Saturday, July 28, 2007

Bottom Quest

The worst thing about last week is that the there are so few signs that a bottom is near.

Both the mainstream press and the blogosphere offer some excellent reading this weekend. There's also large amounts of windbag speculation and a fair amount of bellyaching. While this all makes for titillating journalism, an evidence-based approach to the market tends to be more productive for your portfolio.

As investors ponder a sea of red, below are five sets of market-based observations:

1. Panic

All rumors to the contrary, there was no evidence of panic selling this week. Even though the Dow shed 585 points in its worst week since 2003, trading was orderly, and from an historical perspective, the declines were chump change. This correction might accelerate into panic, but it hasn't happened yet.

On the left below is a chart of this week's NASDAQ. On the right is the Composite during the LTCM scandal in the summer of 1998. From the July high in 1998, the NASDAQ tumbled 33% in 12 weeks. In the last two weeks, the NASDAQ skidded a sickening 23% (imagine the Dow falling 3100 points in eight days).That felt panicky, especially because --doh! -- everyone thought the bottom had occured four weeks earlier (red arrow).

Image Hosted by

2. Bottoms

I received numerous e-mails, comments and questions this week about identifying bottoms. Of particular interest was whether the IBD100 offered an edge. The short answer is no, the IBD100 is essentially useless at calling bottoms. Most leading stocks are bloody stumps at market lows. Most recover; many don't.

However, other indicators are helpful. When these are used in combination, they're generally accurate (but look again at the red arrow on the 1998 NASDAQ chart just above).

While some are crying oversold! this weekend (and the odds for a ST bounce next week are actually pretty good), almost all of the popular "bottom finders" show a market still above a Queen-approved bottom. Below are three classic proximity alerts. There are others, but those aren't there yet either.

Despite the HUGE surge in New Lows this week, the blue 10-week on both the NYSE and COMPQ High-Low Indexes remains well above important bottoms. Below are 5-year charts of the High-Low Indexes, which include the Oct 2002 bottom as reference. Historically, the blue 10-week line crosses below 40 -- and even below 30 -- before a meaningful bottom is in.

Image Hosted by

The percentage of NASDAQ stocks trading above their 50- and 200-day is still high for a bottom. NYSE versions of this metric produce similar results.

Image Hosted by

NASDAQ market breadth as measured even by the unforgiving NASI is still above the rising trendline off the Oct 2002 bottom. NYSI offers a similar take.

Image Hosted by

3. Crash Alert?

Regular readers know that I'm hardly an alarmist, but -- believe it or not -- important market evidence emerged this week that raises the chance for a 1998-type stock market crash. First of all, we're not there yet and the odds are still long. [Re-read the previous sentence]. However, as I was preparing notes on this, I read Adam at Daily Options Report quoting Jason Goepfert on the same thing.

NYSE New Lows spiked to severe levels this week. In fact, New Lows have been this high only four times in the past 10 years. Making matters weirder, the NYSE is just 7.1% below an all-time high.

Also, the Dow wasn't the only chart that set a 4-year record this week. The VIX surged an ear-popping 43% to a new 4-year high, even though the SPX fell a relatively meager 4.9%. In fact, the SPX is up 6.9% from the March low, but instead of falling, the VIX has skyrocketed 160%!

The VIX is now parked 36% above its 10-day, a very unstable spread for the VIX. This week, option traders priced in something ferocious -- and still unseen. Given the New Low spike and a 24-handle VIX, my observations mirror Goepfert's about what stock market history says are three plausible scenarios:

--- the worst is over, and it's time to go long for the mother-of-all short covering rallies

--- this isn't just a correction, but the 4 1/2-year bull market is done and we're grinding lower into a lengthy bear market

--- the market detects something bad on the horizon, and we're about to see a rapid, convulsive selloff

The odds of a crash are long, but the market itself has thrown these scenarios onto the table, and not just the guys at Safehaven. Below is a 10-year chart of NYSE New Lows, along with the VIX.

Image Hosted by

4. Decent Economics

Juxtaposed against the doom and gloom is that the underlying US economics are surprisingly good. GDP is recovering, earnings are solid, valuations are nominal, employment is high, inflation is stable, and yields are low. If the market does plunge wildly lower, the economics point to it being a superb buying opportunity.

It's worth noting that, as subprime contagion and a rising yen threatens the US economy, the 2-year/10-year yield spread detects nary a whiff of recession. In fact, it's printing a 3 1/2-year low.

Image Hosted by

5. The Penalty Box

Below are the six worst performing groups this week. Housing is no surprise, but it's noteworthy that the metals continue to offer no refuge despite a dollar perched at the abyss. The next two losers -- energy and cyclicals -- are the most troubling economically. Utilities took out the 50- and 200-day in five sessions, and small caps lived up to their reputation as the style to avoid.

Image Hosted by

Pullbacks separate the wheat from the chaff, and one useful feature of this correction is that it started during earnings season. Companies that beat estimates are holding up better than the broader market, and the fresh earnings data is useful in preparing high-quality watchlists. No single investing strategy works for everyone, but stocks that pull back the least are typically the ones first out of the gate when the selling stops. Historically, they also run the farthest.

The week's best performers? Biotech, down 1.8%, followed by the NDX, down 3.9%.

Also, it's useful to track which stocks closed higher on a week like this. It's a great set of stocks with which to start a new watchlist.

--- On the NDX, 13 stocks closed higher on the week: AAPL, AMAT, AMLN, AMZN, BIIB, BMET, CELG, CHKP, GRMN, GENZ, ISRG, TLAB, VRTX, WYNN.

--- On the OEX, 6 stocks closed higher on the week : CL, IBM, MRK, PEP, PG, T

--- On the MID, 23 stocks closed higher on the week: ADVS, AJG, BEC, BRL, BRO, BSG, CCMP, CVD, DCI, EXBD, FFIV, GGG, GPRO, MFE, PAS, PLT, RFMD, ROL, RSG, VARI, WOOF, WPO

It should be a lively week ahead, as investors learn how good they are at following the ball in the other direction.

Hope everyone's having a great weekend.




Bill Luby said...

One of your best yet, dk. Bravo!

dk said...

Thanks, Bill.

dan said...

Check the weekly chart and look at the volume on AMLN too. Something's up. Some obesity trials with Symlin and with Symlin + Leptin are due before the end of the year, plus LAR Byetta is due that time too. Could be big money placing bets on that, but then, can't they just "quietly accumulate"? Something must be up near term. My guess is Byetta scrips will break out this week or next, as this is about 1-2 months after the Avandia news and the American Diabetes Association meeting. Switching from Avandia to Byetta would likely require a doctor visit so Byetta didn't show the immediate pop in scrips that the oral drugs did. Also, AMLN gives the first month free so new users dont' show up for a month. Also, about a month ago, there was huge call volume on January 40 and 45s. This volume was on two days where the pps was flat to down, and these calls were going up by $.05-.15. Who in their right mind would buy millions of dollars of calls with conviction, for 6 months away unless they thought the pps wasn't at a near term low? You could say they were playing this latest move of upgrades (Tuesday) but then, why pay the January time premiums?

(tamasad from iv)

beam said...


There is another market bottom indicator for reference: 90% downday.

The volume figure is available from Yahoo financial market stats. If anyone knows where to get point figure, please post.

Anonymous said...

Good post dk. Regarding #1, what would indicate to you that panic selling is/has taken place in the market?

dk said...

dan...I've been an owner of AMLN on and off over the years, and keep an eye on it. After a tough 2006-2007, it seems to be working itself up the right-hand side of a long, 18-month base. Interesting observations about the reasoning behind the move. Nice stock move...

beam...another interesting link! Thanks for the Lowry PDF. The 90% selling climax is also a Marty Zweig idea, but the Lowry doc provides Interesting detail and perspective. Thanks for your comments, and I appreciate you stopping by. know it's panic selling by the % of the drop. If you look at the 1998 NASDAQ chart again, you'll notice that the last red candle is a -12.1% drop in just four sessions. In fact, during this period, the Composite dumped 23% in eight sessions.

adam said...

Great stuff.

Anonymous said...

dk, you have a phenomenal blog! I think you are very generous with your time, knowledge and expertise, and I just wanted to say thanks!

Bill said...

Something to think about re: panic and percent down. The 5%+ fall for the SPX over a 5-day period is in the bottom 1% of all 5-day periods since 1950 in terms of gain. If a move in the bottom percentile of all 5-day moves doesn't qualify as "panic" perhaps your definition is too strict.

dk said...

Hi Bill...thanks for the comment. The term panic has an arbitrary nature similar to pornography...hard to define, but I sure know it when I see it - lol. It didn't seem like panic to me, but I skew a little dark.