Wednesday, June 27, 2007

Bulls (Finally) Bounce Back

The bears are still in charge, but they sure are doing it the hard way.

The NASDAQ gapped down to its 50-day on weak manufacturing data, then reversed course and put in its best day in over 9 months. The resulting outside day reclaimed 2600 and erased nearly three days of technical damage in just 6 hours. Volatility works to the upside as well.

The problem is that volume increased less than 1%. Although the mainstream press called Wednesday's action "window-dressing", the mediocre trade suggests that startled shorts fueled a big chunk of the rally. In truth, countless portfolio managers enjoyed a lucky technical bounce.

Because of this, the NASDAQ remains under Monday's Amber Alert. It stays that way until a follow-through day on big volume takes out the old high of 2631. Of course, that's just 1% away, and on the NDX, it's even less than that.

The recent strength in leading stocks is the most credible evidence that further highs are likely. The IBD100 lagged early on Wednesday, but came on strong in the final two hours to match the Composite's 1.2% gain. The critical tell was that the IBD100 saw genuine buying on Wednesday and not just panicked short-covering. 30 IBD100 stocks printed accumulation days, vs. just 17 on the OEX and 16 on the NDX.

Considering the litany of woes facing the financial markets, the major stock indexes continue to hold their own. The indexes are all back above their 50-day, and the NASDAQ's trampoline move was textbook. There's a lot of work ahead before the market looks healthy again, but the bears inexplicably squandered another easy kill.

Image Hosted by

In the "Erasing-Technical-Damage-in-Just-Six-Hours" Department, option volatility clearly wins the prize. I've made caffeine and methamphetamine jokes about the VIX, but after 16 straight days of intraday price swings ranging from 7% to 18%(!), it's not so funny any more. Instead, it has grown disturbing.

What it's saying is that there's tremendous disagreement among option investors about near-term stock market risk. The troubling part is that the intensity of these differences isn't fully visible in the stock market itself. Based on when this recent volatility began, it would seem that "bailing out" a hedge fund is viewed by the option community as a much nastier process than parties are letting on.

Image Hosted by

In Wednesday's durable goods report, the data showed technology as one of the few areas of business spending that's growing. This is consistent with what the stock market has been saying as well.

The chart below compares the Tech Index with the SPX. As you can see, for the past four weeks tech stocks have accelerated rapidly to just below a 15-month high vs. the blue chips. Regardless of how you choose to measure it, technology is steadily resuming a leadership role. At the very least, build a tech watchlist.

Image Hosted by

Everything bounced today, but stocks with big surges in volume are worth particular consideration. These are stocks under accumulation, and they tend to be more disease-resistant. This is an important quality in malarial times like these.

The big news tomorrow is the FOMC policy statement. There's lots of chatter about the Fed officially adopting an inflation metric that includes food and energy. This change seems unlikely, but you never know. Given the current volatility, a change wouldn't be viewed as "market friendly" near-term, regardless if it help the Fed do their job better or not.

Expect continued chop, and maybe even more buying. A 2:15pm fireworks show is a real possibility, and as long as you follow the volume, you should do OK.

I'm out most of the day tomorrow, but will check in later in the evening.

Until then, happy trading.




I.L. said...

Hey dk, great post as always.
I'm just curious, how do you define a distribution day for a stock. Is it when down volume is greater than up volume?

dk said...

Thanks, IL...a distribution day is one in which a stock or index declines on volume greater than the previous session. Not all distribution days are created equally, and the real problem occurs when they appear in clusters. Hope that helps.