TUESDAY, AUGUST 12, 2008
S&P, Nasdaq Divergence Typically Resolves Upward
08/12/08 4:15 PM EST
I'm not a big fan of divergence analysis, since it can be very subjective and hard to pinpoint for timing purposes. There are exceptions, however, with some of the most consistent ones being standout moves by one index over another. We've gone over these in various forms between the S&P 500 and Nasdaq 100 many times over the years, and we witnessed another unusual one today with the S&P getting hit fairly hard while the NDX clung to positive territory.
The last time the S&P 500 fell 1% or more while the Nasdaq 100 finished in positive territory was July 15th, the day of the market low. Historically (back to 1985), such wide divergences on a daily time frame have been quite positive - the following day, the S&P was up 8 out of 9 times by an average of +0.8% (the one loss was -2.6% but was more than made up over the next two days).
Three days later, the S&P was still up 8 of 9 times but the average return climbed to +1.1%, then after that the returns and percentage of time positive fell more in line with random. The performance of the NDX tended to lag the S&P after these instances, showing positive returns only about half the time, with moderately positive average returns. The exception was five days later which showed that index up 7 of 9 times by an average of +2.8%.
The only time I can find the NDX in positive territory on a day the BKX Banking Index got hit for a 5% or greater loss was June 16, 2000. The NDX was up nearly 4% the next day (volatility was much higher then), but then it gave those gains back and basically went nowhere for the next month.
Gold got hit again today, which might be troubling for those still considering whether the commodity is still in a bull market - this is not what we usually see in a healthy market. While pessimism towards the metal isn't quite extreme as we discussed this morning, the HUI Gold Bugs Index was clearly oversold heading into today and it enjoyed a nice bounce in response. The only other time in the past 12 years I can find where Gold went down 1% on the day while the HUI Index surged 3% was on April 30th of this year, which happened to mark a short-term low for the index.
I neglected to mention assets in the Rydex Precious Metals fund this morning, which dropped below $150 million for the first time in three years yesterday. The fund now makes up only 11% of all Rydex sector assets, one of the lowest concentrations since 2003. Our Public Opinion towards Gold will be updated this evening, so we'll have a better feeling for how the latest decline is being taken by fans of the metal.
Rydex traders have undergone something of a sea change in their asset allocations. We're seeing massive movements of funds into the Healthcare, Biotech and Strong Dollar funds, at the expense of Basic Materials, Commodities (which has lost two-thirds of its asset base in a month) and Energy (which is skirting its lowest levels of the past four years). Whether these traders are making these moves at precisely the wrong time as usual is still up in the air, but their record at calling major trend changes is spotty at best.
The stock indices, despite the selling pressure today, have held up fairly well and the S&P is still right at its breakout level of 1285ish that we went over this morning. The selling today was enough to move our most sensitive indicators towards or slightly into oversold territory, so we're at another interesting short-term juncture. The NDX remains relatively overbought as we went over this morning, but now we get some oversold readings for the S&P (albeit on a shorter time frame).
That makes for a confusing mix, and I'm not sure which will win out. My preference is to look for long-side setups as long as the S&P is acting as well as it has been, and the current "oversold on support" setup is the type of thing I want to see. I'm just not sure if one day of selling is enough, especially given the overbought readings on the Nasdaq. Bottom line, I'm not adding to any long positions here, rather I'm going to wait and see if we get more of a shakeout first.
Nasdaq, Gold Bugs On Opposite Ends
08/12/08 9:05 AM EST
Good Tuesday morning...We begin the day with mixed action in the pre-market futures as the news flow is again fairly quiet. Commodities are little changed this morning, as are foreign equity indices.
It's no secret that technology and small-cap issues have been leading the recent advance. After pulling their weight, it's natural for them to be pushing the limits of overbought types of indicators.
According to the Up Issues Ratio for the Nasdaq, issues on that exchange are exceptionally overbought. With a 5-day average just under 60%, a 10-day average at 55% and a 21-day average at 52%, each of the three averages are above their "too far, too fast" trading bands.
The last time we've seen all three averages this high was October 16, 2006 after which the Nasdaq 100 went into a several-week trading range. Going back to 1985, go-forward performance in the NDX was mixed - sometimes it perfectly highlighted times that index was about to top about, other times it indicated the liftoff phase of a strong market that only escalated.
Over the past decade, these kinds of overbought readings have not been kind. The next day, the NDX was up only 23% of the time (5 out of 22 instances) with an average return of -0.8%. By three days later, it was up 36% of the time but by a week later that was at 50% and the further out we look, the more in line with random the chances became that the market would be higher.
However, average returns were still well below average even when looking out a month (-0.8% after these overbought conditions versus just about even for a random return). As that index touches the 62% retracement of the June decline, and the level where it broke down earlier this summer (around 1950), it would be unusual to not see some selling pressure with these kinds of overbought readings.
At the opposite end of the spectrum, we have the HUI Gold Bugs Index. Gold traders have been waiting for a long time for oversold conditions in that group, and like usual now that they've got it, they're not sure they want it.
Going back to 1997, I can find only six dates that show an equivalent level of oversold readings like we're seeing now in the charts we post to the site, based on the percentage of components in the HUI Index that are above their 10-, 50- and 200-day moving averages and the number of new 52-week lows in the components of the index.
The chart below is a composite of the two that we update daily on the site.
Those previous dates for conditions as oversold as the current one were 10/27/97, 11/26/97, 08/26/98, 10/24/00, 05/07/04 and 05/13/05. The last three were all excellent times to get long gold stocks on a short- and intermediate-term time frame, which is almost always the case when we see exceptionally oversold conditions during on ongoing bull market. The 1998 occurrence was a little dicey at first, but led to a 30% rally over a month's time shortly thereafter.
The two instances in 1997, however, were within the midst of a bear market, and simply led to further losses. That's what gold traders/investors need to decide now - if we remain in a general uptrend, then these kinds of conditions are what one waits for to establish or add to long positions; if not, it's just a recipe for loss. The simple trend indicators I use are very similar to what we saw in early January for the S&P 500 - if we reverse to the upside soon, then the index may stave off bear-market status, but we're on the precipice here and it better rally soon.
While the breadth of the HUI Index is clearly oversold to an extreme degree, we're not seeing sentiment reach the same levels of pessimism in the metal itself. Speculators, both large and small, are still quite net long Gold and have just started to come off positions that were near multi-year highs - that kind of speculation will take weeks more (at least) to dissipate. Also, Public Opinion towards the commodity is currently at 64% bullish, about in the middle of the range of the past few years. If it got down to 55% or below, then we could make an argument that we're seeing some actual negativity about Gold's future.
Yesterday was the first time since 1997 that Gold dropped at least 3% to at least a six-month low, and it's only the second time since 1990. Going back to 1980, there have been 17 such instances, and the metal bounced back the following day 13 times by an average of nearly +1%. That percentage of time positive and average return were pretty consistent up to 10 days later, before the returns returned to be more in line with random, so there was at least a moderate tendency to see some short-term bounce-backs from swift declines to new lows like yesterday.
Back to stocks, the S&P 500 tracking fund, SPY, was up over 1% again yesterday, marking the third time in a month that the fund has seen consecutive back-to-back 1% or greater gains. The last time we saw that kind of buying pressure was August 2002.
Going back to 1950, there have been 13 times when the S&P 500 cash index saw three instances of back-to-back 1% gains within a month's time. Returns going forward were mixed up to a month later, but by three months later it was positive 11 times by an average of +5.7%; six months later 10 times by +9.2% and one year later 10 times by +15.2% - all of those are at least double random returns during the study period.
Sentiment-wise, we're not seeing many extremes at all at this point. Our Indicators At Extremes on the Daily Overview page continues to shrink on both sides of the ledger, and now has an equal number of bullish and bearish extremes. On a very short-term basis, the STEM.MR Model for the Nasdaq 100 is above its overbought trading band, which is no surprise given what we looked at earlier in this note.
So we should see prices back off in the coming session(s) as some of these readings get worked out, but I don't see any good setups that would suggest shorting here has a good risk/reward ratio - particularly as the indices remain above their breakout levels of 1285ish on the S&P and 1880ish on the NDX. As long as we remain above there (or really above 1250 for that matter on the S&P), my focus remains with trying to find acceptable places to initiate/add to long positions.
All the best,
President and CEO
Sundial Capital Research, Inc.
Forwarding or otherwise distributing this copyrighted material is a breach of your subscriber agreement. Violators are subject to termination of their subscription with any received subscription fees forfeited. Any references to historical performance are based on data we deem to be reliable, but are based upon feeds from third parties. We do not recommend subscribers take positions based on data presented here alone, but rather incorporate it into a comprehensive investment outlook.
© 2008 Sundial Capital Research, Inc. All Rights Reserved. www.sentimenTrader.com