Minyanville recently reported:

“When the Dow closes below its December closing low in the first quarter, it is frequently an excellent warning sign. Jeffrey D. Saut, Managing Director of Research, and Chief Investment Strategist at Raymond James, brought this to our attention a few years ago. The December Low Indicator was originated by Lucien Hooper, a Forbes columnist and Wall Street analyst back in the 1970s. Hooper dismissed the importance of January and January’s first week as reliable indicators. He noted that the trend could be random or even manipulated during a holiday-shortened week. Instead, said Hooper, ‘Pay much more attention to the December low. If that low is violated during the first quarter of the New Year, watch out!’

Thirteen of the 27 occurrences were followed by gains for the rest of the year – and twelve full-year gains – after the low for the year was reached. For perspective we’ve included the January Barometer readings for the selected years. Hooper’s ‘Watch Out’ warning was absolutely correct, though. All but one of the instances since 1952 experienced further declines, as the Dow fell an additional 10.5% on average when December’s low was breached in Q1.

Only three significant drops occurred (not shown) when December’s low was not breached in Q1 (1974, 1981 and 1987). Both indicators were wrong only three times and six years ended flat. If the December low is not crossed, turn to our January Barometer for guidance. It has been virtually perfect, right nearly 100% of these times (view the complete results at www.stocktradersalmanac.com).”

-The Stock Trader’s Almanac

The December low in question was 12,194, and it was breached by Friday’s closing low. The fact that it was a weekly close probably added some additional significance for the chartists out there. Our readers know we have been bearish, particularly on tech stocks. We looked pretty dumb there for a while, less so now. We’re more interested in finding some good stocks to buy than in looking smart, however.

The major report on the Economic Calendar this week is the employment report on Friday. Consensus expects 95,000 jobs added on a month/month, seasonally adjusted basis. We think that basis is full of beans, and will be expecting something on the order of 2 million jobs (plus or minus a hundred thousand or two) on a year/year, unadjusted basis.

The earnings calendar is also light. None of our Watch List names are scheduled to report. However, Finisar (FNSR - Monday), National Semiconductor (NSM - Thursday) and Tech Data (TECD - Thursday) should offer some further insight into the technology outlook.

For some reason, the latest fashion appears to be issuing convertible notes. The practice allows companies to reduce the apparent share count, but assuming the share price rises (which stockholders are presumably rooting for) the shares willl just come right back. We’ve seen this with Ceradyne (CRDN) and Finisar (FNSR), and now Xilinx prices $900 mln 3.125 pct convertible debentures | Reuters.com:

Xilinx Inc. (XLNX) said it priced $900 million of 3.125 percent convertible junior subordinated debentures due in 2037, and expects the sale to close March 5.The debentures will initially be convertible into shares of Xilinx’s common stock at a conversion rate of 32.0760 shares of common stock per $1,000 principal amount of debentures, the company said.

At recent share prices, the company should be able to buy back between 32 and 33 million shares with the proceeds. If the company does well and the shares rise above $31.18 before 2037, the company will then exchange the bonds for approximately 29 million shares. So, for a net reduction of 3-4 million shares the company will be paying a 1.725% premium (the convertible yield less the dividend yield on the stock) for 30 years.

The additional interest expense relative to the dividends that would otherwise be paid amounts to $15.5 million per year. That would allow the company to buy 600,000 shares annually at the current price, which would give them the same net share reduction after just 5 or 6 years rather than the 30 year term of the convertible bonds.

Or alternatively, the company could buy back shares with the $300 million it used in the December quarter to increase its balance of short term investments. That would allow them to buy back 11.5 million shares and retire them permanently, never having to worry about the dilution again, and never incurring the interest payments on the debt. Over time, that would quadruple the net share buyback and they would recover more than 100% of the initial outlay in interest savings over the life of the convertible. If, for some reason, the company then failed to generate sufficient cash flow to run its operations (which appears to be an unlikely outcome) they could always then turn to the convertible securities market for a cash injection.
Of course, the company won’t have to reissue the shares if the stock does not rise above $31.18 by 2037. But we’re pretty sure shareholders aren’t interested in that potential outcome. No matter how you slice it it is hard to see how this deal is positive for existing shareholders.

In honor of Stock Market Beat’s birthday next week, we are looking through the archives for some of our best posts. As it happens, several of them investigate whether companies are using aggressive accounting practices. Who’s cooking the books?
Is it DELL?

Is it Xerox?

Is it Fidelity National Information Services?

Is it Ceradyne?

Or maybe Finisar?

The comment box is open.

Finisar Corporation’s earnings release started out sounding quite upbeat:

Finisar Corporation (FNSR), a technology leader in gigabit fiber optic solutions for high-speed data networks, today reported record quarterly revenues for its fiscal second quarter ended October 29, 2006 of $108.2 million, an increase of 24.9% over the second quarter of last year, marking the thirteenth consecutive quarter of revenue growth and the ninth consecutive quarter of record revenues.

Sounds great, until one realizes that the consensus expectation was $110 million in revenue. Perhaps the company is making up the revenue shortfall with cost reductions? We don’t know, because their backdated options are going to force a charge, the amount of which they haven’t figured out yet. Not surprisingly, the stock declined significantly after hours.

The revenue shortfall perhaps is unsurprising, as it appears there is adequate capacity out there to cover current demand. The slowdown is not a one-time phenomenon, either, judging from the guidance offered on the conference call:

For the third quarter we expect total revenue to be in the range of 108 to 112 million with Optics revenue in the range of 99 to 102 million and Network Tools revenue between 9 and 10 million. Non-GAAP gross margin should be between 38 and 40%. Non-GAAP operating expenses should be approximately 30 million. DSO is projected to be 55 days for the remainder of the year. Inventory turns are projected to be between 4.7 and 5 for the remainder of the year. Available cash is expected to increase 5 to 10 million to 135 to 140 million. Our fiscal year 2007 guidance remains unchanged with total revenue expected to be in the range of 430 to 460 million. Non-GAAP gross margin of 38 to 40% and non-GAAP operating expenses of approximately 120 million.

Too bad consensus called for $113 million in Q3 and $450 for the year. It’s a good thing the company recently engineered some cash flow flexibility with their shareholder-unfriendly note exchange. The way things are going, they’ll need all the flexibility they can get.

The headline for the durable goods report was not very bright this morning - Durable goods orders slid 8.3 percent in Oct - Yahoo! News

New orders for U.S.-made durable goods tumbled much more than anticipated in October on a big drop in civilian aircraft but were also down unexpectedly when transportation was stripped from the total, a government report suggesting economic weakness showed on Tuesday.
Durables goods — big-ticket items expected to last three years or longer — fell 8.3 percent, the biggest drop since July 2000. The decline was propelled by a 21.7 percent fall in transportation orders, the
Commerce Department said.

But even excluding transportation orders, durables declined 1.7 percent as manufacturing, fabricated metal, and computers and electronics orders all slid.

As is our custom, however, we like to dig a little deeper into the data to determine whether there are any bright spots. Instead, what we saw for the most part was rising inventories and falling orders and shipments. All charts below are based on information provided by the U.S. Census Department and collected by Stock Market Beat.

Although last month we said it looked like it was time to play defense, even that sector is giving back much of its strength.

Also consider computers and related products. Both shipments adn orders are falling through the floor. The potential bullish case is that customers are holding off on equipment upgrades in anticipation of Microsoft (MSFT) Windows Vista. However, given the strong recent performance at Dell and Hewlett Packard (HPQ - Annual Report) this bullish case may be priced in already. That could leave investors holding a heavy bag if the Vista orders don’t come in as expected.

Communications equipment, which had formerly been bucking the trend in technology, has also seen a sharp reduction in order growth.

Semiconductors may be the bright spot, but at this point it seems too early to tell given that the one strong data point is balancing several weak ones. At the least, the strength could support the argument that slowing computer sales and orders are Vista related, and that the semiconductors will be needed to build computers in a few months.

Electrical equipment and appliances also look strong, which is probably due at least in part to strong holiday sales of flat-panel televisions.

There seem to be few places to hide.

A comment on one of our recent articles took issue with our focus on Corning’s (GLW - Annual Report) LCD glass business, saying “It’s telcom division is just at break even now. When the youtube driven video over ip takes over, demand for fiber to the home will explode.” One reason we disagree is this recent press release from Level 3 Communications (LVLT):

Level 3 Communications’ Content Markets Group today announced a multi-year agreement with Photobucket to support the company’s growing online personal media sharing service. Under the terms of the agreement, Level 3 will provide Photobucket with network solutions including Level 3® High Speed IP service via Multiple 10 gigabits per second (10GigE) ports.”With over 28 million global users and adding 80,000 new users per day, Photobucket needs a robust, reliable and highly scalable network provider,” said Darren Crystal, chief technical officer and co-founder of Photobucket. “Photobucket has worked with Level 3 since our company’s founding, and we are confident that they will continue to meet our needs.”

“Photobucket, an online personal media site that enables users to manage and share their digital lives, delivers billions of images and videos across nearly 250,000 different websites every day. An additional 7 million personal images and videos are uploaded to Photobucket.com daily. This relationship gives Photobucket direct access to Level 3’s Tier 1 global network,” continued Crystal.

The fact that Level 3 considers this newsworthy tells us that there is still plenty of fiber capacity out there from the bubble days.  As Om Malik notes:

It is all spin and a blatant attempt to get a little Web 2.0 pixie dust. In fact, Level 3 is spending liberally to get it. They are sponsoring the Web 2.0 conference, and paying top dollars for it. I wonder why they are not one of sponsors (or even an exhibitors at) ISPCon, a conference that is closer to their core business.

Until that YouTube demand starts eating up the old capacity, demand for new capacity is likely to remain muted.

Finisar’s (FNSR) announcement that it was exchanging some of its convertible notes for contingent convertible notes generated some interest in our case study on contingent converts. Given the interest, we figured we would reward it with a study of the specific Finisar transaction. We’ll start with the company’s spin:

Overall, the exchange provides the Company with more flexibility to utilize its cash flow from operations between now and 2010, while also minimizing dilution to shareholders.

If you think this sounds too good to be true, you would be correct. Let’s start by looking a little more closely:

The New Notes contain provisions known as net share settlement which require that, upon conversion of the New Notes, Finisar will pay holders in cash for up to the principal amount of the converted New Notes. Any amounts in excess of this cash amount will be settled in shares of Finisar common stock.

What this means is that the old notes were convertible into stock (270 shares per $1,000 of bond principal) or the company could settle in cash if it preferred. The new notes, by contrast, require the company to settle the first $1,000 of each bond’s ending value in cash rather than shares. This does not provide the company “more flexibility to utilize its cash flow.”  In fact, just the opposite. So we can surmise that some cash flow covenants were restricted “between now and 2010″ to give the company more flexibility in the short term. Specifically:

The New Notes do not contain the put option provisions of the Outstanding Notes which provide the holders of those notes the one-time option to require the Company to repurchase the Outstanding Notes on October 15, 2007.

So instead of being required to repurchase the notes next year, they get a reprieve until the notes expire in 2010.

Now let’s tackle the dilution aspect. The company tells us:

The New Notes also are convertible into 35 more shares of Finisar common stock per $1,000 principal amount than the Outstanding Notes.

Hmm. Instead of 270 shares, the bondholders can get 305 shares worth of value for their bonds. That doesn’t exactly sound like “minimizing dilution to shareholders” to us. What they really mean is that, because the principal will be settled in cash, the reported diluted share count will be lower. This is just accounting sleight-of-hand. The economic reality is that bondholders are getting more value for their money, and this value will come directly out of shareholder’s pockets.

What Finisar has done is this:

1. They agreed to settle the first $1,000 of each bond’s value in cash in exchange for being able to reduce the reported (not the economic) dilution to shareholders.
2. They sweetened the notes by making them convertible into 35 additional shares (a $125 value per bond.)
3. In exchange, bondholders have to wait until 2010 to cash in the bonds instead of having the option to do so next year.

Who do you think got the better end of this deal?