sentimenTrader.com

All rights reserved.

Thursday, August 1, 2002  9:27 PM EST

Today's commentary will be short because from a sentiment perspective, I don't have much to add.  The carry-through from yesterday's strong close was good for a scalp only from the gap down open, but the econ numbers at nine squashed any hope for more.  It seems one more bad economic number was all the market could take, and we have another big test tomorrow morning.  Another bad release should give us the most-discussed retest of a low I can ever remember.

The VIX closed up over 6 full points today, which is one of the larger volatility increases in the history of the indicator.  There have been 21 days in which the VIX closed 6 points or higher from the previous close, and it has proved to lead to interesting market action.  The average return 1, 5, 10 and 30 days after such jumps is positive, beating a random return by about a percentage point.  The market was up after each time frame about 2/3 of the time.  The caveat here is that several times such a move lead to an even LARGER move soon after, coinciding of course with a large market drop.  These are perilous times, and if you have a short time frame, caution is warranted.  Although I am becoming more bullish the further we drop, possibly leading to a retest, I am on alert for the possibility of violent intraday action that seemingly comes out of nowhere.  These jumps in volatility always put me on guard, and I don't like it one bit.

As of this evening, I don't have any particular bias once again.  I am open to moves from both directions, although the bias is for more downside.  We are nowhere close to oversold on a short-term basis, and the constant talk of a retest will probably keep any interested buyers away since they believe they will be able to get in at better prices within a few weeks.  The employment numbers tomorrow could have a large impact either way, so that will also influence the action.  I will be anxiously awaiting this weekend's COT and NYSE Members numbers to get a better handle on the intermediate-term picture.

Wednesday, July 31, 2002  8:42 PM EST

So was today's late-day reversal a month-end markup trick by mutual funds looking to goose performance a bit and stem the flow of redemptions?  Common sense and conventional wisdom would suggest so, but the data suggests otherwise.  I looked at the last 1/2 hour and last 1 hour periods over the past four years to see if there was a tendency for late-day markups on the last day of the month.

You can see that there has been a very slight tendency for the market to rise during the last part of the trading day over the past four years - EXCEPT for the last day of the month!  There is actually a strong tendency for the market to DROP during the closing on the last day of the month.  This is the opposite of what most journalists spew out as common knowledge.  So, now we know that today's late-day rise was unusual.  What good does that do us?  I then looked at past periods where there was at least a 1% rise during the closing hour (not just the end of the month).  The average return during the following 1/2 hour was 0.22% (with a 59% "win" rate and avg win/avg loss of 1.32) and the average return during the following 1 hour was 0.26% (with a 64% "win" rate and avg win/avg loss of 1.18).  So that tells us that when there is a significant late-day rise, the market usually follows through the next morning.  That will have me on the lookout to possibly buy very early weakness or a gap down for a quick trade, although the jobless claims number will impact that to some degree.

I have now created a new put/call ratio based on the information I have been presenting over the past week on the S&P 500 options.  I simply take all the trades at the bid and divide it by those on the ask.  I then take the bid/ask ratio on the puts and divide it by the bid/ask on the calls.  This takes all of our information and expresses it as a single ratio.  When the ratio is high, then that means there has been a strong bias towards selling puts and/or buying calls - a bullish development.  When the ratio is low, then the institutions have shown a bias towards buying puts and/or selling calls, which is bearish.  We don't have enough data to exactly determine what is "high" and "low".  Again, I am generalizing here out of necessity - there is no way to know for sure if these purchases and sales are part of a larger strategy, but I believe the generalization holds water.  I'm not feeling very creative right now, so if any of you have a good idea for a name for this ratio, shoot me an email.  Here are the readings for the past week:

Although I don't have nearly enough data to deem this ratio consistently effective, it makes fundamental sense and it has proven extremely effective during this last week of turmoil.  I think it's best used as a very short-term secondary indicator to clue us in to the institutional mindset.  If the ratio is high after a downtrend, it should give us additional confirmation that a trend change back up may be imminent - vice versa for a low ratio.  I don't think today's reading gives a clear enough bias from either direction to provide much value.

Today's session should be just what the bulls are looking for - bad economic releases couldn't knock the air out of this nascent rally, and the intermediate-term conditions are still oversold.  They have every excuse in the book to keep running this market higher, but as I said yesterday I think another two or three days should put something of a cap on it if we keep rising without a considerable retracement.  Our shortest-term indicators climbed out of overbought with the morning's decline, so there's certainly room from a sentiment perspective for a further spurt upward.  I don't have a particularly strong bias for either direction in the short-term, so I'll be looking both ways tomorrow.

Tuesday, July 30, 2002  8:22 PM EST

After surfing through one of the largest one-week (5-day) losses in history, we now are faced with one of the largest one-week gains in history.  There is really not much to go on as far as historical precedent goes, so we'll need to play this one by ear.  Although the recent action is not unexpected and is a positive sign for long-term market prospects, the short-term is looking more dubious.

I've said that when a new trend is in place, short-term sentiment needs to take a back seat due to numerous false signals.  The longer the trend lasts, however, the more weight I begin to place on extreme readings.  I realize that one day is not a long time, but our shortest-term measures are once again in deeply overbought territory.  The STEM.MR model is close to zero, and the intraday cumulative TICK indicators haven't cycled down very much.  At some point, most likely very soon, these conditions will begin to exert influence on the market.  We don't necessarily have to go down, but at least churn sideways for a few days while these measures work off their extreme conditions.  Of course, the "market" can do whatever it wishes and may just keep rising every day, but I'm trying to weigh the odds here.

In the S&P 500 options, an institution bought 4,600 August calls at the 1050 strike.  On the surface, that looks positive, but the trade went off at a cost of $0.20 per contract, or a total dollar amount of around $92,000, which was almost assuredly covering an existing short position (no doubt for a large profit).  In the August expiration, the call bid/ask ratio was .76 (it would have been 1.50 without that large purchase mentioned above) while the put b/a ratio was .83.  In the September expiration, the call b/a ratio was a large 5.19 while the put b/a ratio was .59.  If we put the strikes together, we get a total call b/a ratio of 2.97 and a total put b/a ratio of .71.  Unlike yesterday, we have a very clear bias today towards selling calls and buying puts, which is a bearish development.  The heaviest skew was in the August 950 and 975 calls, and the September 900 calls.  It looks like they're allowing for a little more upside over the coming weeks before the assumed retest of the lows.  As I regularly mention, it's impossible to know if these are covering transactions, legs of various spreads or outright purchases and sales, but we can get a general understanding of institutional posturing by monitoring the trades going off on the bid and ask.  I realize that the nomenclature is confusing to those of you unfamiliar with options, and I will try to explain more as we go along, but I think it pays to keep abreast of these ratios.  I've never seen them mentioned anywhere else, so if it's confusing, please let me know.

I haven't mentioned the total put/call ratio by itself for quite some time, but it's now getting low enough to be worrisome.  The 3-day moving average of the one-year rank (which is posted to the site) has now cycled down to 45%, which is the lowest since the May peak.  The lowest before that was 20% at the top in March, so there's still certainly room to fall.

I mentioned this weekend that the advance/decline line would likely cycle up to neutral and then overbought very quickly with anything resembling a rally.  We've already reached the neutral stage and about three more days of relatively moderate to strong readings will take us up to very overbought.  It would be difficult to imagine much more upside after that before something of a retracement back down.

The intermediate- to long-term indications continue to look rather positive since the Wednesday reversal.  The short-term picture, however, is now growing more cloudy and any further rally attempts will likely be looked upon as aggressive shorting opportunities for a trade.

On a side note, I am in the process of revamping the signal process that has been in place for several months now.  There is just too much confusion as to what the signals are meant to convey.  You will receive an email with the details of the changes.

Monday, July 29, 2002  8:34 PM EST

One might think that such a large up day would necessarily lead to profit-taking and another leg down, as the bandwagon-jumpers who were late to the party lose like usual in the short-term.  That would be wrong.  I looked at the top 1% of percentage increases in the NYSE Composite Index since 1966 and their follow-through action.  The top 1% turned out to be moves greater than 2.3%.  "Random" is a random return during the study period.  "Avg" is the average return in the NYSE Comp the given number of days after a large up move as defined above.  "% Pos" is the percentage of time the Comp was higher the given number of days later.  I think the other rows are self-explanatory.

I then looked separately at those days where the index was below the 12-month moving average and the big percentage move was accompanied by at least 90% up volume (which is what we had today).  The results improve somewhat.

The entire point of the exercise was to determine if days like today, when the index was in a downtrend (as defined simply by being below the long-term moving average) and it had a large up day on hugely positive volume, pinpointed a short-covering climax that ultimately failed.  Although the study is fairly crude, I think it does show that these days have historically NOT been bad for the market.  In fact, these days prove to be fairly good BUYING opportunities, not opportunities to sell or even worse sell short.

The institutional investors we've been monitoring in the S&P 500 options don't seem to be taking advantage of this rise (which they were very well positioned for) to sell calls.  They actually seem to have a slight bias towards buying calls, particularly in the August expiration.  I find that highly unusual after such a large move and clearly defined resistance ahead.  The puts in the August and September expirations were fairly evenly balanced between buying and selling and I don't think there's much to read into there.

I suggested in the weekly commentary that the VXN put in a classic reversal pattern which usually preceded a further drop in that index with a corresponding rise in the Nasdaq 100.  That happened in spades as the volatility indices got clocked today, losing over 15% each (both the VIX and VXN).  These moves are quite drastic and there isn't enough data to determine what happened in the past during such declines.  I don't think it's a headlong rush back into complacency.

Obviously, short-term sentiment is at an overbought extreme.  The STEM.MR model is stuck at zero and the intraday cumulative TICK indicators are in extreme overbought territory.  The NYSE TICK indicator actually hit a new all-time overbought reading this morning.  But just like on the way down, these indicators will probably be quite ineffective on the way up unless they cycle back down to oversold.  Like I said in the weekly commentary and again in the signal alert, I would NOT fight this short-term trend.  I realize the longer-term trendlines all still point down, and we will probably have a rest period, but this market was like a tinder box just waiting to be lit.

- Jason Goepfert