The IPO has loomed so large over the market's landscape that it's easy to forget there's an alternative, less flashy route to going public: the one-step spin-off. With the IPO window shut and market valuations severely depressed, this form of divestiture -- in which a company simply distributes shares of a subsidiary to its own stockholders -- is coming back into vogue.

Untethering fast-growing divisions from the slower-growing whole theoretically unlocks shareholder value. That need, while slackening, has not disappeared with the IPO market itself. For instance, Cabletron Systems (NYSE: CS), which earlier this year sold a stake of subsidiary Riverstone Networks (Nasdaq: RSTN) through an IPO, is now planning a one-step spin-off for its network management business, Aprisma Management Technologies, and another for its wireless networking division, Enterasys Networks. Elsewhere, Adaptec (Nasdaq: ADPT), with its digital media software division Roxio (Nasdaq: ROXI), and Conexant Systems (Nasdaq: CNXT), with its Internet infrastructure business Mindspeed Technologies, shelved IPO plans in favor of one-step spin-offs.

The re�mergence of one-step spin-offs is a real sign of the times. During the IPO frenzy, a company could sell a small fraction of a subsidiary to the public, raising a huge amount of capital in the process. Today, with valuations much lower, a parent company would need to sell a higher percentage of a subsidiary to raise enough capital to attract the interest of big, institutional investors. But there's a limit to how much a parent can sell. To keep the IPO proceeds tax-free, Internal Revenue Service rules dictate that a parent company cannot sell more than 20 percent of a subsidiary. That leaves many would-be issuers in a bind.

Astute managers are therefore turning to the one-step spin-off, hoping to unleash the value of a fast-growing subsidiary. "One of the drivers of one-step spin-offs is maximizing shareholder value," says Paul Chamberlain, cohead of the West Coast technology banking group at Morgan Stanley. "Technology spin-offs have become increasingly common in situations where individual business segments are growing at vastly different rates than that of their parent."

AN EASY CHOICE For Roxio, which filed for an IPO in September only to watch its valuation drift steadily south, the IRS rules made the decision to go with a one-step spin-off easy. The company couldn't raise enough capital to make an IPO worthwhile and still retain a tax-free ruling from the IRS. "Because of the IRS rules, we wouldn't have been able to attract the quality of institutions that we wanted to participate in the IPO," says Roxio president and CEO Chris Gorog.

One-step spin-offs do have several clear disadvantages. Unlike an IPO, the process doesn't raise fresh capital, so the new company must rely on the parent company's largesse. Roxio, for instance, will have to live with a $30 million parting gift from Adaptec. Also, because Roxio had no IPO road show, management must work hard to raise its profile.

So how should investors view one-step spin-offs? A spin-off is going to be priced into a stock to some extent. But market inefficiencies can and do still exist, and Roxio, which analysts expect to earn 85 cents a share on revenue of $138.7 million for the fiscal year ended March 31, 2002, is a case in point. Through customers like Dell Computer (Nasdaq: DELL), Hewlett-Packard (NYSE: HWP), Philips Electronics (NYSE: PHG), and Yamaha, the company has a 70 percent market share of the CD-burning software that comes installed on PCs. That gives it a big edge over Ahead Software, based in Germany, which only recently signed on its first original equipment manufacturer, and Prassi Software USA, currently being sued for patent infringement by Adaptec and Roxio.

A keen investor could have bought into Roxio's public debut by purchasing Adaptec shares for as little as $9 in the week preceding the April 30 date of record for distribution. The board had set the Roxio dividend at .1646 shares per share of Adaptec. So, on June 13, that investor would own Adaptec, which closed at $9.32, but also Roxio, which closed at $11.95. That's a return of 59.4 percent (see "Dance lessons").

The IPO window might be closed, but a side door to the public markets is still ajar.

Write to letters@redherring.com.

September 26, 2000: Adaptec files Securities and Exchange Commission registration statement for a $57.5 million IPO of Roxio, a division that makes CD-burning software. Adaptec's stock closes at $19.06.

January 4, 2001: Citing market conditions, Adaptec withdraws the IPO registration statement. Its stock closes at $11.44.

April 12: Board approves spin-off and sets the Roxio dividend at .1646 shares per share of Adaptec.

April 26: Adaptec shares close at $9.07.

April 30: Roxio hosts first investor conference call. Adaptec shareholders of record entitled to receive Roxio dividend.

May 4: Roxio stock begins Nasdaq trading on a "when-issued" basis, closing at $15.99. Adaptec's stock closes at $11.71. Shareholders who bought Adaptec on April 26 see a 59.4 percent return.

May 11: Roxio shares distributed to Adaptec stockholders.

May 14:Roxio (Nasdaq: ROXI) shares officially begin trading.

Source: Bloomberg

Click here to return to main article.

In light of these developments, it would seem a perfect time for Global Power Equipment Group (proposed NYSE: GEG), a prime beneficiary of that new plant construction, to go public, right? Yes and no.

The near-term financial outlook for Global Power couldn't be better. The Tulsa, Oklahoma-based company's heat recovery steam generators (HRSGs) are used in combined-cycle power plants. Because the hot exhaust from the gas turbines that power these plants can be routed through an HRSG to produce more electricity, combined-cycle power plants are far more efficient than the simple-cycle plants that dominated the industry prior to 1999. Of the 146,000 megawatts expected to come online over the next three years, 98,000 megawatts will be generated by combined-cycle power plants, estimates Arnold Leitner, a consultant to the electric utility industry at Boulder, Colorado-based RDI Consulting.

IN DEMAND As a result, demand for the company's products is at an all-time high -- its current backlog of orders stands at $634 million, up 36 percent from year-end levels -- and stock valuations for comparable companies are soaring. So, too, is the short-term outlook for the company: after generating pro forma net income of $21.6 million, or 47 cents per share, on revenue of $447.8 million last year, the company should easily exceed those results in 2001. Worst case, we would estimate Global Power's 2001 net income at about $50 million, or $1.11 per share, on revenue of $648 million.

The problem, according to some industry observers, is that demand for HRSGs, which comprised 53 percent and 41 percent of Global Power's 2000 revenue and first-quarter 2001 revenue, respectively, may be unsustainable. Electric utilities are responding to current fears by building too much new capacity too quickly. If that's the case, Global Power customers such as General Electric and Mitsubishi Heavy Industries could cancel orders and leave public investors buying in at the peak of the market without realizing the potential bust cycle that may lie ahead.

"Right now we're seeing an unprecedented demand for heat recovery steam generator equipment," says Mr. Leitner. "But that won't be sustainable going forward because the demand for electricity is simply not there to meet these levels of capacity additions."

Joint lead-managers Credit Suisse First Boston and Salomon Smith Barney, which are looking to price 7.35 million shares on Thursday for trading on Friday, are aware that it may be premature to prey on investor demand for the latest hot industry. The Global Power IPO is expected to price at a substantial discount to the valuations of similar companies that are benefiting from new power-plant construction. Based on the $17 asking price, the company's shares would price at 13.5 times 2001 earnings, a huge discount to Shaw Group (NYSE: SGR), Jacobs Engineering (NYSE: JEC), and Fluor Corporation (NYSE: FLR), all of which are more diversified companies. Shares of Shaw Group, Jacobs, and Fluor trade at 36.2, 21.4, and 27.9 times 2001 consensus earnings, respectively.

The steep discount is entirely appropriate, in our opinion. Global Power is too focused on HRSG sales, a product category that will provide plenty of upside but offers little room for alternative sources of revenue should demand dry up. Moreover, management has few other products in development. All of the $113 million that the company is taking in from the IPO will be used for debt reduction, and its debt covenants cap capital expenditures at $10 million annually. After the offering, Global Power will have just $2.2 million in available cash.

By the end of the year, Global Power's IPO investors will begin to feel the effects from the over-construction of power plants. From 2004 through 2010, utilities will only need to add 5,000 megawatts of power annually to keep up with demand, estimates Mr. Leitner. Long before then, Global Power and its stockholders will realize that the boom will have gone bust. And, after all, don't we all invest in IPOs for the long haul?

Easy enough to say, but more difficult to live by in today's tech-heavy market. Outside of the investment banks themselves, Wall Street has had precious few opportunities to follow Peter Lynch's age-old maxim. However, Instinet Group's (proposed Nasdaq: INET) upcoming IPO and the wave of rival electronic communications networks (ECNs) expected to follow in its path will give The Street plenty of investment alternatives.

A subsidiary of the Reuters Group (Nasdaq: RTRSY), Instinet's trading platform is used by more than 700 broker/dealers and virtually all buy-side institutions. By eliminating the costs and time involved in executing trades through a market maker, Instinet has permanently changed the way institutions buy and sell shares, and the ECN rakes in commissions for each trade executed on it. As a result, the company's $369 million IPO is obviously attracting a lot of interest. "If these guys just maintain their market share, this is an extremely attractive deal," said one attendee of the company's road show in San Francisco last Wednesday.

Clearly it's more than the free lunch -- even though those are also rarities in these IPO-less days -- that draws packed crowds of buy-side investors to Instinet's road show. The company's share of Nasdaq-listed volume crossed 15 percent in the first quarter, up from 11.3 percent in the year-ago period. Instinet's share of NYSE volume also jumped, hitting 3 percent, compared with 2.4 percent last year.

The casual observer might think that a company like Instinet, which derives the majority of its income from trading, would suffer during the current bear market. The casual observer would be wrong. Despite the turmoil in the major market indices, one fact hasn't changed: trading volumes continue to rise -- and that's good for Instinet's business. "One of the things that really lends some comfort is that they've been able to weather the market's upheavals," says Randall Roth, a research analyst at Renaissance Capital, advisor to the IPO Plus Aftermarket Fund (symbol: IPOSX).

As a result, Instinet's first-quarter net earnings came in at $50 million, or about 21 cents per share, on revenue of $431 million. Based on management's 2001 revenue guidance of 20 percent growth from the $1.44 billion it generated last year and net margins of 11.6 percent in the first quarter of 2001, Instinet would net $200 million, or 85 cents a share, in 2001. At the $12.50-per-share asking price for the Instinet IPO, its shares would price at 14.7 times our 2001 earnings estimate.

Although there's a limited universe of companies to draw comparisons against, the Instinet offering is priced to sell. Shares of Investment Technology Group (NYSE: ITG), which specializes in large block trades, trade at 21 times expected 2001 earnings, whereas shares of Knight Trimark Group (Nasdaq: NITE), a Nasdaq market maker, and LaBranche & Co. (NYSE: LAB), an NYSE specialist, trade at 22 and 23 times forecasted 2001 net earnings, according to First Call.

As a result, lead managers Credit Suisse First Boston and Salomon Smith Barney should have little trouble attracting buyers. Based on the steep discount, we would expect these banks to hike the offering's $12.50 asking price. We would be buyers of Instinet shares at levels as high as $17.50, which would translate to a price-to-earnings ratio of 20.6 on 2001 estimates, still below its comparables.

WHERE DO YOU GROW? In addition to a sluggish IPO market -- Instinet would be only the second tech IPO brought by CSFB this year, down from 23 placements at this point in time last year -- one of the reasons for the steep discount is that competition is squeezing the company's profit margins. Whereas the 30-year-old company once enjoyed free reign over institutions, REDIbook, Bloomberg Tradebook, and other well-financed ECNs have forced Instinet to reduce the per-share transaction fees it charges to execute trades over its system. Last year, these transaction fees fell 7 percent to $0.0164 per share, estimates J. Cathleen Boucher, an analyst at Putnam Lovell Securities. Although that might not seem like a lot, it is when you consider that Instinet executed trades of 66.7 billion U.S. shares in 2000 -- this amounts to about $1.1 billion in U.S. trading revenue.

This discounting is clearly a response to the industry's attempt to take a slice of Instinet's 45 percent market share. By way of comparison, REDIbook increased its share of the ECN market to 12 percent in 2000, up from 3 percent in 1999, whereas Instinet's share of ECN trading volume slipped from 65 percent in 1999. REDIbook is backed by a consortium that includes Goldman Sachs, CSFB, Fidelity Investments, and Charles Schwab. And with the expected launch of Nasdaq's own ECN, SuperMontage, in 2002, the long-term prognosis is not going to get any better.

At the same time, Instinet is losing its grip on big block trades to the likes of Investment Technology Group and startups like Liquidnet. As a result, the average fee per transaction has fallen 19.5 percent to $15.81 in 2000. Lower per-share fees, smaller transaction sizes, and the decimalization of stock quotes, are all expected to continue to drive up operating costs.

Yet even if Instinet's market share of equity trading continues to decline -- something that appears to be inevitable -- the company will still benefit from a continued increase in trading volumes. (In the first quarter of 2001, the average number of shares traded increased 2.9 percent sequentially to 2.1 billion shares per day, despite a 25.5 percent decline of the Nasdaq Composite in the first quarter.)

And Instinet has aggressively invested in new products to stay ahead of the competition. Last year, for example, Instinet spent $62.2 million on its fixed-income trading operations, which the company launched in March 2000. This unit is expected to be a big contributor to 2001 revenue expectations, according to attendees of the company's road show.

Even in fixed income, however, Instinet will butt heads with companies like eSpeed (Nasdaq: ESPD), as well as newer entrants such as BondBook and BrokerTec Global.

But these competitive pressures will likely lead more private ECNs to go public in the near future. In the wake of the Instinet offering, we would expect to see IPO filings from Island, Tradescape, REDIBook, and finally Nasdaq itself. All we can say is: it's about time.

To get this column sent to your inbox, subscribe to the e-newsletter.

Part six of seven.

Former Supreme Court Chief Justice William O. Douglas served as Securities and Exchange Commission chairman from 1937 to 1939 and was fond of comparing the SEC to a shotgun kept behind the door: "Loaded, well-oiled, cleaned, ready for use, but with the hope that it would never have to be used."

Now, as reports of potential abuses in the allocation of IPOs circulate, the investment community is asking why nobody bothered to reach for the shotgun during the height of the IPO boom -- when insiders were aware of the tie-ins, excessive commissions, "flipping," "laddering," "spinning," and other questionable activities -- and why the trigger is being pulled now that the market has cooled.

The SEC's basic mandate is to promote the fair movement of markets and the full disclosure of information, and to ensure general fairness between buyers and sellers. Susan Woodward, the former chief economist at the SEC, says that the commission doesn't step in until "there's a great public outcry," as there has been by retail investors who felt cheated by artificially inflated IPO prices.

SEC investigators "tend to follow a pretty similar pattern," explains Patrick McGurn, director of corporate programs at Institutional Shareholder Services, a shareholder advocacy group. "They start to think that there's some systematic pattern of abuse. They call around before they go public. And they allow specifics of the case to hit the press to shake out more information."

Take the SEC's bulletin from last August, in which it warned banks against tie-ins -- the questionable practice of allocating IPO shares to institutional investors who promise to purchase additional shares in the aftermarket at predetermined prices. By that point in time, the practice of tie-ins had become so widespread, sources say, that the bulletin was merely a prelude of the bigger investigation to come. Specifically, the SEC stepped in to remind underwriters that they are "prohibited from soliciting or requiring their customers to make aftermarket purchases until the distribution is completed." It's common practice for investment banks to ask about levels of interest to purchase additional IPO shares in the aftermarket, but they are prohibited from soliciting purchases of unregistered securities -- or in this instance, aftermarket shares before the IPO has occurred.

"Normally, you wouldn't put out something like this unless the SEC had found abuses that warranted further action," says Mercer Bullard, former assistant chief counsel at the SEC and current CEO of Fund Democracy, a Chevy Chase, Maryland-based mutual-fund shareholder advocacy group.

Beginning in December of last year, the SEC and the Justice Department have been following up the SEC's August bulletin and examining specific trading practices, namely share allocations and commissions. The agencies have asked for the trading records of institutions that received IPO allocations and subsequently traded 10,000 or more shares, at commissions of at least 10 cents per share, from several investment banks. Firms whose records have been requested include Credit Suisse First Boston, Goldman Sachs, Bear Stearns, Lehman Brothers, and Morgan Stanley Dean Witter. Salomon Smith Barney and Merrill Lynch also received more general requests for their trading records, according to sources. So far no actions are known to have been taken as a result of the broad inquiry.

INVISIBLE HANDSThe SEC's preference for allowing stock markets to operate independently gives investment banks great leeway. Take commissions that can be charged for executing stock trades. The SEC provides no direct supervision on them, instead leaving guidance to the exchanges themselves. The National Association of Securities Dealers (NASD), for example, provides only cursory instruction. NASD's Conduct Rule IM-2240 suggests that underwriters can charge commissions of up to 5 percent of share's price. But this is only a guideline, not a hard rule.

When dealing in large-cap, liquid stocks, like IBM or Texas Instruments, the 5-percent guideline made it easy for institutions that received IPO allocations to offer inflated commissions as a token of appreciation for IPO shares. In some cases, investors paid as much as 40 or 50 cents per share -- about ten times the usual nickel per share. As one manager of a small hedge fund boasts, "I absolutely paid commissions in the 50-cent-per-share range, and there's nothing illegal about it."

The SEC's approach during the frenzied IPO market was not terribly effective in other ways -- to the indirect benefit of investment banks and high-tech insiders. In 1998, the commission began an investigation into spinning -- the allocation of IPO shares, under the "friends and family" category, to high-tech executives as an incentive to bring their financing business to the underwriting bank. The investigation ended almost as soon as it started, however. Wall Street firms successfully lobbied the federal regulators that they could police their own ranks and put an end to spinning by basing allocation solely on the level of commission business. "Remember that the SEC is captured by its regulatees, " says Ms. Woodward, explaining that only when the public raises hell will "something that works for [the banks] be changed."

Further evidence of the SEC's reluctance to litigate can be found in the way it dealt with online grocer Webvan (Nasdaq: WBVN). During its IPO roadshow, the startup allegedly violated the SEC's pre-IPO "quiet period" rules by disclosing financial projections to prospective investors. But rather than bring an administrative action against company officials or its underwriters, the SEC and Webvan agreed to a cooling-off period before its November 1999 IPO.

Except for high-profile cases like the fingering of Ivan Boesky for insider trading and Michael Milken for junk-bond illegalities, the SEC tends to administrate rather than litigate. Consider the agency's resources: it has only 846 enforcement personnel and an annual budget of about $2.5 billion with which to police not just high-tech IPOs, but all aspects of the public markets. In 1999 and 2000, 705 companies from the high-tech sector alone went public. Strapped as it is, the SEC still filed 503 cases in 2000.

TO THE COURTSDepending on the outcome of the SEC's ongoing investigation, the agency's enforcement division could try its case in either of two venues: in an administrative or in a federal court. Enforcement officials, according to sources familiar with the matter, are inclined to try their case in a U.S. District Court because of their ability to seek more restrictive sanctions there.

A federal court judge can grant an injunction prohibiting particular underwriting practice like tie-ins. Given that tie-in arrangements are already outlawed by Regulation M of the Securities Exchange Act of 1934, any investment bank suspected of them is a likely target for the SEC. Violating a federal court order would put an individual or bank in contempt of court, meaning the SEC could levy additional fines and/or imprisonment for any future violations of the injunction.

An administrative court judge, meanwhile, can only issue a cease-and-desist order, which lacks punitive powers. SEC officials would need to petition a federal judge to win an injunction against a particular underwriting practice.

The SEC's current investigation will likely examine mandated ethical standards and disclosure requirements. Under the 1934 Exchange Act, for example, enforcement officials might seek to prove that underwriters exhibited reckless disregard for the public and that investors were defrauded when underwriters failed to disclose side payments from investors in exchange for allocations. Proving this fraud requires both proof of intent and evidence of damages caused by the side payments.

The intent to defraud is a difficult threshold to prove, but failing to disclose the existence of tie-in arrangements could still violate Section 11 of the Securities Act of 1933, according to Mr. Bullard. This course of action does not require proof of intent, only evidence that by omitting information investors were misled.

Despite the SEC's longstanding preference for self-regulation, investment banks may still have a reason not to volunteer any more information than is necessary. Any concession by an underwriter in either a criminal or a civil investigation has the potential to open the floodgates for class-action lawsuits, says Alice McInerney, a partner at New York-based Kirby McInerney & Squire.

She argues that an admission of wrongdoing by the industry would help her re-open an antitrust claim that she brought in 1999, currently under appeal. In the class-action lawsuit, brought on behalf of investment firms that participated in IPOs against virtually every underwriter that does business in New York, her firm alleged that underwriters permitted certain institutional investors to flip their IPO shares, but the underwriters conspired to restrict retail investors from flipping their IPO shares either by revoking access to future IPO allocations, or by imposing penalty bids if sales occurred within a restricted period of 30 to 90 days after an offering.

As more secrets of the IPO process are revealed, the number of class-action lawsuits will surely rise -- even without any admission of wrongdoing by a bank to the SEC.

In part seven of our series, we'll examine the present climate of the IPO market and assess ways that the going-public process might be improved.This series is based on reporting by Eric Moskowitz, Michael V. Copeland, Stephen Lacey, Stephen Lucey, Christopher Locke, Tom Davey, and Dan Briody. Email questions or tips to eric.moskowitz@redherring.com.

Yet after a March move to switch lead underwriters to Morgan Stanley from Goldman Sachs, Tellium CEO Harry J. Carr is on a cross-country roadshow to drum up support for the company's $105 million IPO. Despite Mr. Carr's pedigree -- the former AT&T executive was at the helm at Yurie Systems when Lucent Technologies (NYSE: LU) acquired the switch manufacturer for $1 billion in 1998 -- his latest venture could be his greatest challenge.

After luring Mr. Carr away from Lucent in December 1999 with options to buy 13.2 million shares -- 7 percent of the company -- for $1.07 per share, the Oceanport, New Jersey-based company had little trouble raising $222 million in September 2000. There were 57 institutions that bought in at levels comparable to the company's original $2.9 billion IPO valuation.

But September was a long time ago. After a reverse two-for-one stock split in April and after burning through $60 million since the last infusion, Tellium is now marketing itself to investors at a $1.5 billion market capitalization. Joint book managers Morgan Stanley and Thomas Weisel Partners are looking to price 7.5 million shares at $13 to $15 per share the week of May 14.

The decision to go forward with the IPO screams of desperation similar to that of Agere Systems (NYSE: AGR.A), another Morgan client that was forced to slash its IPO price to $6 per share, a 60 percent discount to what Lucent originally had sought for the unit. The buy-side can smell blood and will likely take advantage of Tellium at its pricing date.

So why would Tellium management elect to go forward with the offering when it still has $166 million in the bank? One possible explanation could come from the nature of the September private round. As part of the original financing agreement, these Series E preferred stockholders will be made whole, regardless of where the company prices its IPO. At $14 per share, the midpoint of the offering price, their original investment is convertible into the same 15.6 million shares despite the reverse two-for-one split, essentially doubling their ownership stake to 14.3 percent.

THE LONG HAUL The real answer, however, may lie in Tellium's focus on optical switch products for long-haul networks. The second generation of its Aurora optical switch, which began shipping in the third quarter of 2000, uses software that enables carriers to upgrade their networks more rapidly between the major hubs that connect cities. By intentionally ignoring the switching of optical signals in the DS-1 (1.5 Mbps) and DS-3 (45 Mbps) levels used within metropolitan networks, Tellium has become a leader in the OC-48 (2.5 Gbps) switch market.

"As you look into the future, these types of switches have a place in the market because of the huge amount of data that's being carried," says Sam Greenholtz, a senior optical networking analyst with Communication Industry Researchers, a Charlottesville, Virginia, optical industry consulting firm. But because its switches can't groom signals down to lower speeds, where many carriers are concentrating their expenditures, the long-term strategy has cost it over the short term.

The problem is that Tellium hasn't been able to attract many buyers for its long-haul switches. Dynegy Connect, a Dynegy (NYSE: DYN) subsidiary, accounted for 70 percent of the $15.6 million in sales the company generated in the first quarter, while Qwest Communications (NYSE: Q) accounted for the remaining 30 percent. In September 1999, Dynegy Connect agreed to purchase $250 million in Tellium gear over the next five years, while Qwest signed a three-year contract in September 2000 to purchase $300 million in equipment. Cable & Wireless (NYSE: CWP), Tellium's only other customer, also inked a $350 million, five-year purchase agreement in September, but those sales have yet to start coming in the door.

While the $900 million of the three contracts is surely a lot of money over the next five years, current income hasn't stemmed the company's short-term bleeding. In the first quarter ended March 31, 2001, Tellium had an operating loss of $52.5 million after a pro-forma 2000 operating loss of $151.3 million.

Tellium's inability to attract other big-name customers -- there's no mention of any company other than Cable & Wireless conducting tests with Tellium switches -- gives other companies such as Sycamore Networks (Nasdaq: SCMR), Nortel Networks (NYSE: NT), and Tellabs (Nasdaq: TLAB) time to refine their optical offerings. Privately held Calient Networks, which also is focusing exclusively on the long-haul optical switch market, may already have usurped some of the momentum that Tellium enjoyed in its September round of financing. In January, the San Jose, California-based company raised $225 million.

At first glance, the IPO is as expensive as they come. With just $27.8 million in trailing 12-month sales, Tellium would originally price at 53.9 times those sales. That compares to just 7.8 and 16.4 times sales for Sycamore and Ciena (Nasdaq: CIEN), two other optical switch manufacturers.

But Wall Street always prefers the promise of the future to the reality of the past. On that basis, Tellium would price at roughly 6.8 times the sales that its existing contracts imply for this year, compared to 7.8 and 9.1 for Sycamore and Ciena, respectively. And it could get better: the company may be forced to reduce the asking price further to compensate new investors for their patience.

That said, despite the fact that Tellium may have the products that will be used in next-generation networks, Mr. Carr & Co. will likely still have a difficult time convincing public investors that the company will be a long-term survivor. Its trip into the public markets will only be the beginning, akin to the return journey of a salmon to its breeding grounds. On average, out of every 1,000 eggs laid, one survives to return and spawn. Tellium's chances are much higher than that, but we'll wait until they're a little farther upstream.

To get this column sent to your inbox, subscribe to the e-newsletter.

As service providers scale back network construction, network vendors industry-wide are beginning to feel the pinch. In announcing that it would reduce capital spending by $2 billion over the next five years, XO joined Verizon Communications (NYSE: VZ), Qwest Communications (NYSE: Q), SBC Communications (NYSE: SBC), and WorldCom (Nasdaq: WCOM) as companies that have slashed their 2001 capital expenditure budgets.

Add the latest spending slowdown to the graveyard of CLECs that have filed for bankruptcy -- Winstar, Advanced Radio Telecom, and Convergent Communications (OTC: CONV) -- and the combined impact on the income statements of equipment vendors has been hard to digest for an already queasy investment community.

Investors have hammered smaller equipment vendors such as Ciena (Nasdaq: CIEN) and Ceragon Networks (Nasdaq: CRNT). Ciena, which sells dense wavelength division multiplexing (DWDM) equipment for XO's fiber-optic networks, has seen its stock tumble 20 percent so far this week, closing at $53.47. Meanwhile, shares of Ceragon, which derived more than 10 percent of its first-quarter revenue from sales of wireless networking equipment to XO, have fallen 25 percent to $2.40. Although Ciena and Ceragon are most directly affected by XO's lower cap-ex guidance, Nortel Networks (NYSE: NT), which sells voice switches and fixed wireless networking to XO; Sonus Networks (Nasdaq: SONS), which provides soft switches; and Triton Network Systems (Nasdaq: TNSI), which sells point-to-point wireless networking, also will be affected.

However, Level 3 Communications (Nasdaq: LVLT) was perhaps most directly impacted by XO's decision to unwind its European expansion and concentrate on penetrating further into the 62 U.S. markets it currently serves. Although XO will lease additional fiber-optic capacity in the U.S. as part of spending reductions, its decision to forgo the purchase of fiber networks in Europe will reduce revenue at Level 3 by anywhere from $150 million to $300 million, speculates Kent Siefers, a research associate at Thomas Weisel Partners. Level 3, which last week trimmed its 2001 revenue guidance to between $1.4 billion and $1.5 billion, has seen its stock fall 14 percent in the last week to $13.04.

DEBT INGESTION In many ways, XO Communications is a textbook case study in what's wrong with the CLEC business model. The company, which has funded its growth through a heavy dose of long-term debt, is finally feeling the weight of that burden in light of a capital-markets shutdown.

"They basically had to do something. XO had painted themselves into a corner," notes Joe Altobello, an associate analyst who follows CLECs for CIBC World Markets. According to Mr. Altobello, XO would have run through its available cash by year-end. Even after the $250 million equity infusion by Forstmann Little, the company still faces a $1 billion shortfall in funding its operations until it reaches free cash flow profitability, which isn't expected to happen until 2006.

Given the inability of CLECs to raise any type of financing in today's marketplace, the company's stock is probably the last thing an investor would want to buy, even if you believe it will survive for the long term and one day achieve profitability. Instead, if investors are confident in XO's ability to get additional financing, we suggest that they take a look at its 5.75 percent convertible bonds that are due in 2009, or its 9 percent senior notes that are due in 2008.

Even if XO management hits its 2001 revenue goal of $1.3 billion, the company will still have a debt-to-revenue ratio of 4.8, given a debt load of $6.3 billion. Not a lot of room to negotiate until it achieves profitability. The net result: the higher you sit in the capital structure the better protection you have if the company fails to achieve profitability.

THE BOND ADVANTAGE While equity buyers were reluctant to buy into the XO story -- the stock fell 4 percent to $4.32 on Thursday -- bond investors were clearly more comfortable with the lower cap-ex spending and equity infusion.

Quick bond lesson. A company typically must pay $1,000 to bondholders once the bond matures. Bonds are quoted on the market in terms of par, or 100. So if a bond is trading at 55, that means a bond that pays $1,000 at maturity would cost $550 to purchase. XO's convertible bonds, which were issued in January and initially traded as high as 120 percent of par and as low as 36 in March, traded today at 40, up from 38 on Wednesday. The more senior 9 percent notes, which traded as high as 80 in January and as low as 42 last month, closed on Thursday at 55, up from 45 on Wednesday.

Bottom line: it seems that there's a better chance of XO going bankrupt than actually making it in the long run. And if the company were to go bankrupt, bondholders at least would receive proceeds from the sale of assets -- while owners of stock are likely to receive nothing. Little wonder that the value of the bonds increased on Thursday.

Bowne & Company (NYSE: BNE), in particular, might have a tough time of it. The 225-year-old printing firm derived $187 million -- 72 percent of its $261 million fourth-quarter sales -- from financial printing related to IPOs, secondaries, M&A tender offers, and regular company reports to shareholders. A year earlier, financial printing generated $218 million -- or 78 percent of Bowne's $279 million in total revenue.

That 6 percent revenue decline, plus the fact that the company reported a net loss from operations of $225,000, or 1 cent per share in the fourth quarter, helped push Bowne's stock down to $10.47 on February 26 -- 27 percent below its 52-week high. During a heated fourth-quarter conference call with investors, Bowne management said the company would shut down antiquated presses, consolidate select operations, and lay off about 3 percent of its workforce in an effort to cut costs by $20 million.

For Bowne, which fed off the $224 billion raised by technology companies through IPOs and secondaries during the past three years, the market downturn has provided quite a crash diet. Whereas until recently the company struggled just to keep up -- putting its 30-year-old presses into service -- now the struggle is of a much different sort.

Write to stephen.lacey@redherring.com.

Fortunately, as was the case with the Securities and Exchange Commission's adoption of Rule 155, which went into effect on March 7 and made it easier for issuers to transition from a withdrawn public offering to raise money in the private markets, the SEC has broadened the relief that it offers from the filing requirements triggered by Section 12(g) of the Exchange Act. "It's a sensible solution," says David Redlick, an attorney at Boston-based Hale and Dorr.

The SEC's new and more flexible guidelines enable companies to seek relief for stock options that are immediately exercisable, as long as they are nontransferable in most cases. Also, employees can remain vested beyond their termination date. In certain circumstances, the relief provisions also cover consultants that participate in an option plan.

"There will be companies out there that will want to amend their option plans to give them more flexibility," says Michael R. Littenberg, a securities lawyer and head of the Internet & New Media Group at New York-based Schulte Roth & Zabel. "In this environment, companies let employees go, but you don't want people to leave on bad terms, and this gives them the ability to potentially preserve some upside."

PUBLIC DISCLOSURE, BUT NO PUBLIC OFFERING For companies that hit the option-holder threshold before the end of last year, the clock is ticking on when they will be able to seek relief from Section 12(g). "There are a lot of companies that could inadvertently blow through the Exchange Act requirements simply because they're not aware of the rules or haven't focused on how many option holders they have," says Mr. Littenberg.

Specifically, companies with more than 500 option holders and assets in excess of $10 million at fiscal year-end have until April 30 to petition the commission for relief from the regulations that would require them to begin filing documents with the SEC as if they were a public company. (Companies that petition for relief but are denied that status have until July 30 to resubmit an application.)

It goes without saying that it would not be ideal for private companies suddenly to have to start disclosing their financial statements, given the difficult state of the economy and an inability to raise money in the private markets. Essentially, these companies would experience all of the pain of being public in today's market, including the added costs of lawyers and accountants for public disclosure, without the benefits of actually having raised money. In fact, being a private company with books that are open for anyone to see could be extremely harmful. "Because of the rocky shape of the capital markets, competitors can use a company's financial disclosure against them in negotiating for business," notes Mr. Redlick.

Although it's difficult to determine how many pre-IPO companies would be affected by Section 12(g), 16 of the 107 prospective technology issuers in registration to go public have more than 500 employees, according to Thomson Financial Securities Data. Those most likely affected by the regulation are companies in labor-intensive industries, like tech consultancy Global Knowledge (proposed Nasdaq: GOGK), engineering consultancy Tality (proposed Nasdaq: TLTY), and laboratory testing provider American Medical Laboratories (proposed Nasdaq: AMLS), all of which are in registration.

In the IPO bull market of 1999 and 2000, investment bankers, lawyers, and accountants were quick to criticize the slow reaction time of the SEC. In a much tougher 2001, it's time to applaud the commission for recent actions that ease the pains of tech companies in today's market, as well as those of their current and former employees.

In what some analysts consider nothing short of an information technology revolution, data center operators, telecommunications companies, systems integrators, and private startups are rushing to market with software packages that automate administration of the software and servers that run the networks of both small and large companies.

"It's the biggest shift that we've seen since the consolidation of telecommunications," says David Tapper, who follows the Internet infrastructure management industry for IDC, an industry consultancy. "We're seeing the marriage of information technology with telecommunications."

It's a powerful message that hasn't been endorsed by the public markets. Early providers of automated management tools like Loudcloud, which makes Opsware, and Genuity (Nasdaq: GENU), which sells Black Rocket, have seen their stock prices flounder as investors wonder whether they'll have sufficient funding to be part of the revolution.

Despite gushing approval from Loudcloud's IPO underwriters on Tuesday -- lead manager Goldman Sachs put the stock on its "Recommended List" -- the company's shares have been unable to gain any ground in the public markets.

Loudcloud's share price has fallen 20.7 percent to $4.31 since Goldman Sachs, Thomas Weisel Partners, and Epoch Partners initiated coverage of the stock. Nevertheless, venture capitalists remain bullish on the sector. Telenisus landed a $45 million Series C round earlier this week; SevenSpace took in $45 million in its second round in February; and Inteq and SiteLite also are busy closing their second rounds of financing.

"I don't think Global 2000 customers are going to buy into a completely automated solution," says IDC's Mr. Tapper. "Eventually, we're going to get to a standardized service, but Loudcloud may be ahead of the curve."

Loudcloud's ability to support its leased data centers, servers, and 586 employees will likely be the least of its challenges. With $243 million in cash as of March 8, the company probably wouldn't burn through its money until the end of fiscal 2003. The problem is, the lucrative nature of outsourcing has attracted the attention of much larger technology companies. Monthly service contracts that can exceed $200,000 per month for enterprise customers has attracted collaborative development efforts from Microsoft (Nasdaq: MSFT), Compaq Computer (NYSE: CPQ), and Intel (Nasdaq: INTC), as well as stand-alone projects from IBM (NYSE: IBM) and Sun Microsystems (Nasdaq: SUNW).

Recent alliances between these developers and telecommunications carriers to implement managed solutions for existing and new data centers underscores the potential of IT outsourcing. Within the past month alone, IBM and Qwest Communications (NYSE: Q), Sun and Sprint (NYSE: FON), and Microsoft/Compaq/Intel/EDS (NYSE: EDS) have all announced collaborations in this arena. Qwest and IBM, which agreed to build 28 new data centers over the next three years, will split the $5 billion of revenue the companies expect from the multiyear agreement.

That type of financial stability may prove to be the largest obstacle for Loudcloud and a host of privately held companies targeting the infrastructure management space. "You're outsourcing your business infrastructure, so you have to know that they're going to be around," notes Corey Ferengul, an industry consultant at the Stamford, Connecticut-based Meta Group. "Were telling our clients that you must have a guaranteed service contract of at least 12 months and assurances that they have enough cash to last 18 months."

ECONOMIC REALITIES Given the public market's demand for immediate profitability, smaller private companies have been tweaking their business models so they can become profitable sooner. To keep a lid on costs, Rancho Santa Margarita, California-based SiteLite funnels all of its business through an outsourcing agreement with Exodus Communications (Nasdaq: EXDS). Unlike Loudcloud, SiteLite isn't burdened with leasing the server and hosting space, but instead has teams that monitor customer-specific applications.

Although in-depth knowledge of a customer's software applications may not be scalable, the strategy makes operational sense to Exodus as it transitions from a co-location company toward lucrative value-added services. "One of the problems is that some of the managed service providers are trying to offer an extremely automated, one-size-fits-all solution," says Peter Fortenbaugh, senior vice president of strategic planning at Exodus. "The reality is that many of our customers have very large, complex Internet operations that require custom solutions."

Although Exodus has rolled out its own completely automated solution -- developed by Sun, Compaq, and Cisco Systems (Nasdaq: CSCO) -- outsourced relationships like the one with SiteLite reduce the company's reliance on in-house engineering talent and enable it to be quicker to market with a menu of managed services. Other Exodus partners include Peakstone for applications management and Mirror Image for content distribution.

Concentrating on monitoring a customer's existing applications, rather than forcing them to adopt a cookie-cutter solution, is seen as a more economically viable business model. Oakbrook Terrace, Illinois-based Nuclio, which has 70 clients, has followed that prescription to profitability. As a result of its unique position of being profitable, the company is looking to consolidate a crowded field of managed service providers (MSPs).

"We believe there's tremendous opportunity to roll up some of this industry," says Nuclio CEO John Jazwiec. "There's a lot of players going after a lot of business, but the reality is that you have to get to the $25 to $30 million per year run rate to cover the fixed costs of providing these services. We see a lot of people in our space that are running out of cash."

Having a cash-flow-positive parent in privately held Forsythe Technology to bankroll those acquisitions certainly doesn't hurt. And there are plenty of targets. At least 100 smaller MSP companies may have no alternative but to seek a buyer.

And unless Loudcloud can convince Wall Street that its execution is as good as its idea, it too may wind up as prey for a larger technology firm. Chairman Marc Andreessen has played this game before. Netscape, his last great idea, did wind up selling out to America Online, after all.

The pact with the world's largest server maker comes on the heels of an oversubscribed round of venture capital in January. Ensim set out to raise $45 million but ended up with $64 million -- and still had to turn VCs away. The round is all the more impressive given that venture capitalists as a whole are getting tightfisted with their money. Among the top-name VCs that participated in the funding were New Enterprise Associates, Onset Ventures, and CMEA Ventures.

THE IRON GIANT While Ensim is not unique in its ability to set up virtual servers and remotely monitor application usage and billing, aligning itself with the world's largest server vendor distances it from similar solutions provided by Sphera, Computer Associates's (NYSE: CA) iCan SP, Xevo, and PlateSpin.

"With the shakeout in the service provider industry, customers want a high level of service and the ability to provision applications in a rapid manner," says John Madden, an analyst at Summit Strategies, a Boston-based consultancy that tracks ISPs, application service providers (ASPs), and data center operators. "Ensim's software gets them off the profitless pit of prosperity of adding dedicated servers for each new customer."

The Compaq deal could be huge for Ensim. Last year Compaq, which controls 34 percent of the server market, became the first company to ship more than 1 million servers. For each server that comes equipped with Ensim software, Ensim will receive a recurring monthly fee, in addition to one-time, up-front license revenue. Although none of Ensim's 60 customers is in danger of defaulting on that fee, CEO Rosen Sharma says the Compaq distribution channel will allow his company to tap into the top-50 type of global service provider that will ensure its long-term financial viability.

DIVIDED WE STAND Ensim's ServerXchange software increases server utilization by subdividing each physical server into multiple private servers. While customers want the reliability and flexibility of a dedicated server, the Ensim solution allows service providers to operate their server farms more efficiently. At the same time, they are able to give the end customer the perception that an individual server is dedicated to their needs.

While dividing servers into increasingly fine layers comes at a cost to computing power, the solution is particularly attractive for providing services to small and midsize enterprises that may not always utilize a server's complete capacity. Frequently, these customers only need a fraction of the server's potential computing power to run their e-commerce sites, administer email, collaborate internally, or stream video and audio. "[The Ensim solution] aligns costs much better with the financial model of service providers because it provides for a 'pay as you grow' type of scenario," says Mark Linesch, vice president of Compaq's service provider solutions.

Another valuable aspect for ISPs, ASPs, and data center operators is that Ensim's software gives them the ability to create new virtual servers and modify existing ones through a "drag-and-drop" graphical user interface. There is no programming required to monitor, maintain, and adjust to a customer's capacity or application needs. The fact that everything can be managed remotely also makes economic sense to service providers.

"Part of what is critical to reducing the costs of operating data centers is to get people up and online very quickly with a customized suite of applications," Mr. Linesch says.

While we wouldn't expect to see Ensim in the IPO pipeline anytime soon, it has plenty of time to get itself ready for a road show. Even before Monday's deal with Compaq, Mr. Sharma said his company had enough capital to last three years. He expects Ensim to be profitable by the second half of 2002. Depending on which pundit you talk to, the company might be just in time for a new wave of IPOs.

Those regulations dictated that for six months after pulling an IPO, a company could only raise money from qualified institutional buyers (QIBs) -- those with $100 million under management. The problem was that the private investors who knew the company's story best might not qualify as QIBs.

To remedy the problem, the SEC enacted Rule 155, which became effective March 7. It reduces the safe harbor period from six months to 30 days and expands the pool of investors a company can tap in that period to include investors classified as accredited -- institutions with $5 million in total assets and individuals with a net worth of $1 million and income in excess of $200,000 during the last two years. "There's a severe liquidity crisis going on, and Rule 155 is a response to that crisis," says Curtis Mo, an attorney with Brobeck, Phleger & Harrison.

In a bull market, few private-company executives would care about Rule 155. But now that the IPO window is nailed shut, the new rule is an important way to safeguard the financial health of young companies forced to delay their public debut.

That said, the 182 technology companies that bowed out of the IPO process in 2000 might reasonably have hoped for swifter action from the SEC on a rule that's taken nearly three years to wend its way to approval.

Write to stephen.lacey@redherring.com.

With just nine technology IPOs in the quarter, down from 99 placements in the same period last year and 33 in the fourth quarter of 2000, investment banks turned to the debt market to offset the shortfall in equity underwriting fees.

"With interest rates coming down and given the negative sentiment in the equity markets, there was strong demand in the quarter for debt offerings because of the attractive yields," notes one fixed-income banker at Credit Suisse First Boston. For Credit Suisse, which was shut out of the tech IPO market in the first quarter after underwriting a Street-high 56 technology offerings last year, the debt market's receptivity has helped soften the blow of an otherwise rough quarter for its investment-banking operations.

INVERSELY PROPORTIONAL While the number of companies going public fell to a ten-year low and the number of follow-on offerings are at a two-and-a-half-year low, the amount of tech-related debt issuance climbed to an all-time high of $55.3 billion by 59 issuers during the quarter, according to Thomson Financial Securities Data. That's up from $10.5 billion in the fourth quarter and $22.8 billion in the year-ago period.

"The initial demand was from the traditional telecommunication issuers, and that had a pull on lower-grade technology paper," notes the Credit Suisse First Boston banker. Companies that issued at least $1 billion's worth of above-investment grade debt included Motorola (NYSE: MOT); Qwest Capital Funding -- a subsidiary of Qwest Communications (NYSE: Q); AT&T Wireless (NYSE: AWE); France Telecom (NYSE: FTE); and SBC Communications (NYSE: SBC).

With the equity markets all but closed, the fact that there were buyers for debt securities suggests that the liquidity crisis is not as severe as many people feared. And as AT&T Wireless and SBC Communications roll out third-generation (3G) wireless services later this year, capital will flow from their balance sheets and onto the income statements of vendors like Nortel Networks (NYSE: NT), Cisco Systems (Nasdaq: CSCO), and Lucent Technologies (NYSE: LU).

Yet as earnings warnings continued to roll in over the course of the quarter, debt buyers grew increasingly wary of riskier high-yield placements. Although the market absorbed offerings from wireless operators such as Triton PCS (Nasdaq: TPCS) and Nextel Communications (Nasdaq: NXTL) -- which took in $350 million and $1.25 billion, respectively, from debt offerings in January -- lower-grade issuers have been forced to turn to the bank debt markets to secure financing.

Cash-strapped companies such as data-center operator Equinix (Nasdaq: EQIX) and application service provider USinternetworking (Nasdaq: USIX), both of whose stocks trade below $2 a share, were able to turn to the banks to secure $150 million and $320 million, respectively, in January. With these companies pledging their assets as collateral, the bank loan market extended into the private markets in early March as privately held data-center operator InFlow secured a $75 million facility.

BANKING FOR ALTERNATIVES For the investment banks, which took in gross proceeds of just $253 million from tech-related IPO issuance in the first quarter, down from $701 million in the same period a year ago, the debt placements have provided a valuable source of underwriting fees. Led by Salomon Smith Barney, which placed $11.8 billion of debt for 21 issuers; J.P. Morgan Chase, which placed $7.3 billion for 17 companies; and Credit Suisse, which placed $6.6 billion for 14 issuers, tech-related debt issuance more than offset the shortfall from equity placements. Salomon, J.P. Morgan, and Credit Suisse all failed to place a single IPO in the first quarter.

As technology stocks continued to fade this year -- the Nasdaq composite shed 25.5 percent of its value during the first quarter and has fallen 59.8 percent in the last year -- so did investors' appetite for IPOs. "The public institutions have seen their portfolios plunge to such a great extent over the past 12 months that there's a lot of proven companies to invest in," notes George Bischof, a general partner at Charter Growth Capital and former investment banker at Robertson Stephens.

With the exception of Morgan Stanley Dean Witter, which was lead manager on the $2 billion offering for KPMG Consulting (Nasdaq: KCIN) and the $3.6 billion IPO of Agere Systems (NYSE: AGR.A), it was a brutal quarter for technology bankers. Other than Morgan Stanley, only Merrill Lynch was able to land multiple offerings during the quarter through deals for Exact Sciences (Nasdaq: EXAS) and SureBeam (Nasdaq: SURE).

While it's clear that the IPO market will remain closed well into the second quarter -- there are only four technology deals scheduled for the month of April -- the debt markets should continue to provide a much-needed source of liquidity, not to mention a way for investment bankers to earn their keep.

Indeed, for much of 2000 it seemed preordained that Loudcloud would have the hottest IPO since Netscape itself, the company that started all the IPO madness in the first place. But Loudcloud's hype machine had its gears stripped after failing to raise a much-needed $110 million in a fall IPO, and detractors are starting to pile on. Mr. Andreessen's vision of providing scalable, completely outsourced systems for Internet infrastructure management is too expensive, they say; too few customers are signing up, the competition is gaining, and cash is dwindling. While it's clearly not a market that is Loudcloud's for the taking, we disagree with those who are suddenly predicting its doom.

Cash is not exactly a small issue, of course, and only Loudcloud's investors know the current story for sure, but based on its most recent Securities and Exchange Commission filing as of press time, Loudcloud had a six-month cash operating loss of $23.8 million that dropped its available cash to $142.2 million as of July. Its burn rate of $10 million, according to that filing, might have lowered Loudcloud's cash on hand to $80 million by the fiscal year ended January 31. (The contracts signed by customers in the fall require no significant up-front payment.) What this means is that if the IPO window doesn't open soon, the company will likely need bridge financing. But in fact, Loudcloud planned a late-February road show that could pave the way for a March offering.

Loudcloud's viability depends on recurring revenue from customers -- monthly payments that range from $50,000 to $500,000. As of October, Loudcloud had secured contracts worth $76 million over a year and a half. Most critically, the company is winning lucrative contracts from the likes of Blockbuster (NYSE: BBI), News Corporationя(NYSE: NWS), Univision Communicationsя(NYSE: UVN), and Brocade Communications Systemsя(Nasdaq: BRCD), which all signed up this year. Add to those customers Nike (NYSE: NKE) and Fannie Maeя(NYSE: FNM) -- the type of not-dot-com clients able to pony up $500,000 per month -- and it looks like Loudcloud is executing a key strategy implemented in the fall: branch out from the dot-com companies that dominated its client roster of 29 as of October. (Names like DrSpock.com, eDeploy.com, and Work.com hardly inspire awe these days.) Obviously, too, those new customers don't believe that Loudcloud's service is as expensive as the critics might have you think. "If you project the costs out two years, Loudcloud was by far the cheapest option," says Sunny McRae, a vice president of engineering at Xign, an electronic payment systems provider that signed a January service contract for $100,000 to $150,000 per month.

Loudcloud is buoyed by the fact that companies are showing the same willingness to outsource their Internet operations this year as they were to allow EMC (NYSE: EMC), Brocade, and Network Applianceя(Nasdaq: NTAP) to handle their storage needs last year, says Alex Arnold, Internet and IT services analyst at Adams, Harkness & Hill. But Loudcloud is by no means out of the woods, and competition is certainly heating up. Totality, which secured $100 million in private financing last July and manages Kmart's BlueLight.com Internet site, heads a list of a dozen private companies that have flooded the nascent sector. Meanwhile, Exodus Communicationsя(Nasdaq: EXDS), Genuity (Nasdaq: GENU), IBM (NYSE: IBM), and EDS (NYSE: EDS) are also targeting managed services through the rollout of complex data centers.

Loudcloud clearly doesn't have the divine right to rule over the managed service provider business as some once thought, but it still has a leadership position and is securing big contracts. Mr. Andreessen may be nervously checking his rearview mirror, but he's not looking at any tailpipes.

Write to stephen.lacey@redherring.com.

Two weeks ago, Agere management said that revenue for fiscal 2001 (which ends in September) would come in at $5 billion, a 5 percent increase over the $4.75 billion it posted in fiscal 2000, according to one buy-side source. Late last week, however, management revised that guidance downward by calling for flat revenue growth this year, according to an amended filing last Thursday -- the sixth of seven amendments. As large purchasers of Agere's optical networking components -- like Lucent, which accounted for 21.5 percent of sales last year, Nortel Networks (NYSE: NT), and Cisco Systems (Nasdaq: CSCO) -- continue to lower their own earnings projections, the prognosis for Agere is likely to only get worse.

"I'm even more bearish given the latest guidance," says Patrick Comack, an analyst at Guzman who initiated coverage of Agere with a Neutral rating and a $4 price target. That target is well below the $6 per-share price that lead manager Morgan Stanley Dean Witter was able to place on Agere's 600 million-share offering.

Agere's problems run deeper than just a question of slowing sales. As inventories back up, there's evidence that the company is being forced to discount equipment. Deteriorating prices were evident as early as the end of last year. Although sales for its December quarter increased 41 percent year-over-year to $1.35 billion, gross margins fell 590 basis points to 42.1 percent. And odds are that these lower-margin sales will ultimately have an effect on Agere's bottom line.

As a result, Mr. Comack is advising clients to avoid the stock ahead of what he believes will be a "train wreck" in the March quarter. He conservatively projects that Agere will lose 6 cents per share in the March quarter and 14 cents per share for the year. In Agere's IPO filing, management said that it expects to post a "significant operating loss" for the year.

Such veiled language is what you'd typically find in an IPO prospectus. And the frightening thing is that the expected losses could be just the tip of the iceberg. Restructuring, which may include layoffs, and other related one-time charges could take operating results even lower. The following passage is also from Agere's most recent prospectus: "In the near future, it is possible we may take additional action to lower capital expenditures and reduce costs and expenses in areas such as general and administrative, research and development, marketing, sales, and manufacturing. These actions are likely to include reductions in our workforce levels."

JUNK STATUSThe stakes are so high for both Lucent and Morgan Stanley that there was talk that the lead manager was an active buyer of the company's stock in the aftermarket. Vinny Slavin, a trader at Cantor Fitzgerald who advises clients on IPOs, speculates that Morgan Stanley exercised its right to sell more than the 600 million shares. Known as a "green shoe" provision, the over-allotment option gives Morgan Stanley the right to short sell an additional 90 million shares, which would have made the total offering 690 million shares, 15 percent above the original offering size.

If the stock goes up, the underwriter is able to fill that short position from shares offered by the issuer at the offering price. If the stock goes down, the underwriter can cover its short position by purchasing shares in the aftermarket in order to stabilize trading. "That's the only reason it's holding a $6 bid," says Mr. Slavin. "You could probably call up Morgan and still get in at six dollars."

Faced with a lack of buyers, Morgan Stanley apparently stepped in to lend support to Agere's stock. Morgan Stanley's syndicate desk declined comment on whether it exercised the over-allotment option. The stock traded in an extremely narrow range, with a high of $6.16 and a low of $6, before closing at $6.02, a mere .3 percent gain. Sellers are likely to test Morgan Stanley's staying power over the next few days, and we would expect to see Agere stock head lower.

The great lengths that Morgan Stanley went to in order to get the offering done points to even more trouble ahead for stockholders of both Lucent and Agere. The original structure of the Agere offering, which would have valued the subsidiary at $23.4 billion, was expected to be the deleveraging event Lucent needed to avoid having its debt ratings lowered to "junk" status. A month ago, Lucent would have shed $5.1 billion in debt. In the completed offering, which values the subsidiary at $9.8 billion, Agere assumes $2.5 billion in debt.

Whether Morgan Stanley did or didn't exercise the over-allotment option on the Agere offering could have additional consequences on Lucent's debt reduction efforts. Morgan Stanley currently holds $840 million's worth of Lucent commercial paper. The debt held by the underwriter represents the remnants of an aborted debt-for-equity plan that would have allowed Lucent to shed an additional $2.5 billion. Last Thursday, Morgan Stanley held $1.6 billion in Lucent debt, down from $2.3 billion on March 21.

This tortuous financial restructuring has us convinced that Lucent and Agere will see their debt ratings downgraded to junk status. And as this story continues to unwind, we'd avoid both stocks and let Morgan Stanley continue to pull the strings.