sentimenTrader.com

All rights reserved.

Thursday, December 19, 2002  8:27 PM EST

Quick administrative note:  Due to multiple requests, I have begun posting the Down Pressure indicator for the S&P 500 and Nasdaq 100 to the site, and will do so daily from now forward.  For those who are new, please see the November 11th daily commentary for a review of how this indicator is constructed.  I am reconfiguring the site, so I will have better and easier to follow descriptions and examples of each indicator soon.  Also, as a sign of things to come, each day I am also including Rydex ratios for Treasury bonds and precious metals in the Indicator section.  Analysis of these areas will be expanded in the coming month.

As for today's market, I hate to sound like a broken record, but once again tonight we're neutral across the board on our indicators and models.  About the only movement we're seeing is in the short-term STEM.MR model and its components.  That model is now tagging is upper standard deviation band, which suggests that short-term sentiment is stretched as of today's close.  I wouldn't quite yet call it extreme, but a large gap down open or severe early selling pressure tomorrow would likely make it so.  If either event happens, I would be on the lookout for an intraday reversal.

On Tuesday, I mentioned that I was watching S&P 885 and NDX 1000 as important support levels.  Although we dipped a bit below that on the S&P, today's closes were S&P 884.25 and NDX 1006.05 so obviously there's a whole lot of interest in these levels.  If we cannot sustain trading above these zones, I believe the selling will intensify and likely last at least a couple of days before a sizable upside correction would materialize.  The light year-end trading and seasonality (is anyone besides me sick of hearing about the positive seasonality yet?) might throw a wrench into this scenario, but typically the follow-through would last a couple of days.

I've been struggling to find an edge lately, and still have not found one.  I don't know if that's good because we're just bouncing around or bad because we're about to make a big move, but I don't believe in pressing when there's not a definitively positive risk/reward scenario.  Until such a time arises, I continue to urge caution for both longs and shorts.

Disclosure:  long QQQ calls, long QQQ puts

This disclosure is not intended as trading advice in any form.  It is meant as a note to subscribers that I have a position directly affected by my market outlook.  Although I take great pains to remain objective in my commentaries, I believe it is only fair that readers should know that I have taken positions in accordance with my market outlook.

Wednesday, December 18, 2002  9:00 PM EST

I've received a few requests to reprise the Thanksgiving bias tables for the Christmas and New Year's holidays, so they follow.  I've included this year's applicable date in parentheses, as well as a table of the trend over the past 10 years for this year's corresponding trading dates.  Again, the study period is 1950 - 2001 for the S&P 500.  "Uptrend" is defined as an up-sloping 20-day moving average while "downtrend" is a down-sloping 20-day moving average.  "Exceeded high" means that the day rose above the prior day's high, while "exceeded low" is defined as a violation of the previous day's low.  So, for example, on the 5th trading day before December 25th (which fell on December 18th this year - today), the average daily return has been 0.21%, it has closed higher 55% of the time, and it had a 58% chance of exceeding yesterday's high and a 42% chance of breaking the low.  Like the Thanksgiving bias, there is no real clear daily bias until we get closer to the holidays themselves.  Even then, about the only days that show a true trend that I wouldn't want to fight are the couple of days after Christmas and the second trading day of the new year.  The day before Christmas, widely believed to be very positive, did not prove to be especially so, although it has been up the past three years.

Christmas Day minus 5 (12/18/02):

Christmas Day minus 4 (12/19/02):

Christmas Day minus 3 (12/20/02):

Christmas Day minus 2 (12/23/02):

Christmas Day minus 1 (12/24/02):

Christmas Day plus 1 (12/26/02):

Christmas Day plus 2 (12/27/02):

Christmas Day plus 3 (12/30/02):

Christmas Day plus 4 (12/31/02):

Christmas Day plus 5 (1/2/03):

Christmas Day plus 6 (1/3/03):

Christmas Day plus 7 (1/6/03):

Random:

Recent trend:

Although many of the individual days are nothing spectacular, the roughly two-week period just begun has historically been very positive as a whole.  The period from 5 days before Christmas through the first three trading days of the new year has been positive 83% of the time over the past 40 years, with an average return of just over 1.6%.  16 of the past 20 years have been positive, as has 8 of the past 10.  During the past 40 years, the minimum return during this stretch has been -3.9% while the maximum was 8.9%.  The average negative return was -2.6% while the average positive was 2.5%, however you must keep in mind that only 7 periods were negative and 33 were positive so an average positive return is stretched across many more samples.  If you think of this seasonality stretch in trading terms, it has a positive expectancy of 1.63%.  That is, ((% gain x avg gain) - (% loss x avg loss)) = ((83% x 2.5%) - (17% x 2.6%)).  So, if you simply bought the S&P on the close of the 6th day before Christmas and sold on the third trading day of the new year, you could reasonably expect to earn over 1.6% on your money year after year.  The bottom line is that this time of year is historically very positive, and I would be wary of fighting it.  On a few of the days, as outlined above, I would caution against being short.

In a rather remarkable show of stubbornness, the respondents to the Investor's Intelligence survey continue to stick to their bullish guns.  During the latest reporting period, the percentage bullish actually rose .1% to 50.6% while the percentage bearish rose a tad to 25.8% (from 24.2% last week).  This dropped the bullish ratio to 66.2% from 67.6% last week and 67.1% the week before.  Even though the S&P is lower by 5%, the bullish ratio has barely budged.  This is a relatively rare event, as over the past 20 years, when the S&P drops by between 4%-6% over a two-week period, the average reduction in the II bullish ratio is 5%.  So, with history as a guide, the bullish ratio should have come in around 62% this week as opposed to 66%.  Bullish opinion in the face of declining prices is NOT a recipe for an upside reversal of large proportions, so this continues to be a bearish sign going forward. 

There continues to be a real lack of developments on the sentiment front, on all time frames.  Like yesterday, virtually every one of even our shortest-term measures remains in neutral territory, and until we break out of the ranges I mentioned yesterday, that will likely continue.  I do believe that once the range is broken, there will be a quick spike in that direction, and probably another day or two of momentum.  If we break down, I would be cautious of becoming too aggressively short for more than a couple of days, for no other reason than the seasonality I mentioned.

Disclosure:  long QQQ calls, long QQQ puts

This disclosure is not intended as trading advice in any form.  It is meant as a note to subscribers that I have a position directly affected by my market outlook.  Although I take great pains to remain objective in my commentaries, I believe it is only fair that readers should know that I have taken positions in accordance with my market outlook.

Tuesday, December 17, 2002  7:51 PM EST

There is really not much at all to go over tonight from a sentiment perspective.  All models and almost all indicators are neutral (or slightly negative), so from a longer-term perspective nothing has changed.  In the short-term, the overbought conditions I mentioned yesterday are gone (including the bearish bias in the S&P 500 options), and even our shortest-term measures are also neutral.

While there is now room for the indexes to move higher, the longer-term volatility measurements I outlined a few days ago are beginning to become troublesome.  These are the average daily S&P 500 range, average daily TICK range and contraction in NYSE volume.  If we do get a week or so of generally rising prices (in keeping with the positive seasonality), each of these volatility measurements should firmly enter a territory that has coincided with intermediate-term highs almost without exception during this bear market.  If that happens, in combination with excessive bullishness from the sentiment surveys and bearish overtones from the Commitments of Traders data, the intermediate-term outlook at that point would almost certainly become extremely negative.

I've heard and read a lot from fellow traders about how boring the market has been over the past week.  This is not surprising, as I pointed out a few days ago that December has the narrowest average daily range of any month of the year.  The past week's volume has run about 90% of the yearly average, but the last 5 trading days of the year usually run about 70% of the yearly average, so the pace will only get slower as we approach the new year.  There should be a couple of spikes before the year closes, however, so stay on your toes and don't let the pace lull you into carelessness.

As I said, virtually every one of our indicators is currently neutral, reflecting the recent trading range.  Now that the breakout/breakdown levels are fairly well-defined (approximately S&P 885 - 910 and NDX 1000 - 1055), a break of these prices should cause a quick spike in that direction.  My gut instinct tells me that we'll spike higher above these levels at some point before the year is out, so that will be something to watch.  Until then, I don't see an edge that would justify aggression in either direction.

Disclosure:  long QQQ calls, long QQQ puts

This disclosure is not intended as trading advice in any form.  It is meant as a note to subscribers that I have a position directly affected by my market outlook.  Although I take great pains to remain objective in my commentaries, I believe it is only fair that readers should know that I have taken positions in accordance with my market outlook.

Monday, December 16, 2002  8:50 PM EST

In the intraday note today, I suggested that it would be best to be careful if trying to pick a top today, because these types of moves tend to feed on themselves.  We saw proof of that today as both the S&P and Nasdaq barely paused on their march higher.

I didn't see much evidence sentiment-wise to support a rally of such duration, but two of the short-term positives coming into today have completely reversed.  The NYSE intraday cumulative TICK was at an oversold -569 at the open but closed the day at +4067, just below the +5000 overbought threshold.  The Nasdaq cumulative TICK hasn't reached overbought, but has cycled from -2096 at the open to +194 at the close (readings closer to +2000 could be considered overbought).  The price oscillators have also cycled from oversold to overbought today, egregiously so on the S&P with a reading of 70%.  Again, the price oscillators are just a way to "score" price bars.  A bar that opens at the low and closes at the high -  a bullish pattern - gets a score of 100.  Conversely, a bar that opens at the high and closes at the low - bearish - gets a score of 0.  A neutral bar which opens and closes in the middle of its range (a doji star for candlestick aficionados) is scored as 50.  The oscillators themselves are 13-period moving averages of 30-minute bars, and when they reach 60% or 40%, it often coincides with overbought and oversold market conditions, respectively.  Today's closing reading was 70% on the S&P and 62% on the Nasdaq - both overbought and suggestive of a short-term pause at least.  It depends on how you choose to define a trend, but it appears as though the cumulative TICKs and price oscillators are overbought (or nearly so) in the context of a downtrend, and that is a high-probability short setup.  That's not to say that it always works, but it is high-probability, and I fully expect a decent short opportunity tomorrow, for a short-term trade.

The lowrisk.com sentiment survey came out today with a drastic reduction in bullishness.  The percentage of poll respondents who stated they were bullish dropped to 24% (from 43% last week) while those bearish rose to 53% (from 45% last week).  The neutral camp also rose to 24% from 13% last week.  This combination dropped the bullish ratio to 31% from 49%, which is one of the larger one-week drops in the survey's history (a little over 5 years).  However, the four-week moving average of the bullish ratio, which has been quite effective at pinpointing intermediate-term market turning points, is still at a high 42%.  As I state every time, this survey shows a pretty close correlation to the more widely-known Investor's Intelligence survey which is released mid-week, so we can often get a heads-up as to what that survey may show.  Since this week's lowrisk.com survey showed such a large change, I do expect to see a larger-than-normal drop in bullishness in the II survey, but it's doubtful that it will be as drastic as the lowrisk survey.  After two weeks of steadily declining prices, it's about time we saw some pessimism begin to creep back into public opinion.  Perhaps last week's apathetic drop was enough to scare some people into not taking the end-of-year rally for granted.

The S&P 500 put/call bid/ask bias ratio came in today with an extremely bearish reading of 0.12, the second-lowest reading in the past five months.  For those of you who may be new, the bias ratio is constructed by comparing the trades in SPX options that go off at or below bid to those that go off at or above ask, for both puts and calls in all expirations.  If there is a clear bias towards buying puts and selling calls, the ratio will be low and is considered bearish.  If the ratio is high (say above 3.0 or so), then the institutional traders who frequent these options are showing a bias towards buying calls and selling puts - a normally bullish phenomenon.  Although there is no way to determine the underlying strategy behind the trades - and in fact what may appear bearish could actually be part of a bullish overall position - when the bias ratio reaches one extreme or the other, it often precedes a good move in 1-2 days.  Today's reading would suggest that there will be a decent move down tomorrow or Wednesday, although I believe that expiration is having a huge impact on the indicator and it could be giving a false indication.

While we're beginning to see some signs we need to see to be more confident of an intermediate-term low, there is much, much more work to be done before I would consider the risk/reward skewed to the long side, or even balanced.  At this time, it is still skewed to the short side, now across all time frames.  Although I am very aware of the positive bias around the Christmas and New Years holidays, that is about the only positive sentiment-type measure I can find.  All others are neutral to negative, so I believe rallies should be used to adjust positions accordingly.  For short-term traders, a large gap up opening or early-morning upside run tomorrow should serve as a high-odds opportunity to initiate short positions once price begins to stall (for a trade).  For longer-term traders, I believe higher prices should be used to sell, not buy.

- Jason Goepfert

Disclosure:  long QQQ calls, long QQQ puts

This disclosure is not intended as trading advice in any form.  It is meant as a note to subscribers that I have a position directly affected by my market outlook.  Although I take great pains to remain objective in my commentaries, I believe it is only fair that readers should know that I have taken positions in accordance with my market outlook.