiment indicators

Brinker has also said he looks at Put/Call and cash levels of mutual funds.

-- posted by MadFrank

Top 336. May 19, 2000 9:34 AM

» Gene - Sentiment Readings

Rande, thanks for the excellent article and I agree it is MUST reading. It helps to get one grounded as to the validity or fallacy behind many of the indicators. Also very interesting discussion here on what can or cannot be interpreted from the various Sentiment Indicators. I guess Investing is just like Baseball, a GREAT statistical sport!

First of all, I come at my investing from a Fundamental analysis perspective - earnings, P/E, GAARP and Value Line Analysis etc since it was what I cut my teeth on when looking at inidividual stocks. Maybe that is one of the reasons I find the analysis of folks like Kirk and Jim Jubak on individual stocks very interesting and on solid grounds. Likewise your consistent and solid advice on Asset Allocation and staying the course since that will get Most folks the Best returns for their Long Term financial goals.

I freely will admit to occasional forays into Market Timing, but then I think Kirk's "Periodic Bouts of Profit Taking" and Rande's
"Tactical Allocations" (which I agree with) are for pseudo timing :>)

I became interested in the Sentiment Readings around 1993-1994 when the Fed was last doing their numerous tightening for a term that I heard for the 1st time - to produce a "Soft Landing" and after hearing Brinker and a few others talk about it in their investment models. I became more interested in the "contrarian nature" aspect of it with sayings like, "when the average folks came in, time to get out", "distribution was the smart money selling to the dumb money" etc. The very idea that they took the survey of "Intelligent Investors" of Financial Advisors and their Bullish / Bearish leanings as a contrarian indicator that warranted from 25% to 33% of anyone's model was too interesting to ignore!

Over time, I came to look at it as more of a "trend and directional indicator" and it probably is no more than say 10% weighting at max for me personally. With most other pieces of data being of an objective nature, it seemed reasonable to factor in a subjective piece of data, IMHO. My posting of the readings around the 1998 Market turmoil was meant "See what happened in Hindsight" for these Readings.

I admit to using it as a starting point to then peel back other pieces of data at different times, much as Greenspan lurches from data point to data point! the old saw, one set of data leads one to look at another etc etc. For instance, it kept me in the market during the 1995 - 1998 time frame when it seemed to me the Bullishness was warranted after researching and looking at all the projected S&P earnings etc. It might seem obvious now in hindsight to most folks but since I had been in & out of the market since 1968 and had learnt my investing under a different set of rules, the Sentiment Reading was a kind of a catylst to look deeper into Sentiment if I had a problem understanding some aspects of it.

For instance, for the last 3 to 5 weeks, I have been focusing on Cash flow and liquidity. Why? After the record amounts of $$ flows into the Markets with a VERY high Bullish reading during the Jan - April time frame only producing negatives in all Indices, the high readings became a BIG negative interpretation to me. Also since I was predisposed to thinking liquidity is critical to sustaining this market after the NASDAQ crash (MHO) in March / April, I was even more curious for that continuing Bullish Readings. During the very high Bullish % in late 1999 and early 2000, I was checking out all the Distribution, Insider sales and decided it was very negative also.

Lastly, for many of us overly analystical types, it can be combined with all kinds other pieces of data as Kirk and others mentioned - VIX, Put/Call etc to thoroughly confuse one into total inaction while spending all the time producing charts, graphs and tomes:>)

Since it is a rainy old day hereabouts, along with Double Witching Friday, the markets in some sort of swooning and I have this week's Sentiment Readings (courtesy of a Kirk email), time to prepare this week's Sentiment Readings Posting!

Gene

-- posted by Gene

Top 337. May 19, 2000 10:57 AM

» Gene - 5/22/2000 - Investor's Sentiment Readings

Well, it looks like that slight uptick on last week's Bullish reading was nothing but more than a head fake when this week's Sentiment Readings came in from the "Surveyed". The drop to 41.9% Bullish with the subsequent drop to a ratio of 55.50% was certainly one of the steepest in such a short time. Looks like many of the Correction folks jumped off the fence this week. Then again, maybe they are trying to "hurry up" whatever is supposed to happen in this here doleful Market to get going! Kind of like when that "Analyst" last Fall had projected QCOM to go to $1,000 when it was already at $500, so everyone piled in figuring they would make a real fast buck and then sell higher to somebody else. Or as Jay pointed out a few weeks ago, maybe as a result of what is actually going on in this market with a buyer's strike, they are truly a lagging indicator. But then again, maybe another head fake? So much data to consider, so little brain power to expend.

We continue to believe the economy is in great shape, hopefully can withstand the latest Fed moves without the dreaded "R" word happening and will help in the Long Term. It certainly has made several of the Fixed Income types hereabouts happy as they have been extremely pleased with what they are getting for CDs, Treasury Direct etc for those stock market profits they squirreled away these last few months.

Interesting WSJ article yesterday about how many companies are now able to pass along costs and "tack on a little extra" to make up for not being able to do so the last several years. I know this would sound like heresy hereabouts and to the Fed BUT maybe that isn't so bad IF it allows some productivity to then seep into the system to reward the folks doing the work!

Frankly, we are hoping these moves do the trick as we had gotten very tired of all the Graphs and Charts showing us ad nauseum how this great Bull market was similar to 1929 etc. But then as Rande's article on Sentiment Readings pointed out, sometimes we get conflicting proverbs and sayings, along the lines like "Those who ignore history are doomed to repeat it"!

One cannot deny we have a most interesting market on our hands and we hope folks stayed tune to S101 to read, post and participate in the best all around advice on the I-net Financial dial.

The Rolling 4 week Sentiment average covers the period from 5/1/2000 thru 5/22/2000.

It is also important to remember that Bob Brinker considers a 70% ratio number to be neutral and is only one of his four main indicators with the other three being Economic, Monetary and Valuation.

From Kirk email (thanks) 5/17/00:

Investor's Intelligence

Bulls = 41.9% Last Week = 48.6% Two Weeks = 47.6%
Bears = 33.6% Last Week = 31.2% Two Weeks = 31.4%
Corr = 17.3% Last Week = 20.2% Two Weeks = 21.0%

% Bulls/(% Bulls + % Bears)= Sentiment %

41.9%/(41.9% + 33.6%) = 55.50%%

Rolling 4 week average = 59.98%

Weekly Bullish Sentiment %:

Week 05/15 = 60.90%
Week 05/08 = 60.25%
Week 05/01 = 63.28%
Week 04/24 = 65.99%
Week 04/17 = 68.06%
Week 04/10 = 65.28%
Week 04/03 = 64.79%
Week 03/27 = 67.84%
Week 03/20 = 64.33%
Week 03/13 = 65.93%
Week 03/06 = 64.84%
Week 02/28 = 64.42%
Week 02/21 = 66.63%
Week 02/14 = 65.60%
Week 02/07 = 63.78%
Week 01/31 = 68.17%
Week 01/24 = 67.08%
Week 01/17 = 67.37%
Week 01/10 = 64.84%
Week 01/03 = 67.07%
Week 12/27 = 64.41%
Week 12/20 = 64.43%
Week 12/13 = 63.83%
Week 12/06 = 64.87%
Week 11/29 = 64.16%
Week 11/22 = 62.61%
Week 11/15 = 55.29%
Week 11/08 = 52.70%
Week 11/01 = 51.70%

07/20/98 Bulls = 52% Bears 24% for 68.42%
10/12/98 Bulls = 38.5% Bears 42.7% for 47.41%

Gene

-- posted by Gene

Top 338. May 19, 2000 12:23 PM

» Demogremlin - Statman

Well, I sent an email to Professor Statman asking about his raw data on the sentiment numbers. He got back to me the first thing this morning.

However, all he could do for me was to tell me that Investors intelligence charges for the data, and he couldn't remember how much.

I'll try to find out from II how much they'd charge for their data. I'm not inclined to pay much, however.

I guess that leaves us with finding another source, or gradually and slowly extending our data back in time.

We have data from June of 98 on, thanks to Glen and Kirk (I guess. I'm still trying to figure out who does what around here.)

If someone would like to collect data from July of 97 through June of 98, I'd volunteer to collect data from July of 96 through June of 97.

Any takers?

By the way, Gene, I'm sure you'll catch this eventually, but the numbers for this week don't seem to me to add up to 100%. You might want to double check the numbers. I hope I'm not losing my mind. That's possible too.

Panspar

-- posted by Demogremlin

Top 339. May 19, 2000 12:40 PM

» JenL_2 - Post of the Week Nomination

Gene - Fantastique!

I nominated your two messages above as "Posts of the Week":

Thanks Gene

.....Jen

-- posted by JenL_2