Wednesday, January 29, 2003 8:25 PM EST
The most recent Investor's Intelligence sentiment survey came out with another basically unchanged reading, despite the savage beating the market took during the survey period.
During the past 15 years, there have been 30 two-week periods where the S&P 500 dropped more than 5% (Friday close to Friday close), not including our current drop of just over 7%. Of those 30, only 7 showed an increase in the II bullish ratio. And of those, only 2 occurred when the bullish ratio was above 60%. I am defining bullish ratio here by the following formula:
Bullish Ratio = BULLS / (BULLS + BEARS)
One instance was the two-week period ending 3/16/01. Following that, the S&P quickly dropped another 6%, before swinging violently over the next few weeks, eventually leading to a new leg higher (but not before the bullish ratio dropped to nearly 50% first). The other instance was the two-week period ending 7/28/00. After that, the S&P immediately took off on a 7% rally, but that failed within a month and obviously took us to new lows. However, both of these cases had a bearish percentage of at least 30%, compared to the current low reading of 26.1%. In fact, this bearish percentage is the second-lowest in the past 15 years (after the market dropped more than 5% over the previous two weeks), eclipsed only by a reading of 25.8% on 12/13/02.
So, despite the S&P suffering its ninth-worst two-week period in 15 years, and despite the index being down 44% from its high, the bearish percentage is in the bottom 7% of all the readings over that time frame. To me, that sounds like a "it can't go any lower" mentality, and I do not think it's healthy. Put another way, the bullish ratio has been above 60% for 14 straight weeks. I don't like to fit indicators to match the data, but the chart below shows a 14-week moving average of the bullish ratio over the past 13 years. Although it is a badly lagging indicator, it does show the stark contrast between our current situation and past major, multi-year bottoms. Look at the readings below 40% which corresponded to the major lows in 1991 and 1994. Then look at readings above 65%, which is where we are today.
Because bear-market rallies tend to originate from panic selloffs and are usually extremely quick and forceful, an indicator with this type of a lag does not perform well for timing intermediate-term moves in our current environment. However, the intent is to show major turns that last for several months or more, and the conclusion from this indicator is that we are nowhere close to a new bull market.
As I said in the comment this weekend and again in the intraday note today, we continue to have several sentiment indicators that are on completely opposite extremes. This very often precedes extreme volatility in both directions, and that has played out so far this week. I have no reason to think that will change anytime soon, so we are in a rather difficult spot right here. On the one hand, we got the "buy" signal from the STEM.MR model today (see Monday's commentary), which would argue for higher prices. Along with that, we have seen extremely aggressive Rydex asset flows into the leveraged bear funds, panicky-type readings in the VIX, oversold breadth readings and overall positive action in the S&P 500 futures (i.e. Commitments of Traders data). Also, the Composite model is approaching 70% for the first time since the October low. All of these would suggest that we are close to an intermediate-term low (if we haven't already seen it).
On the other hand, we have relatively complacent readings in the put/call ratios, new high/new low readings nowhere near where they "should" be, and still-too-optimistic readings from most of the sentiment surveys.
I think the great majority of us have been trading long enough to know that we are rarely - if ever - going to get everything lined up perfectly, with bells going off telling us to buy. So, the optimal buying point for an intermediate-term rally may have already passed us by. But I think when the larger (monthly and weekly) trends are taken into consideration, we have too much evidence suggesting complacency is a problem, along with the fact that the market has not become nearly as oversold as it has at most points preceding intermediate-term bear-market rallies. Therefore, I do not believe Monday's low will hold. I will have to reconsider that conclusion if we are able to trade through and hold above the 870 level on the S&P, which is the point where we broke through a triple bottom. If so, it would suggest that we just saw a false breakdown, and could mean that higher prices are imminent. Since we are now approaching overbought on many of our shortest-term indicators, and have just rallied to resistance, I don't like the odds of being long right now. I think this term is way over-used, but how the market reacts around this 870 number truly will be "pivotal" to the longer-term health of the market. Until this is resolved, I am comfortable holding our small intermediate-term short position in the model portfolio.
Disclosure: long QQQ puts
This disclosure is not intended as trading advice in any form. It is meant as a note to subscribers that the author may have a position directly affected by the market outlook reflected in the commentary. Although the author takes great pains to remain objective in any commentaries, it is only fair that readers should know that the author may have taken positions in accordance with his market outlook. Positions can and do change at any time, without notice to the reader.
Monday, January 27, 2003 7:50 PM EST
The STEM.MR model set a new two-year high today, with a reading of 69%. Obviously, this is greater than the readings at the September 2001, July 2002 and October 2002 lows. Readings above 60% have only occurred four distinct times over the past two years: 9/17/01 through 9/21/01; 7/15/02 through 7/24/02; 8/5/02; and 10/4/02. As of today's close, the model has been above 60% for 12 straight half-hourly periods - this is exceeded only by 19 consecutive readings from 9/20/01 - 9/21/01. What must be noted is that these readings are not buy signals unto themselves. If you simply bought when the model reached a hyper-extreme reading (above 60%), you would have suffered what anyone would consider an unacceptable drawdown before the trade became profitable. As I often say, these models are not to be used in isolation - you must at least wait for price confirmation before becoming too aggressive in your trading. For example, if you had waited for the first violation of a previous day's high after the model gave a reading above 60%, each signal would have lead to nice gains with minimal drawdown.
There is one slight cheat here - on October 8th, the S&P violated a previous high by a barely perceptible 0.64, and held there for just a few minutes, so I didn't count that. A better entry signal would have come on October 10th, where we firmly violated and held above the previous day's high. With only four signals, there is not nearly enough of a track record here to say with any degree of confidence that we should expect the same kind of results this time. Obviously, I would never suggest you bet the farm the next time we violate a previous day's high just because of this model, but I think it is intriguing enough to take a shot with a small amount of risk capital, with a stop loss below the entry day's low (if you even want to give it that much room) and a tight trailing stop. I will most likely take this trade for the short-term portfolio.
Adding some credence to this signal is the fact that the STEM model, 10-day advance/decline line, 10-day up volume ratio and NYSE cumulative TICKS (daily and intraday) also entered oversold territory by the close today. If we add in today's (finally) elevated put/call ratio, spiking VIX and extremely aggressive Rydex shorts, there appears to be quite a bit of fuel that could be added to an upside explosion if (when) it comes.
There will be plenty of potential catalysts this week, and one of them I touched on this weekend. So far, we're going according to the (admittedly tenuous) plan I laid out regarding State of the Union addresses. The market seems to react poorly in advance of the speeches, reflecting the hesitancy usually evident in front of important scheduled news events when we're in a downtrend. Recently, however, the market has tended to recover within two days after the speeches are given. This suggests that we may find a short-term low sometime tomorrow (if we haven't already done so). Refer to Sunday's weekly commentary for the details.
As I said above, I will be looking for a violation of a previous day's high tomorrow and Wednesday to go long in the model portfolio. I may conceivably take an entry before that time if we get a large gap down similar to this morning, so that is a possibility as well. In the intermediate-term portfolio, I further pared our short position into the gap down this morning and we are left with 1/4 of our initial position with a stop above last Thursday's high at the moment. Since my belief is that there is still a considerable amount of longer-term optimism to be wrung out of the market, I would look for any long trades taken in the short-term to eventually give way to another short position in the intermediate-term.
- Jason Goepfert
Disclosure: long QQQ puts
This disclosure is not intended as trading advice in any form. It is meant as a note to subscribers that the author may have a position directly affected by the market outlook reflected in the commentary. Although the author takes great pains to remain objective in any commentaries, it is only fair that readers should know that the author may have taken positions in accordance with his market outlook. Positions can and do change at any time, without notice to the reader.
© 2003 Sundial Capital Research, Inc. All Rights Reserved.