Russell Rebalancing Felt Far and Wide

Editor's Note: This is a bonus story from Alan Farley, whose commentary usually appears only on RealMoney. We're offering it today to TheStreet.com readers. To read Farley's commentary regularly, please click here for information about a free trial to RealMoney.

---

Friday's Russell rebalancing act triggered a flurry of last-minute spikes that left hundreds of stocks with bizarre, wide-range price bars. These spikes took place in both directions, with price shooting well above or below levels that were traded for most of the day. I spent the weekend looking at what happened and gauging the impact this event will have on affected stocks.

The "what happened" part is hard to determine from the outside looking in. We know money managers shifted around millions of dollars to accommodate the rebalancing. I've heard that many transactions took place through the Archipelago ECN (ARCA). Those who trade through ARCA understand how this operation could turn an orderly transition into absolute chaos. In their defense, I've seen no confirmation that they were the culprits behind the huge spikes.

Russell rebalancing has a nasty reputation for spiking prices, but this year's moves were off the charts. Somehow this occurred despite high-profile changes in Russell's reconstruction methodology. The real irony is that these modifications were supposed to dampen volatility at the time of the changeover, not send it through the roof.

Huge Numbers

Friday's statistics were out of this world. Looking at the top 50% of all exchange-traded stocks by volume, 274 had daily ranges greater than 10%. An astounding 115 had daily ranges greater than 20%. Compare these aberrations with Thursday's sedate numbers, when 124 stocks had daily ranges greater than 10% and a mere 12 stocks had daily ranges greater than 20%.

This phenomenon was confined to the Nasdaq. In fact, only eight of the 274 stocks moving 10% or more were NYSE-listed issues. At least in this case, that troubled exchange showed the value of the specialist system when it comes to supporting an orderly market. Their Friday book of business does show a minor uptick in response to the rebalancing, but it was nothing unusual.

There were real transactions taking place at crazy levels, at least according to the data centers. The above chart of Unitedglobalcom (UCOMA Quote) -- which isn't even a Russell component -- is one example. In fact, I cross-checked my numbers on TC2000, eSignal, Prophet.net and Stockcharts.com. All sources tell a similar tale: huge block trades in excess of 1, 2 or 3 million shares just before Friday's close. In many cases, transaction levels exceeded 200% of a stock's average daily volume.

I've debated the impact of "out of place" price bars with technicians over the years. In other words, can a price bar that doesn't match the underlying pattern really affect an ongoing trend? My argument is always the same: Appearances are far more important than reality when it comes to price prediction.

In other words, phantom spikes will affect trading decisions, even when caused by bad or ridiculous data. The reasoning is simple. There are legions of technical traders who make their decisions by looking only at the charts. These folks could care less about why weird price bars are sitting there. Their considerable influence becomes even stronger over time, as the story behind this particular trading day fades into history.

At its core, technical analysis is a chicken-and-egg thing. In a week, everyone will forget the reasons for the wide range bars and use them as swing pivots. In fact, they'll be hard to ignore for months and will affect virtually every signal that relies on daily range data. They'll also affect lists of new highs and new lows to some extent, though that's hard to measure at this time.

The flip side is that many downward spikes should help weaker price patterns build more stable-looking bases. This also suggests that short-sellers will be more reluctant to jump in and push these stocks down, because massive rinse and recovery bars are staring them in the face.

Let me clarify my observations with Cytyc (CYTC Quote), an active pick in my newsletter, The Daily Swing Trade. This stock was grinding out a breakout pattern until Friday's bizarre turn of events, as I wrote in a Columnist Conversation post over the weekend.

Now it shows a staggering graveyard doji, one of the most potent reversal signals in candlestick analysis. As long as this bar stands, I have little choice but to remove the stock from the newsletter.

That's not to say Cytyc won't eventually trade up and through Friday's spike. The problem is, the aberrant high will undermine legitimate demand for weeks or months to come. Consider all of the traders flipping through their charts in coming days looking for good breakout patterns. Many of them will see that doji candlestick and decide it's not the best place to risk their money.

I understand the supply and demand calculation may be bogus on these stocks, but I'm not sure it matters. What does matter is that there are hundreds of other stocks that don't show these crazy spikes. As a technician and stock-picker, I'd rather avoid the risk of these odd patterns and apply my capital to the vast pool of unaffected issues.

One final observation about last Friday drives me crazy. Neither of the stocks I've highlighted here are on Russell's addition or deletion lists. So I'm baffled how this event hit scores of unrelated stocks and unsuspecting shareholders.