February 19, 2008

Software industry M&A and IPO roundup 2007

by Gordon Graham, Editor, SoftwareCEO

2007 was a turbulent year for the economy, with rising oil prices, the subprime credit meltdown, and the slipping American dollar.

But if you attended our Software University class recently delivered by Corum Group, you know that the deal-making continued.

The total world value of M&As in 2007 was $4.7 trillion, with both Europe and Asia setting records for deal volume.

Going public, while an exciting thought, represents just the crumbs from this table.

All told, there were $268 billion worth of technology M&As in 2007, versus $15 billion worth of technology IPOs. That means roughly 18 times as much money flowed through M&As as through IPOs.

"Once again, we saw that the real path to liquidity is through an M&A, not an IPO," says Ward Carter, Corum's president. Many more firms go the M&A route because there's so much more money flowing there.

In the software industry, Corum saw M&A valuations of 2.5X to 3X revenues, while SoftwareCEO's numbers show IPOs ranging from 6X to 10X revenues.

The rule of thumb holds: You can make two to three times as much going public as through an M&A, IF all the conditions are perfect.

How to build value, and then exit...
For anyone wondering about an exit in 2008, Corum has a few simple tips.

Sell when you have momentum, and when your sector is trending up.

And don't be the last man standing, i.e., the last company to be acquired in your space. After all the big players have made their play, your firm could be the odd man out, with no partner in sight.

"The most attractive software companies have strong recurring revenues, either through SaaS or a large stream of maintenance contracts," says Carter.

"Even though people tend to focus on the technology, it's not just technology that drives a deal. There's a lot of other things that buyers look at very carefully."

Among these, he names:

• Business model
• Installed base
• Team/domain expertise
• New markets to expand into
• Distribution

You also have to consider culture, he says.

"It's certainly easier to be acquired by a customer or a partner you've done business with before. It's sometimes hard to get a deal done when you're eight or nine or 14 time zones away."

Speaking of the clock, timing clearly plays a big role in the success of any M&A or IPO.

"You get a real sense of the maturity of a company when you start looking at the total revenue, revenue per employee, and profitability," says Dave Sommer, publisher of SoftwareCEO.

"And when a company is ready to go IPO, hopefully the market is receptive to it. When money is a bit tighter, you see more M&As as opposed to IPOs."

Money may be a little tighter in 2008, with fears of a recession and the January stock market downturn. But with lots of buyers from around the world, Corum thinks 2008 may be even stronger for M&As than 2007.

Paid site members can find slides from the seminar in the SoftwareCEO Downloads Library, if you scroll to the section called "Software University Slide Sets."

Our Downloads library also contains copies of Corum Group's quarterly M&A report, starting from 2001. Scroll to the section entitled "M&A and Financing" and look for "M&A Briefings, 2001-2007."

2007 software IPO market: The best in years.
Last year we said 2007 should continue the "return to normalcy." But if anything, IPO activity got hotter than we ever expected, with more deals, better valuations, and more investors who didn't need to see profits yet.

Early in the year — before so much talk of the "R" word (recession) — conditions were quite favorable for an IPO.

In all, 23 software and service companies went public during 2007, up from 16 in 2006. These software IPOs raised a median of $100 million, up from about $81 million in 2006 and $86 million in 2005.

Across all industries, Renaissance Capital's IPOhome.com says there were 234 IPOs in 2007, with an average deal size of $229 million and total proceeds of $54 billion. And all three of those numbers were the highest in seven years.

Maybe we're finally out from under the shadow of the dot-coms. (We won't mention it again here, we promise: It's history.)

Technology was the largest single sector for IPOs, with more than a quarter of the year's total. And investors were reasonable in their expectations.

"2007 saw a rise in the number of unprofitable IPOs, as investors were attracted to fast-growing technology companies, particularly on-demand software," says IPOhome's 2007 Annual Review.

"But while investors were willing to overlook a lack of profits to date, they wanted to see companies that had demonstrated the scalability of their business models, and were close to booking profits."

Our numbers confirm these observations. And they see this trend continuing in 2008, with technology IPOs remaining strong.

What is a software company, anyway?
For the past few years we've been questioning what is, and isn't, a software company.

You may object that not all the IPOs that SoftwareCEO tracks are "pure" software companies. Or there may be companies on other industry lists that we didn't include.

You may be right. But at some point, we have to make a call.

For example, is TechTarget a software company?

In the old days, it would have been called a publisher, or a marketing agency. But these aren't the old days, and this firm relies on its web-based portals and the custom-wrapping on its downloaded content — webcasts, podcasts, and white papers — to deliver quality leads to IT clients.

So we call it a software service company, as opposed to a "pure" software company, and we include it in our analysis. You may not agree.

We reserve the "software" category for those with a more purist definition. But even that gets dicey sometimes.

For example, Constant Contact helps small businesses manage e-mail marketing, send out e-newsletters and surveys, and track customers through dirt-simple online CRM. And it does it all through its portal (aka SaaS).

So is it a software company? We say yes, but you may not.

For 2008 and beyond, we've started to tag software IPOs in two new ways. We're tracking SaaS versus licensed software, and vertical-market versus horizontal apps. We shall see, a year from now, what insights emerge from these new ways of slicing and dicing the IPO numbers.

Twenty-three software IPOs in 2007.
All in all, 2007 was a good year for doing a software IPO.

"In fact, the criteria for going public became a little more forgiving in 2007," says Sommer. "In 2006, the median income was break-even. But last year, even if you were losing $5 million, you were still OK."

Sommer says the numbers tell us there was more money out there, and more opportunities to invest in software IPOs.

At the beginning of 2007, these eight companies had already filed their intent to go public, but not yet set a share price:

• Bridgeline Software (NASDAQ: BLSW)
  web service development
• Buy.com
  computer and software retailer
• Comverge (NASDAQ: COMV)
  energy/electrical industry software
• GenuTec (NASDAQ: GENU proposed)
  messaging/conferencing IVR software
• Local Matters (NASDAQ: LOCL proposed)
  document management software
• Picis
  hospital management software
• Sourcefire (NASDAQ: FIRE)
  security intrusion protection software
• Veraz Networks (NASDAQ: VRAZ)
  VoIP software switch

Of these eight coming into 2007:

• Three went public during the year (Comverge, Sourcefire, and Veraz Networks)
• One went public early in 2008 (Bridgeline)
• Two are not yet public (GenuTec and Local Matters)
• Two IPOs were withdrawn (Buy.com and Picis).

Table 1 shows when each software IPO took place in 2007. They were distributed throughout the year, except for January.

Table 1: 2007 Software IPOs by Month

The outlook for 2008: More of the same? Or not?
In 2008, SoftwareCEO is tracking 14 further software companies who filed for an IPO but are not yet priced:

• Accuro Healthcare Solutions
• Aprimo
• ArcSight
• Convio
• Danger
• GlassHouse Technologies
• INFONXX
• Initiate Systems
• Intelius
• LogMeIn
• SS&C Technologies Holdings
• Synacor
• Syngence
• Varolii

This is a big pipeline, and if all of these go IPO in 2008, we should end up with a nice crop.

But anything could happen with the R-word in the air, not to mention the U.S. election, when both sides will no doubt claim that voting for the other guys will wreck the economy. These two factors may make investors pause in 2008.

How will that affect software firms seeking to go public in 2008?

"Software firms will probably have a tougher time doing an IPO in 2008 because of the economy," says Sommer.

"Once it's generally acknowledged that we're in a recession, that will have a cooling effect, and investors will hold on to their cash instead of investing in new companies going public."

If that happens, the alternatives could be painful.

"If a company founder can't see any other way, they can hold back their IPO, they can lower their IPO share price, they can do an M&A with another company, or they can decide not to exit at all."

We're hoping it doesn't come to that in 2008, but it could.

Investors were generous in 2007.
In 2007, our median IPO company lost <$4.8 million> on revenues of $46 million. Yet it still raised $100 million in exchange for about one-fifth of the company.

Table 2 tells the story.

Table 2: 2007 Software IPO Companies (millions of US$)

Investors were rather generous in 2007, reversing a trend for the past several years where IPO companies needed real revenues and real bottom lines.

In 2006, this trend softened, and in 2007 it melted away completely.

Our IPO firms did have some revenues coming in: more than $25 million on the low end, not quite $50 million on the median, and over $75 million on the top end.

But only the top quartile had any profits to speak of, and those were fairly modest.

"In 2006, you could break even, but last year you really didn't have to be profitable. You could have revenues of $46 million, lose almost $5 million, and still go IPO," says Sommer.

Revenue OK, but profits down in 2007.
The traditional guidelines recommend four consecutive quarters of profits and $50 million in revenues before going public.

But in 2007, the IPO market forgave the firms who hadn't yet reached these milestones. We believe this is in part because the very active M&A market is still softening these guidelines.

Table 3 shows the median software IPO revenue and net income since 1999.

Table 3: Software IPOs Revenue and Income
(millions of US$, medians at time of IPO)

In terms of revenue, in 2007 we had mighty VMware with sales over $700 million, trailed a long way after by Deltek with $228 million.

The only other IPO firm that topped $100 million last year was Virtusa with $124.7 million. Solera Holdings was close to this mark, with $95 million.

The others ranged down through the double-digits to the smallest, SoundBite Communications, with just $2.5 million.

So the median revenue for 2007 came in at $46.0 million, about the middle of the pack for the past five years.

On the income side, things are more intriguing.

Profits ranged from $85.9 for VMware and $19.0 million for Virtusa at the top end, to a net loss of <$23.4 million> for Netsuite and <$18.9 million> for Solera at the low end.

But in fact, almost three-quarters of our firms — 17 out of 23 — were in the red, which pushed the median income down to a loss of <$4.8 million>.

And that didn't seem to slow down anyone's IPO.

The highest multiples in four years
2007's revenue multiples were up right across the board, to the highest point since 2003.
And that's looking at three times as many IPOs: 23 in 2007 versus only eight in 2003. This means Wall Street had a good attitude to software IPOs last year, and investors were willing to take a risk on our sector.  
Table 4 tells the ongoing story.

Table 4: Software IPO Revenue Multiples

The revenue multiple is a key metric that shows whether investors believe that a particular company can grow and prosper. And the higher the multiple, the bigger the payoff for company founders.

In 2007, the typical software IPO was valued in the range of 6X to 10X revenue. Many were much higher, with nine firms who hit the jackpot with multiples topping 15X.

Don't forget, it now costs an estimated half a million dollars to go public, due to the tighter regulatory environment imposed by laws like Sarbanes-Oxley. But in 2007 our IPO firms seemed to clear that hurdle with little difficulty.

Once again, we called it too conservatively last year. We didn't think 2007 would be as robust as 2006, yet the multiples continued to climb.

Going into 2008, we have a strong pipeline of IPOs, balanced against fears of a recession and the wildcard of the U.S. presidential election.

It's hard to believe that multiples can continue to rise in 2008. But if they do, more power to the software CEOs who cash in.

Revenue-per-employee dips somewhat in 2007.
Another key metric for any company going public is revenue-per-employee (RPE), also known as "productivity."

For five years straight, productivity rose steadily among our IPO firms until 2006, when this number seemed to stall around $170,000. Then in 2007, it fell to not quite $150,000.

This is another metric proving it was somewhat easier to go public in 2007. The numbers in Table 5 show the employee metrics.

Table 5: Employee Metrics at Time of IPO

Of course, some of our companies had absolutely stellar RPEs.

Monotype Imaging had an impressive $574,667, Limelight Networks had $382,738, and six more firms had RPEs above $200,000. Those numbers should be acceptable in any climate.

On the low end, China's Perfect World had 430 employees and an RPE of only $30,000. SoundBite Communications and Virtusa both had RPEs less than $35,000.

That's just not sustainable for any company with any employees in North America or Europe. Sooner or later, these companies need to get the cash register ringing more heartily.

Looking at RPE by quartile shows a big range for RPE. The lower quartile showed $103,894 in 2007, with the upper quartile more than twice that at $226,772.

Many factors influence RPE.

For early-stage IPOs, the revenue may simply not be there yet. Or they may have hit a mother-lode that gives a very high RPE from a small number of employees.

Later-stage companies may have a moderate RPE, reflecting a more mature segment. Or, they may have a large RPE, flowing from a commanding market share.

In any case, any company considering going public needs to be conscious of its RPE metric, and how the market will view its numbers.

What happened since they went public?
We always find it interesting to look at how the recent crop of IPO firms did after going public.

Considered together as a category, the software IPOs fared well. By December 31, 2007 the median was up 23.8 percent.

That means these software IPOs heartily outpaced the overall NASDAQ, which was up 9.8 percent for the year.

But it was a split decision: 14 companies were up, while nine were down. And while the top firms gained a median of about 75 percent, the bottom lost close to 15 percent between IPO and year-end.

Table 6 shows these results.

Table 6: Share Performance, 2007 IPO to Year-End

Note: We calculate the medians from our complete spreadsheet data, so the deltas in the bottom row may not compute. For example, in the lower quartile in Table 6, the share-price difference looks like <9.5> percent ($11.25 to $10.19); but the differences calculated from all the IPO share prices in the bottom quartile come to a <14.6> percent decrease.

Who's hot and who's not?
Let's take a quick look at the winners and losers in 2007.

At first glance, the biggest winner looked like VMware, whose stock price almost tripled to give it a market cap of $31.9 billion by year-end — a jump of $21 billion in less than 21 weeks.

Of course, it didn't stay that high forever; see the note after this.

The next big winner was athenahealth, whose share price doubled after its IPO to push its market cap to $1.1 billion.

Five more firms rose 70-something percent by year-end, and two more were up 50-something percent. Wow.

On the downside, the worst performers were Glu Mobile and Limelight Networks, which each lost half their market cap by year-end. In dollars, Glu Mobile was down $444 million, and Limelight lost $635 million.

Sourcefire and Veraz Networks were each down around 40 percent in market cap, losing $154 million and $126 million respectively. Ouch.

So although our median IPO firm was up 23.8 percent at year-end, there was a lot of play on both sides of this number. The winners won big, and the losers, well... let's hope they turn it around in 2008.

A note on VMware's stumble...
Some called virtualization software maker VMware, a spinoff from EMC, "the most anticipated tech IPO in years."

It started off with a bang in August: Initially priced at $29, the shares started trading at $57. They hit a high of $125.25 in late October, before ending 2007 at a very positive $84.99.

That made VMware for a time the fourth most highly-valued software company in the world, surpassed only by Microsoft, Oracle, and SAP.

Then came the dark day of January 28, 2008, when the company announced its Q4 2007 earnings.

Net income more than doubled, and revenues were up 80 percent over last year... but that wasn't good enough for analysts who'd been expecting 82 percent.

Add in some muddled comments from company execs — who mentioned the dreaded concept of "slower growth" now that the company was so big — and investors fell all over themselves to punish the stock.

After one day of frenzied trading, the shares nose-dived to $54.87.

Some call it a market correction. Some call it jittery investors. Some call the stock a bargain at its new, lower price. It's all part of the game of being publicly traded.

"New IPOs tend to be more risky," says Sommer. "This could be partly the volatility in the market, and partly that VMware is a relatively new offering, where the risk is much higher than with established companies."

If you're interested in more coverage of this, here's what Business Week had to say.

And here's the investment site The Motley's Fool's take on VMware.

Far-ranging market caps.
Our 2007 market caps on IPO date ranged from $120 million to $11 billion. We'd call that a pretty wide spread.

At the high end, we've already discussed VMware.

Next in line was SaaS firm Netsuite, with an IPO day market cap of $1.5 billion, growing to $2.3 billion by year-end. Netsuite raised $161 million for just 10.4 percent of its stock, close to the smallest slice of equity for any IPO firm in 2007.

On the bottom end, SoundBite Communications offered 34.7 percent of its equity at IPO, for a market cap of $120 million. By year-end, its capitalization was down to $103 million; on both measures it was the smallest of all the 2007 software IPOs.

Looking at Table 7, valuations were up across the board to the highest point in four years.

Again, this seems to reflect a willingness by investors to pay more to get in on the action with a software IPO.

Table 7: Software IPOs Market Cap, 2004 to 2007

Software versus service: The flip-flop continues.
From one year to the next, there are intriguing similarities and differences between the IPOs we tag as "software" or "service."

As usual, four out of five of our "service" firms from 2007 are web-based, including:

• Limelight Networks, which provides high-speed networks for delivering content from clients over the web
• Perfect World, a Chinese-based service with online games and a virtual world called "Second World"
• Salary.com, which provides salary survey information through the web
• TechTarget, which operates portals for IT-related content, such as white papers, webcasts, and podcasts.

The fifth, Virtusa, is a more traditional outsourcing firm that does contract programming and consulting.

Remember, our definitions are open to debate. With that in mind, Table 8 shows the financial metrics of the 2007 software IPOs against the service IPOs.

Table 8: Software vs. Service Firms Doing IPOs in 2007

As you can see, 2007 brought bigger revenues and higher profits (or at least smaller losses) on the service side.

For the past few years, we've seen a flip-flop between these two categories. In 2004 and 2005, median sales and profits were higher in service. In 2006, they were both higher for software.

In terms of revenues-per-employee, software beat out service in 2007, with close to double the productivity. This is another switch, since before that RPE had been higher in service firms for two years running.

Many of the other metrics lined up about the same between the two groups.

Both median market caps were up at year-end, but look at this: Software was up 21.5 percent versus service, up only 6.3 percent.

What does this tell us?

It looks like the market pegged the service firms about right, but it significantly under-priced the software IPOs. You could even say that the pure software firms left some money on the table at IPO time.

Of course, hindsight is always 20-20.

A mixed year for software CEOs who went public.
Most software CEOs did well at IPO time: Their median holdings were worth $22.1 million, up a cool $2 million from 2006.

And by year-end, their holdings had increased by 23.8 percent, giving the median CEO an extra $4.3 million.

But not everyone got this nice little Christmas bonus.

The top quartile did particularly well, gaining 74.1 percent from the IPO date to year-end. This gave CEOs in that quartile a paper gain of almost $11 million each.

But in the bottom quartile, things weren't so rosy.

Share prices of these companies actually retreated 14.6 percent by year-end, erasing $3.5 million from the median CEO here. Ouch.
Table 9 shows you the money.

Table 9: Rewards for Software CEOs from IPOs in 2007

Note: How can the year-end value of the CEO's stake in the top quartile ($42.6 million) be 74.1 percent greater than the value at IPO ($31.7 million)?

Again, this is a quirk due to our calculation method. First we do the math for each line, and then the medians. That means we multiply each CEO's ownership x share price to get a number — the range in 2007 was $445,864 to $232.5 million — and then we figure the median.

To come up with the percentage change, we first calculate each CEO's gain or loss, then run the median on all those percentages.

Our analysis shows the median software CEO owned only 3.7 percent of the company they took public. This is way down from 8.5 percent in 2006, 5.2 percent in 2005, and 6.1 percent in 2004.

But it may be no cause for alarm.

"Yes, the CEO share is a little lower this year, but so much of this depends — particularly on the high end — on the individual circumstances of each company," says Sommer.

The CEO's slice of ownership depends a lot on the company history: how long the CEO has been there, how many rounds of financing the firm's been through, and how big the company has grown.

"There are typically one or two CEOs who have a higher percentage, depending on the history."

For example, the biggest CEO stake was 26 percent of China's PerfectWorld for founder, chairman, and CEO Michael Yufeng Chi , worth $232.5 million at IPO time.

The smallest stake in 2007 was 0.3 percent for Diane Greene, co-founder, president, and CEO of VMware. Her share was worth $32.9 million at IPO time.

Only two other CEOs owned in the double digits of the outstanding shares:

• Kent Plunkett, founder and CEO of Salary.com at about 25 percent (worth $37.7 million at IPO)
• Kris Canekeratne, co-founder, chairman, and CEO of Virtusa at 11.7 percent (worth $37.5 million at IPO).

There's another factor as well: CEOs of service firms tend to own more of the company than CEOs of software firms.

In fact, the top three owners mentioned above are all from service firms. In 2007, the median service CEO owned 11.7 percent of the company, compared to only 3.2 percent for software CEOs.

"To invest in developing software, you typically have to raise more money and give away more equity," says Sommer.

"Service companies tend to be more self-funded. So the CEOs of service companies can often hold onto a higher percentage than the CEOs of pure software companies."

We wish all those CEOs who went public in 2007 the best of luck with their companies.

Likewise, we wish any of you who decide that 2008 is the year to go public, or to do an M&A, all the very best with your plans. We hope that next year at this time, we have a lot of good news to report.

Editor's Note:
This report uses data from SoftwareCEO's M&A/IPO Tracking spreadsheet, which is available for download in the M&A and Financing section of our Downloads Library. The Excel file contains detailed M&A data from 1999 through 2003, and IPO specifics from 1999 through 2007.