Recent initial pub–lic offerings by Canadian high-tech firms have exploded out of the gate, vastly outperforming the tech-heavy Nasdaq composite index along the way. As a result, the damage done by the disappointing tech IPOs of 2004 and 2005 is finally receding from investors’ memory.

The leader in the field is Waterloo, Ont.-based Sandvine Corp., which went public on the Toronto Stock Exchange last October at $1.90 a share. This developer of tools for broadband Internet service providers has seen its share price more than triple, to $6.55 at the close of trading on July 9. This represents annualized growth of almost 500%, vs 18.8% on the same basis during the same period for the Nasdaq composite index.

As well, two Ottawa-based firms — Dragonwave Inc. and Espial Group Inc. — have seen their share prices jump at an annualized rate of more than 100% since their IPOs earlier this spring.

All three firms are riding a wave of enthusiasm for companies that supply bits and pieces of next-generation networks. The latter two companies are capable of dealing with the flood of new traffic created by bandwidth-hogging applications such as video-sharing Web site YouTube.com and social networking site Facebook.com. Dragonwave develops microwave gear for high-speed wireless networks, while Espial makes software products for delivering Internet-protocol television.

The rapid rise in share prices for these companies is great news for anyone clever enough to have bought shares shortly after trading began. Investors now thinking about jumping in, however, face an entirely different situation. Not only has the easy money already been made at Sandvine, Dragonwave and Espial, but the odds of a sudden, sharp reversal must be rated as good.

To understand why, consider the track record of an earlier class of seven high-tech firms that went public on the TSX during a 15-month stretch that ended in April 2005. This group’s experience not only holds important clues about how to play the latest IPOs, but at least two of the firms are buying opportunities in their own right.

To date, the IPO class of 2004-2005 has certainly not impressed. Only one member — Toronto-based Q9 Networks Inc. — has seen its share price outpace the Nasdaq composite index since the start of trading. In sharp contrast to Sandvine et al, the majority in the earlier group repaid early investors with substantial declines from their IPO price.

A key factor in their rapid deterioration in share value has to do, in most cases, with those companies’ lack of diversification. The majority of their revenue tends to be generated by less than a handful of customers. This means that a change in buying patterns by just one of these clients can have a dramatic impact on earnings, as discovered recently by Ottawa-based March Networks Corp., a developer of digital video recorders for securing corporate premises.

March Networks announced on Jan. 15 that its top customer — Wal-Mart Stores Inc. — had sharply and unexpectedly dropped its purchases of March Networks technology in favour of another supplier. Because Wal-Mart accounted for roughly one-third of March Networks’ total sales, shares of the firm tumbled 44% the day of the announcement to $11.84 a share from $21.13.

The plunging price was all the more shocking because, just a few weeks earlier, March Networks had trumpeted its 13th consecutive quarter of revenue growth. Nevertheless, the aftermath of the announcement presents investors with an opportunity to acquire its shares at a relative bargain. Four of six analysts who have issued reports on March Networks since early July rate the company as a “buy.”

Greg Reid, an analyst with Wellington West Capital Markets Inc. in Toronto, increased his 12-month price target for March Networks on July 9 to $17.50 a share from $11 after the company announced a $20-million contract from an unnamed U.S. retailer. This was on top of $6 million worth of new business from Wal-Mart, which continues to buy some gear from March Networks.

In his July 9 research note, Reid predicted March Networks will achieve revenue of $84 million in the fiscal year ending April 30, 2008, down from $87.6 million in fiscal 2007. He is also calling for a net loss of 2¢ a share in fiscal 2008, down substantially from a net profit of 34¢ a share in fiscal 2007.

@page_break@These numbers admittedly look weak, but analysts, including Reid, view fiscal 2008 as a transition year in which March Networks is likely to succeed in its attempt to diversify its revenue base away from Wal-Mart. The company’s non-Wal-Mart business is increasing much faster than the industry’s 20% growth rate. Reid anticipates March Networks’ sales in fiscal 2009 will jump by more than 30% year-over-year to $109.2 million, resulting in earnings of 58¢ a share.

Sandvine and Dragonwave are even more exposed to the whims of individual customers than March Networks was. Sandvine pointed out in a regulatory filing that just two cable operators accounted for 80% of its sales in its first half ended May 30, 2006. This was an increase from a 61% dependency during the same period a year earlier.

On the other hand, Dragonwave’s single biggest customer accounted for 36% of the firm’s sales during the nine months ended Nov. 30, vs 48% in the same period a year earlier.

Espial has the most diverse base of customer revenue of the three most recent IPOs, with its largest account making up only 20% of total sales in recent quarters. However, Espial illustrates another serious impediment to growth in market value: it’s not big enough yet to attract a following of independent analysts.

Dave Furneaux, a founding partner of Boston-based Kodiak Venture Partners, doesn’t like to take companies public unless he can justify a market capitalization of at least US$400 million. “Otherwise,” he says, “they just don’t have the visibility.”

Of the three most recent IPOs, only Sandvine, with a recent market value of $812.9 million, has attracted a solid following among analysts — six, according to Bloomberg LP. Espial and Dragonwave — with market caps of $78 million and $118 million, respectively, as of July 9 — had just one analyst report between them. And that was by the lead underwriter for Dragonwave’s IPO.

In contrast, Q9 Networks and March Networks each have at least 10 analysts to heighten retail interest in the firms. However, none of the other members of the class of 2004-05 had more than five.

The advantage brought by analyst coverage becomes readily apparent when the companies put on a good performance. This seems the case now for Sandvine and Q9 Networks, both of which lease Internet infrastructure and services for a fee. Nine of 11 analysts covering Q9 Networks rate the company as a “buy” following the publication of its second-quarter results in June. The investing thesis in a nutshell: Q9 Networks has invested heavily in recent quarters in new facilities and equipment, and is adding customers at little extra cost.

Scott Penner, an analyst with Toronto-based TD Newcrest, has a 12-month share price target of $17, based on his estimate that revenue will jump 23% to $70.2 million in the fiscal year ending Oct. 31, 2008, vs $56.7 million in fiscal 2007. He is projecting fiscal 2008 earnings of 25¢ a share vs just 3¢ a share in fiscal 2007.

Sandvine, given the recent rapid run-up in its share price, probably carries more risk than Q9 Networks. But Sandvine is still being recommended by three of five analysts who have issued reports since early July. A major factor is the company’s superlative growth prospects: the consensus estimate calls for sales of $119.6 million in the fiscal year ending Nov. 30, 2008, up 65% from $72.1 million in fiscal 2007. Earnings are expected to reach 21¢ a share in fiscal 2008 vs 16¢ a share in fiscal 2007.

Of course, this assumes Sand-vine’s major customers stay the course. IE