A Profitable Pattern

I have noticed a profitable pattern.  I am going to cast this from the bearish perspective, but in theory it could work either way.  I have a few other historical examples, but let us focus on the present.

Background

Investors have many time frames, and so do those managing money.  When you see someone interviewed on CNBC it could be a mutual fund manager, a hedge fund manager, or a trader.  It is natural for everyone to talk his book, but the time frame is relevant.

A mutual fund manager likes to see holding increase, get publicity, and attract new assets.  Having said this, these managers are not likely to sell into rallies.  Instead, new money in their funds gets invested at current prices in current holdings.

A hedge fund manager or trader is different.  Their disclosures warn that they may change opinions and trades in a heartbeat.  They can, and should, do so.  A five percent move in a stock is a major trading profit.

The Process

A portion of the hedge fund community takes a position.  For this example, it is a short in semiconductor stocks.  The position is supported with plenty of reasoning they deem to be rigorous and also by technical analysis.  It is especially desirable to find a situation where a sector has made a big run.  A chart of the SOX tells the story.

I haven't added any lines in this, and readers should ignore the last decline, since that is the end of the story.  The point is that analysts could have a field day with this chart — before the decline.

Cocking the Hammer

Those establishing a bearish position wish to highlight major concerns.  In this case the "hammer" was the question of double booking of orders.  This occurs when inventories are so low that those building mother boards want to be assured of needed parts.  To do this, the theory holds, they place multiple orders for needed parts in short supply, canceling some orders when their needs are met.

It is an argument that raises a question about the validity of reported orders, and creates fear that the orders will decline.

Pulling the Trigger

One might think that actual evidence of order decline would be necessary to trigger the fear.  That assumption is incorrect.  It is easy to grasp any news as evidence that this prophecy has been fulfilled.

Yesterday the B of A/Merrill analyst downgraded the entire sector.  The resulting report, which I have read carefully, is based upon an inventory model and some assumptions about where we are in the cycle.  It does not specifically refer to any double booking, and the resulting forecasts are still above many other Street observers.  It was a big call, getting a lot of play, and knocking down the stocks by 5 percent or so, as the chart indicates.

Bearish observers incorrectly cited the double booking argument, not actually mentioned in the report.  It talks about a small "overshoot" in inventories, not mentioning any double booking.  It is based upon a theoretical model, without a lot of supporting evidence.  Since few actually read these reports, the inaccurate summaries gain credence.

The Result

Anyone with a short position can cover for a nice trade.  It was a nice setup.  Any downgrade could be cited as evidence.  When the analyst from a big firm downgrades, everyone pays respect, at least in the short term.  That is long enough for traders and hedge fund types.

The Facts

There is plenty of evidence contradicting the double-counting argument:

From iSupplyCorp we have the following:

In their earnings announcements, leading semiconductor suppliers,
including Texas Instruments, STMicroelectronics and Intersil, all
reported they saw no signs of double-booking during the third quarter,
said Ciriello. “This should give the semiconductor industry confidence
that the magnitude of the current recovery accurately reflects real
demand levels.”

A balanced article at the WSJ mentions the double-booking issue:

Another analyst, FBR Capital's Craig Berger, also cited signs of
weakening PC demand, though he said in a note, "We do not want to get
overly bearish on the PC space given improving global demand and
still-unknown holiday sell-through trends."

Berger said he remains "bullish" on the semiconductor
sector. "We think fears of a first half of 2010 estimate cuts from
double ordering are overblown, that global supply chain inventories are
at or near all-time lows, that global demand trends will continue to
recover in 2010." 

Astute tech stock observer Bob Faulkner writes as follows (see the full article for his nice inventory charts and complete analysis):

I've tracked inventory in the electronics-supply chain for years. While
it’s too early in earnings season to draw any final conclusions, I
think it’s very safe to say that inventory is quite tight on a relative
basis.

Steve Ballmer reports that Microsoft sees "fantastic sales" for Windows 7.  If this is accurate, it should stimulate a new cycle of purchases.  More people and businesses replace computers — especially those who avoided VISTA — than upgrade existing machines.  The current hardware may not be adequate and is difficult to configure.

Our Take

The trading setup is a real dilemma for managers with a long-run horizon.  It was completely predictable that something — anything — would be seized upon as confirmation for the double-booking argument.  There was no need to wait for actual evidence.  Even knowing this, should the manager blow out all of his chip holdings, expecting a pull back of unknown size?

The real answer for those with a longer perspective is to focus on the actual time frame and ignore the bumps in the road.  The traders will make their profits, and eventually the results will tell the story.  It is an opportunity to add to positions for those who disagree with the thesis.

An Interesting Exercise

There was a similar setup — hammer pulled back — on car sales after the Cash for Clunkers expired.  Perhaps this is a good topic for an example where we now have more evidence — an immediate decline, followed by a restoration in sales.

I invite other examples — both long and short — from readers.  Look for the setup — what we all should watch for– and then look for evidence.

Our position:  Long INTC, MSFT, and AAPL.