瞻博網絡 (JNPR) 2004 Q2 法說會逐字稿

Welcome to the Juniper Networks, Inc. second quarter financial results conference call. During the presentation, all participants will be in a listen-only mode. Afterwards, we will conduct a question and answer session. At that time, if you have a question, please press the one followed by the four on your telephone.

As a reminder, this conference is being recorded Tuesday, July 13th, 2004. I would now like to turn the conference over to Ms. Michelle Levine, Senior Manager of Investor Relations for Juniper Networks. Please go ahead, ma'am.

Good afternoon, everyone. Thank you for joining us. With me today is Scott Kriens, our Chairman and CEO and Marcel Gani, our CFO. If you have not yet seen the press release, it can be retrieved at www.Juniper.net or off of First Call or Business Wire.

Today Scott will discuss Juniper's second quarter performance, as well as provide an update on the integration of NetScreen and comments regarding Juniper's vision and strategy, the networking industry and our outlook. Following Scott's comments, Marcel will review in detail the financial results for the second quarter ended June 30th.

Before I turn the call over to Scott, I'd like to remind you that the matters we will be discussing today may include forward-looking statements and as such are subject to the risks and uncertainties that could cause actual results to differ from those contained in the forward-looking statements, including those risks and uncertainties discussed at our most recent form 10-Q filed with the SEC. We are also presenting some non-GAAP financial information. A reconciliation of GAAP to non-GAAP items can be found on our Investor Relations webpage.

Juniper Networks assumes no obligation and does not intend to update forward-looking statements made on this call. Scott, I'll turn the call over to you.

Thanks, Michelle. Good afternoon to everyone.

Today I'll be talking about the results and the execution behind the second quarter numbers that you've seen. The progress we've made on integration of the NetScreen acquisition, as well as an update on what we see taking place in the industry and how we are positioned, and then following my remarks, Marcel will discuss the financial results in more detail, talk about our stock buy back program, also announced today, which is a measure of our confidence in our cash generating capabilities and we'll of course provide our guidance.

So first, to the second quarter's results, as you have seen from the numbers in the press release, we had another strong quarter across our business in both infrastructure and security products. This is not only measured by revenue, but on many other metrics as well, which Marcel and I are about to review. We reported revenue of 307 million and non-GAAP EPS of 8 cents, GAAP EPS was a loss of 2 cents for Q2. As you'll recall from our discussions last quarter, we must abide by certain purchase accounting rules that apply to Juniper's acquisition of NetScreen and later in the call, Marcel will walk you through a reconciliation of our financials. Please see the press release on our website for the reconciliation of non-GAAP to GAAP results.

With that said, in accordance with those rules, we recognized revenue of 307 million, but one metric that's not effected by purchase accounting rules is cash generation.

Our cash flow from operations was $119 million, as we invoiced and collected cash from our customers in exchange for products and services as in any other quarter and this result easily is your passed our goal as the quarter began of exceeding $100 million. With the acquisition of NetScreen, we now group our products into two main categories, infrastructure products, which are the traditional router products and security products.

Total revenue for infrastructure products and services was 255 million, which was up 14% from last quarter and up 55% from the year ago period. Total revenue reported by Juniper for security products and services was $52 million. Marcel will provide details later in the call for our estimates of security products results independent of the purchase accounting rules. For infrastructure products, we recognized revenue on a total of 1 ,551 units this quarter and we shipped 28 ,864 ports.

I'm pleased to report that both units and ports are up from last quarter. We have seen more demand this quarter at both the edge and the core, as we see the increased momentum of broadband expansion. Once again, this demonstrates the positive business cycle and the continuing interconnection of demand from core to edge as seen for the last several years. Security products also made a strong contribution to the quarter, as that business continues to grow its presence in the marketplace. We're pleased to report that Siemens again represented more than 10% of total revenue during the quarter, and this quarter as total revenues have grown as a result of our security acquisition is our only 10% customer.

In the case of both Erickson and Lucent, we've seen increases in sales in absolute dollars when compared to last quarter and we continue to be pleased with both the strength and breadth as well as the momentum behind both relationships. These results have been achieved while simultaneously executing on the integration tasks associated with the NetScreen acquisition. We have taken a very aggressive strategy on integration, which was chosen with full awareness of the possible risk to the short-term results; however, as the numbers indicate we've been able to both integrate and execute to a very high level and I'll talk more about this in a few minutes.

This quarter there was an unusually high amount of speculation and sometimes even silly rumors around our performance, our integration and our partnerships, and has always been our practice we spent no time rebutting any of this speculation, but chose to let the results of today be our answer. To the extent that the rumors were generated by our wishful competitors, I'll share the conclusion I took away from the chatter.

There are a lot of people hoping and praying that we don't execute on the integration and realize the potential we now have at Juniper, which to me is indicative of both the scale of our opportunity and the competitive advantages that we're enjoying.

Our customers, on the other hand, have spoken clearly, often and with loud voices and they're voting with their purchase orders. Expect the rumors to continue and know that we will continue to ignore them and that customers will continue to be the priority at Juniper.

So with that said, let's move to some of the messages that those customers have delivered recently. There were several announcements during the quarter and I'll quickly provide some highlights from theaters around the world. the Americas, our router and security platforms continue to be the right products at the right time, demonstrated by announcements this quarter, including U.S. LEC, Expedious, University of western Ontario, Stanford University, Hydrochloral, and ASI Guardian.

Additionally, Time Warner Telecom launched the largest VPLS deployment in the United States with their multipoint extended metro Ethernet services enabled by our routing platforms. In EMEA, we announced Italy's fast web and catch communications and significantly, BT group announced that it plans to switch voice calls to IP as its 21st century network takes shape.

Our partner Siemens is supplying a solution comprising its metro media gateways, plus Juniper's core and edge routers, notably including the TS 640 and the BT's trial network for that project. A.P.A.C. continues to innovate with Juniper Networks as a foundation for many leading edge services, demonstrated by announcements this quarter, including Shanghai Telecom and Jay Jing Telecom.

Aside from customer activity in the various theatres, we also made several partner product and technology announcements during recent months. Some of those highlights, last month at Supercom in Chicago where we collaborated with several key technology partners in delivering various solutions to the market, as one example Microsoft announced the demonstration for delivery of premium video quality for their IP TV software solution over the Juniper E series platform.

We announced an alliance with Turion, which further brands our strategy of partnering in cable to enable the delivery of advanced cable broadband services. And in May, at N+I Polycom showcased a video services demonstration, which highlights the integration between our SDX service deployment system and Polycom's past navigator platform.

We announced that our NetScreen secure access appliances are the industry's first SSL VPN's to integrate the popular security assertion markup language, or SAML standard. And we also announced that we are teaming with leading industry partners, such as IBM, Netegrity, Oblix and RSA Security to help expand secure extranet access.

This morning, it was announced that Juniper and Microsoft are collaborating to enable Juniper products to enforce network access and security policies based on client security profiles provided by Microsoft's network access protection. The customer's biggest risk of infection of worms and viruses is from remote access and W/LAN users and by collaberating with Microsoft, Juniper extends existing network access control capabilities for remote access and WLAN users. There are also several events on the product and technology front.

In April we marked two years of T series deployments with a growing customer base of most of the largest production networks worldwide and also announced new intelligent logical router services to compliment our rich JUNO software feature set. Customers who have transformed their networks with T series include companies such as China Telecom, China Unicom, Trans Telecom, NTT Communications, SAIC for the DISA GigB program, SK Telecom, TCom, a division of Deutsche Telecom, Telecom Italia, Telephonica, Telio Senora, and I could go on. The T series is the core router in some of the largest research networks in the world, including Internet 2's Abilene network, Dante's Geat Network, the National Science Foundation's Tara grid project network, and Tokyo university.

We continue to lead in the delivery of technology to solve the problems of our customers and our competitive advantages and our real world lead have never been more apparent.

In June, we introduced the J series services router family of products and the NetScreen 5GT ADSL security appliance designed to meet the needs for improvement in the networking of widely distributed business critical applications. We recently announced the availability of our security products operating system, screen OS, with production-ready support for IP version 6. Juniper's security products, with the new IP version 6 enabled screen OS represent the first in the industry to provide IP version 6 support for both state inspection firewall and IP SEC VPN capabilities.

We rolled out J voice with our new high availability features so carriers can migrate their traditional voice services onto cost effective IP networks with flexibility to offer differentiated VoIP services according to customer needs. And we extended the functionality of the SPX 300 system to include support for JUNO's based NT and the new J series routing platforms.

Finally, that Juniper has been awarded the prestigious PL 9000 certification by the quality of excellence for suppliers of telecommunications forum, or quest. And Juniper is distinguished in being the first and only North American IP routing vendor to achieve this honor in the router category, a testament to the company's focus on excellence in meeting the needs of its customers. So measured by any standard, it was a strong quarter and one in which we executed on many fronts. We delivered products, we added customers, we exceeded our financial goals.

We added to our competitive advantages in multiple markets and geographies and all of this was done while integrating a major acquisition that adds enormous reach and brand opportunities for Juniper. I would like to spend a few minutes on that acquisition and I'll provide an update on the integration activities. As you'll recall, we closed the acquisition on April 16th, less than three months ago.

There are three general approaches one can take when integrating two companies.

The first is to prioritize the commercial performance of the business, protect the run rate and accept the tradeoff of moving slower on integration. Buy now, pay later, if you will.

A second approach is to move quickly and integrate comprehensively and accept the consequences to the financial results and the customers, of disrupting the business immediately upon taking control of the acquisition.

And the third approach, which is my favorite, and the strategy we're succeeding with at Juniper, is to move quickly and decisively on the integration to maximize the leverage available in the new markets with the new teams and to do so without missing a beat in hitting the financial and customer goals we've set for ourselves. Several specifics to illustrate my point, we have hit each of our goals on the financial performance both revenue and cost for our security products and infrastructure products, which Marcel will cover in a minute. All organizations have been integrated.

As you may remember, immediately after the close of the acquisition, we announced Kittu Kolluri as the leader of our new security products group, as well as the leaders of all the other functional organizations across Juniper following the acquisition. Today's update is that each of those leaders has defined and rolled out their teams and each person in Juniper is clear about their roles and responsibilities.

Budgets, strategies and goals across all of Juniper for the remainder of 2004 have been developed and delivered throughout the company. We've held kickoff meetings worldwide over the last quarter with our newly integrated sales force in each theater and also separately with our partners in meetings around the world with regional affairs in Spain, Thailand and the United States. We've met with our partners to understand the best strategies for our companies going forward and we've followed that with the rollout of our new J partner program worldwide.

There have been and will continue to be cross training programs for both employees and partners as we continue our investment in channel expertise to better serve our customers. We have fully integrated our information systems and are taking orders and shipping on an integrated system for all functions, and have disconnected all parallel systems acquired as a result of the NetScreen transaction. This is an incredible accomplishment and a testament to our CIO, Kim Pernicue, and her team in their planning and execution.

So to summarize on integration, quite simply, on budget, ahead of schedule, and ahead of plan. Juniper set a new standard for speed, clarity, focus and execution and from a NetScreen perspective, it's a testament to the quality of the company and the team of talented people that have joined us.

Separately, there's a lot happening in the networking industry today and I'd like to close with a perspective on the communications industry as we see it and Juniper's strategy and role as that industry reshapes itself into the new reality. The industry is adapting quickly to a new equilibrium. That's represented in part by initiatives such as the infranet, which is a good model for the directions we see.

The Infranet initiative, which we have spoken about before, is now well beyond Juniper itself, as major players such as IBM, Oracle, HP and many of the world's largest carriers, including Qwest, Deutsche Telekom, China Unicom, Orange, NTT, and many others have come together to agree on a model for networks and applications that run across those networks that provide users with an assured experience they can trust for its reliability, security and performance. This is what will be required industry wide for us all to deliver on users expectations. And a measure of the importance of this vision is the speed with which it's progressed.

From the announcement to the infranet concept at Telecom Geneva last October to the most recent meeting at Supercom in Chicago a few weeks ago, the group has grown dramatically in size and stature and has been defining a reference architecture upon which submissions to standards bodies will be based, that will provide a universal agreement on how we as an industry will provide networks and services across private and public domains. The growth of the membership and the speed of decision-making underscores the importance of the subject matter. This work is being carried out by the infranet initiative counsel, of which Juniper is only one member and further information on the group can be found at www.infranet.org where the council explains in detail its progress and plans.

But the message is clear. Make networks smarter, faster, and safer. Smarter so that individual needs can be supported on a personalized basis; safer so I can run my business large or small across a shared infrastructure to take full advantage of the economies of scale that represents, without fear for the safety of my information or that of my customers, and faster so that my online games, my music, my movies, and my telecommuting needs can be served as if I were directly connected, though I may actually be anywhere in the world at any time connected by either fixed or mobile means. It is so obvious that this is where the world is going that sometimes it's frustrating at Juniper that it's not all happening faster. But it is happening and the pace is accelerating.

Our strategy, quite simply, is to further accelerate this inevitability, to make the network the full partner to us that it will undoubtedly be, and to make that happen today and not tomorrow. We see the market segmenting into clear areas of opportunity that will form the basis for the new industries that will be born.

Much as microprocessors, software, and personal computing were born, from the early day of the now oversimplified concept of basic computing, the networking industry is similarly oversimplified today and will segment as well. We see at least four distinct and separate opportunities and successful comments competitors focus on them separately from one another.

As the world of applications through the software that interfaces with us all as we network to our information and to each other, and then there are the devices, the cell phones, laptops, PDA's, IPODs, servers, all blending more and more with today's concept of consumer electronics to deliver us high volumes of standards based, low margin, cost effective tools to run our applications, and all of these devices will connect to a network processing layer responsible for the intelligence needed to deliver the assured experiences we will all pay for in accordance with our needs and our individual requirements. And this network processing layer is where Juniper will continue to define and lead a new industry.

Finally, the network processing layer will connect to the physical transport required to actually connect us to each other around the world. The distinction between public and private networking will diminish, as intelligence responsible for connections and experiences deliberately removes the distinctions of location and network ownership from the user domain with increased intelligence and security. And we are accelerating the blurring of these lines with both our acquisition of NetScreen and our recent announcement of our J series products, simply extending the intelligence of the network processing layer to reach the end points of the network where they can connect directly and securely to devices.

Rear view mirror analysis of our strategy has attempted to force fit our expansion into yesterday's definition of carriers and enterprises, while we serve a vision of delivering network intelligence and services, which reaches from device to device across any physical definition of the network that's necessary to serve user's needs. This is how users use the network and since users are paying everyone, public or private, service or technology provider, we prefer a strategy aligned with the users as our highest priority. An excellent illustration of the new world is the recent news of the Defense Department's new Gig B project. The worldwide network with the highest priority on security and reliability, where utmost performance is needed to deliver 3-D battle field imagery and information to troops in real time. There is no distinction as to the physical location or control of assets.

The worldwide, the mobile, by definition required to be flexible without compromise, and this is one of the best examples of Juniper's product portfolio at work as we build the state of the art network for serving our troops around the world securely, quickly and reliably.

We're enjoying these successes today by deploying modular software, version 6.4, built from six years of production experience in the world's most demanding networks, and we're competing with software and products based now on similar concepts, but at version 1.0, making its production network debut sometime later this year.

I wouldn't go so far as to suggest we therefore enjoy a six-year lead, but suffice to say that we're more excited than ever to invest in our advantages and to extend our lead, now that we've all agreed on modularity as the best way to solve today's technology problems.

So in summary, a few final comments. Our strategy is to focus relentlessly on the emerging industry we define as network processing and to be everywhere our customers require the virtual network to be.

Our belief is that success will come with focus. This has been our belief since we started the company in 1996 and has not waivered as we've grown our cash flows from operations as a business from zero a few years ago to today's announced results where we're now generating over $1 million a day with plans to continue that growth. We are in an industry which will recover selectively. Not all the sellers will recover because not all the buyers will recover.

Only the best companies, who focus on the correct definition of the future, and build their franchises there in advance of others, will enjoy the full benefit of the industry's realignment. Gains will come at the expense of others. There will be losers as a result. Growth, cash and financial fundamentals have returned as the basic measurement of success and this will ultimately separate the participants.

There is a great deal of work yet to be done. When the history is written, these will be the best of times for the winners. The time when the destination and the goal was clear and before it has been even understood by everyone, let alone achieved. We have the opportunity to build one of the great companies at Juniper and we will not let that opportunity pass us by.

I would like to thank and recognize all of our employees for their continued commitment and incredible efforts on achieving these results and on the execution of our plan and the strong results this quarter. This has been an especially challenging and exciting time for everyone, as we've had to take on a second job of completing our integration and I'm especially proud of everyone in our combined company for their focus and commitment. We'll benefit in numerous ways from the execution of so many teams of talented people and I appreciate it very much.

I would also like to thank our customers, our long-term shareholders, our partners, and our suppliers for their continued confidence in Juniper networks.

Before I pass the call over to Marcel, I would like to provide one final update, in this case, regarding Marcel himself. As many of you know, Marcel has been responsible for Juniper's financial results as our CFO since 1997 and has been in the industry for more than 30 years. Marcel will continue to serve in his capacity as CFO and VP of business systems for Juniper for the remainder of 2004 and will certify the audited financial results of the company for our full fiscal year.

But we are providing advance notice to you today of our plans for Marcel to change roles in the company in 2005 and to relinquish his current responsibilities and to assume an ongoing role as an individual contributor and a full-time employee. This is in part a personal life-style decision for Marcel to achieve a better work/life balance, which I'll let him comment on in a moment, and in part, a way for us to repurpose his skills and experience on the opportunities outside the day-to-day operations of the company, as we better leverage our position and our potential in the industry we're building.

On the one hand, I'm of course sorry to see Marcel approach the changing of his current role, but I am both confident we'll fill this gap and I'm very excited about the possibilities Marcel can bring with his abilities applied on a full-time basis to expanded opportunities at Juniper. I have tremendous trust and confidence in Marcel and I look forward to his counsel and his contributions at Juniper for many years to come.

With that, I'll turn it over to the man himself, and he'll comment further and we'll of course also review the financial results and our guidance. Marcel?

Thanks, Scott. First, let me state that this is a bittersweet decision for me. Over the last year, I've had many discussions with Scott regarding my personal interest and my career and how to achieve those goals. I want to achieve a better balance between the time spent with my family, my personal interest and the business, while insuring a smooth transition for my existing position. Today's announcement reflects my feeling that the company is extremely well positioned.

I believe that we have done an excellent job of managing through the steepest business cycle I've seen in 30 years, as Scott was nice enough to remind me, and as today's results show, the company is very well positioned to take advantage of the shift towards virtual networks. I'm very pleased with the broadening of our market reach through the acquisition of NetScreen and especially with the success of our integration efforts to date.

I believe that the transition for my position can be successfully accomplished over the next six to eight months without any impact on the business, and in such a way to add talent to our already strong management team and prepare for the increased complexity and scale of our business. I remain committed at Juniper to completing the requirements of 2004 financial year and the subsequent transition of my duties, as well as my new role looking forward. With that said, I'm back to business as usual and I will now move on to the financial results.

I would like to start by stating that I'm very pleased with the financial metrics for this quarter. Total reported revenue for Q2 was 306.9 million, an increase of 37% from last quarter and 86% from a year prior. As a reminder, Juniper closed the NetScreen acquisition during the second quarter on April 16.

As you will recall, due to purchase accounting rules, no company may take revenue for a sale that they did not incur the cost for. For Juniper, this includes old NetScreen deferred product revenue prior to the close of the acquisition. Deferred service revenue is also impacted by the purchase accounting rule for the same reason. It is important to also know that revenue and expenses for the security product group were only counted for the last 75 days of the quarter. We expect that the numbers will begin to normalize in Q3. We estimate that NetScreen would have recognized in excess of $100 million of revenue as a stand-alone company in Q2.

For the quarter, Juniper reported $52 million of security product and service revenue. The amounts, we cannot recognize due to the purchase accounting rules, are 38 million of deferred revenue, and 11 million of revenue that was recognized prior to the close of the transaction. There will also be 9 million and 6 million of service revenue that will not be recognized in Q3 and Q4 respectively due to the same reasons.

Infrastructure product had a strong quarter. We reported 255 million for Q2, which is up approximately 14% from the prior quarter and 55% from the prior year. Infrastructure service revenue was 35.4 million, up from last quarter. We are very pleased with both unit and port counts in Q2. The edge represented more than half of our revenue and we're pleased with the growth in our core business, which grew in absolute dollar from last quarter.

I would like to take a minute to review our revenue recognition policy. Recognized revenue on identified and user shipment, whether they are direct or through VARs, were recognized revenue sold through distribution based on sales out or point of sale report received from the distributors. The total book to bill ratio was greater than 1 in the quarter. The following financial metrics are based on our reported financials that included the impact of purchase accounting rule.

As Scott mentioned, our channel partner Siemens was a 10% or greater customer this quarter, generating approximately 16% of total revenue, based on over 40 end customers. We continue to see diversification of our channel partner. Both Erickson and Lucent were strong contributors this quarter ,and showed an increase in revenue from the prior quarter.

As we have stated in the past, our business will be lumpy by application, as well as by geography. We're pleased with the growth across each region of the world. The Americas represented 49% of total revenue. We're also pleased to state that the U.S. has now shown sequential growth for five quarters in a row.

Europe represented 27% of total revenue and Asia represented the remaining 24%. Revenue through our direct sales force consisted of 31% with the remainder going through global and country specific distributors.

Gross margin was 68.5%, up from 67% last quarter. However, without the effect of purchase accounting rules, gross margin would have been 70%. The increase from last quarter is due to the addition of higher gross margin from our security product. Total service margin was 47%.

Please note that the following operating expense include only a prorated amount from the security product group due to the effect that we closed the acquisition on the 16th of April.

R&D expenses were 58.1 million and accounted for 18.9% of total revenue, which compares to 46.6 million or 20.8% last quarter. The dollar increase is primarily due to the addition of R&D personnel expenses from the acquisition of NetScreen.

We also had head count addition in both the infrastructure product group and the security product group.

Sales and marketing expenses were 75.7 million and accounted for 24.7% of total revenue, up from 43.5 million, or 19.4% last quarter. The dollar increase is due to the addition of sales and marketing personnel expenses from the acquisition of NetScreen and higher commission expenses, due to the overachievement of targets, and the addition of marketing programs to build our business and get leverage from our new channel.

G&A expenses were 17.4 million and accounted for 5.7% of total revenue, which compared to 8.9 million or 4% of total revenue last quarter. The dollar increase is due to the addition of G&A personnel expenses from the acquisition of NetScreen, as well as litigation expenses, including some one-time settlement expenses.

Old non-GAAP references that I discussed exclude the amortization of purchasing tangible deferred compensation restructuring charges, in process R&D and integration expenses and loss on redemption of convertible debt. Please see the press release on our website for the reconciliation of non-GAAP to GAAP results.

Operating expenses increased to 49.2% from 44.2% of total revenue last quarter and we're up to 151.1 million compared to 99 million last quarter, mostly due to the acquisition of NetScreen and head count additions. As expected, our operating margin as a percentage of revenue showed a decrease due to the impact of purchase accounting rules. Operating income was 59 million, or 19.2% of total revenue, compared to operating income of 51 million, or 22.8% of total revenue last year. Net interest and other income totalled 3.8 million compared to 2.5 last quarter. This increase was due to the redemption of the remaining 4.75% convertible subordinated notes and our increased cash balance.

We have been able to achieve a position where the effective tax rate for the combined company for the second quarter remained at 32% and the resulting tax provision is 20.1 million. Net income increased for the quarter to 42.7 million, or 13.9% compared to 36.4 million or 16.2% last quarter.

Diluted earnings per share decrease in Q2 to 8 cents versus 9 cents in Q1 due to the effect of purchase accounting rule. On a GAAP basis, which includes the amortization of purchase intangible, deferred compensation, in process R&D restructuring expenses, integration expense, and loss on the redemption of convertible debt for a total of 67.9 million in Q2, our operating expenses totalled 214.9 million and the net loss was 12.6 million, or 2 cents per share, compared to net income of 33.5 million, or 8 cents per share in Q1.

Now, a few comments regarding the balance sheet. Cash, cash equivalents, short and long-term investments total approximately 1.5 billion. We're very pleased to announce that we generated approximately 119 million in cash flow from operations during the quarter.

As Scott mentioned earlier, today we announced that our board of directors approved the new program to repurchase up to 250 million of common stock. We're very comfortable with our cash position and cash generating abilities and we feel that this program will help reduce the validity of impact of issuing stock. We'll remain committed to responsible financial management and the maintenance of a strong balance sheet.

Account receivable was 140.7 million and days sales outstanding was 42 days, up from 36 days last quarter. It is a little higher than our target range of 30-40 days, due to the impact of purchase accounting on our revenue, while we keep invoicing and collecting from our customer. We expect this number to fall back to our normal range in the third quarter. Deferred revenue was 162.5 million, up 62.2 million from last quarter. This increase is primarily due to the addition of the remaining deferred service revenue from NetScreen, new deferred created by shipments to distributor, as well as an increase in deferred revenue for infrastructural products and services.

Capital expenditures was 12.1 million and depreciation was 10.1 million during the quarter. We ended the quarter with 2 ,552 in total head count, up from 1 ,585 at the end of last quarter. This is attributed to the addition of 913 employees from the acquisition of NetScreen, with the reduction of certain positions due to the duplication or refocusing of certain activities. This means a net addition of 54, mostly in engineering and in the field. We expect this trend to continue, as we will move to take advantage of the opportunities created by the acquisition while maintaining the focus on our financial metrics.

With regard to the acquisition, I was extremely pleased with the integration efforts given that the close process went as smooth as it ever has. Despite the added complexity of the integration, we were able to close the quarter in two days. I would like to thank the entire organization for an outstanding job in consolidating all of the system and for flawless close process.

Now, for our goals and guidance, we'll continue to focus on our financial fundamentals going forward. We can't predict a level of business each quarter, but we're managing through financial planning and would like to share those plans with you. I will now provide with you guidance for Q3 and for second half of the year. Q3 is historically a seasonally slow quarter in Europe; however, we are forecasting total revenue of 366-370 million, which excludes approximately 9 million of lost service revenue due to purchase accounting rules, resulting in approximately 11 cents of non-GAAP EPS.

We expect gross margin will be up in Q3 in the range of 69-70%. We expect shares in the range of 590-600 million as we count the weight of the shares from the NetScreen acquisition for the full quarter. This share count also assumed continued dilution from the 0% convertible, as some of you know the rule on contingent convertible are currently being reviewed by the EITF. A GAAP EPS target is not accessible on a forward-looking basis due to a high variability and low visibility with respect to the non-recurring charges which are excluded from the non-GAAP EPS estimate. We'll discuss a long-term financial model for the business by the end of the year once we have established trends and move past the impact of purchase accounting.

From a balance sheet perspective, once again we expect to generate in excess of 100 million in cash from operations in the third quarter. We're increasing guidance for the second half of the year given the strength of our business from 705 million-720 million in revenue to approximately 760 million-770 million in revenue. Which excludes 15 million of lost service revenue due to purchase accounting rule. This guidance results in approximately 23-24 cents non-GAAP per share for the second half, reflecting accretion from the NetScreen acquisition in Q4.

These forecasts are forward-looking statements based upon estimation and the actual results can vary for a number of reasons, including those mentioned in our most recent quarterly report on form 10-Q filed with the SEC.

Finally, we will continue to focus on our objective of delivering high quality financial metrics. Now we would like to take questions. Please limit yourself to one question per person.

Operator, please instruct the audience regarding the queueing process.

Certainly. Ladies and gentlemen, if you would like to register a question, please press the one followed by the four on your telephone. You will hear a three-tone prompt to acknowledge your request. If your question has been answered and you would like to withdraw your registration, please press the one followed by the three.

If you're using a speaker phone, we ask that you please lift your hand set before entering your request. One moment, please, for the first question.

Our first question comes from the line of Christin Armacost from SG Cowen. Please proceed with your question.

Thank you. Nice quarter. I was wondering if you could give us perspective on what you see out of the Chinese markets, we're hearing some potential slow down with telecom spending and would love to get your perspective.

Christen, we didn't really see that in Asia, generally speaking, or China in particular. I don't think it's that different than the marketplace in total, which is that it's hard to draw conclusions within the growth areas in the market when aggregate comments are made about slow dow,n because that typically translates to mean slowing on the legacy spending, but continuing to be focused in the strategic areas, which is really where literally 100% of our portfolio is placed.

So it isn't to say that that slowdown in some categories might not be occurring, but, or that it might not eventually have some consequence in our business, but we haven't seen that translation in, in at least the business to date.

Thank you.

Next question, please?

Our next question comes from the line of Alex Henderson from Smith Barney Citigroup. Please proceed with your question.

Great. I have actually a question of clarification. Marcel, I think if you add all those numbers up, it's 116 million in NetScreen revenues, is that correct? Then the question is we had heard comments about Cisco implying that network utilization in North America was tightening quite sharply over the last couple three or four or five quarters and that they were concerned about network capacity tightened to the point where there should be an acceleration in demand.

Interestingly, Nortel made an almost identical comment saying they were finding quality of service difficulties when rolling out VoIP installations at some of the service providers because of tightness on some of the routed networks. Can you comment on what you're seeing in terms of the capacity tightness in the network and whether that's in fact a changing construct?

Alex, let me take the first part here. In terms of the NetScreen number, the total is really a little over 100 million and obviously, you know, those numbers are estimate, but we recognized about 52 million in our quarter, in our results.

There was 38 million of deferred revenue that was not recognized that went away, and there was $11 million that was recognized, or was not recognized, that was revenue that happened before the close of the transaction. So these how I get to the 100 million.

So is that excluding the 11 million?

No, that includes the 11 million.

I see. I thought you said the 9 and a 6 as well though.

Those are for future quarters, so the 9 million is service revenue. So the deferred that was lost, 9 million is service revenue that would have been recognized in Q3 as service revenue contracts over a year period.

I get it.

And 6 million are service revenue contracts that would have been recognized in Q4.

Great. That helps very much. Thanks.

so 15 million for the next, the second half of the year. And I'll let Scott talk about the tightness of the network.

Alex, I'm not-- I don't disagree with either of those observations made by Cisco or Nortel.

All of us struggle with trying to measure this stuff because it's somewhat anecdotal and there is not really a repository where one can go and get a quantitative point on all of this, but I think we would certainly agree that as usage continues to go up and by at least most measures in research, it seems to approximately double on an annual basis and of course we're doubling large numbers and some people put that number in the 70-80%.

Some people put it at 100-plus, but it's changing a large number. But the thing that is sometimes not clear about this capacity question that causes the issue around the requirements to quality of service and network intelligence in general is that there are two things. One is the network has to be built to handle peak at a moment in time and so you can have plenty of idle capacity in the aggregate, kind of like traffic jams in rush hour while there's lots of empty roads at midnight. You have to design a network to function 7 x 24 so capacity can go away much more quickly measured by that requirement.

The second thing that happens in any network when everybody's connected to everybody, the burden that places on the central infrastructure, just the mathematics of having any connection go anywhere at any time is that you need a huge amount of capacity in the network, relative to the actual traffic at any one end point or any one node measured at any one instant of time. There is a non-linear impact on the, on the network requirement to support capacity given that people can use it in any mode.

So, I suppose a longer explanation on that for another day, but it translates into meaning that capacity constraints come upon us all much more quickly than it would seem like if someone just measured, you know, the aggregate amount of bandwidth that was out there and made assumptions because those would turn out just in queueing and network design practices to dissipate pretty quickly.

So it's a bit qualitative for all of us, but we certainly see the similar kind of condition occurring and frankly don't see any let up in the demand or use of the networks, so I would expect it to at some rate to continue.

Great. Thank you.

Uh-huh.

Next question, please?

Our next question comes from the line of Stephen Kamman from CIBC World Markets. Please proceed with your question.

Marcel, we're going to miss you. I mean that honestly, but anyway, I wanted to ask about sort of where you go with the gig B and the government markets business. My sense is probably NetScreen may be part of that, but do you see any kind of incremental fall in contracts in the near term? Can you just talk about where that's going and how that connects to infranet, obviously some NDA's there, but any color you could offer.

Steve, couple comments on that. First of all, don't worry. Marcel is not going anywhere. Changing roles, but still here and still available, so no need to miss him.

Secondly, in the government area, actually one of the things that has happened as a result of the NetScreen acquisition is that NetScreen has enjoyed considerable activity and as a result, several customer relationships that have been built across many organizations in government, not only in the, in the defense side of the business where a lot of our focus has been, but also across many of the civilian agencies, so not all those relationships obviously, in fact none of them are as large or significant as gig B, but they all represent now a customer base that is at least by name of agency considerably expanded in government and with the integration of the two teams, the pre-existing Juniper team and the NetScreen team into a single organization that, Debbie Beinhart, our Executive in charge of the government business runs.

We're very encouraged by the opportunities there because people are looking at obviously security issues, but also there is probably a bit of, I would probably put this qualification around this statement to some extent, not a wholesale 100% move across all government sectors to IP, but there is a huge amount of focus on IP and new generation infrastructures and an example like gig B and the defense industry is really a powerful benchmark for an agency that has technical depth and as demanding a requirement as anyone could possibly have and studied the situation intensely for extended periods of time and is now rolling out the network that they have chosen, so we're very encouraged about the opportunities in that market and our team on the field there and continue to invest resources in it.

And just can I ask, IP V 6, is that a key element there and do you feel you still have that lead in terms of rolling that out?

It's, it's a good question, it's a very key element, particularly as a number of devices here just multiplies like crazy and the address space becomes precious. It's an issue in that marketplace.

It's an issue in the security market in general, and that we've addressed with the inclusion of IP V 6 and our screen OS capabilities, and probably most aggressively in terms of infrastructure is in markets like Japan where we've rolled that out some time ago and today enjoy the only integrated IP V6 capability within our product infrastructure straight through the silicon and we think that's going be a significant differentiator for sometime to come, and particularly as IP V6 accelerates in importance.

Thanks very much.

You're welcome.

Next question, please?

Our next question comes from the line of Erik Suppiger from Pacific Growth Equities. Please proceed with your question.

Yes. With regards to Asia, you had commented that you're not seeing a slowdown, but I think that Juniper's standalone was doing something in the range of 29% of revenues there last quarter and NetScreen was something in the range of 28, so given you're at 24, can you just explain what the discrepancy would be?

Yeah, Erik. I think as we mentioned, obviously, you know, it's always hard to look at trends because the business tends to be lumpy, but actually, shipments in all of the regions were strong this quarter. What did happen is that some of the deferred revenue was-- there was more deferred revenue internationally than there was in the US and that's what created that distortion.

Marcel, can you comment as to whether there was any turnover amongst employees in light of the acquisition? In Asia?

We haven't seen much turnover. We've you know, as we mentioned in the script, we have reorganized some areas to eliminate some duplication and refocus, but all of that has been done basically according to the plan that we had set in motion at the start of the acquisition.

Okay. Thank you.

Next question, please?

Our next question comes from the line of Gina Socalough from Buckingham Research. Please proceed with your question.

Thank you. Nice quarter. Can you equate the product and service breakout in revenue and in COGs with an infrastructure and service breakout that you report to us-- I'm sorry, the infrastructure and security breakout that you reported to us?

Gina I'm not sure if I understand your question.

Okay. On the breakout, you give us for revenue products and service. You-- in your script, you gave us infrastructure and security. How much, can you break out infrastructure and security by product and service? infrastructure by product and service, security by product and service, for revenue and then for the COGs, so we can equate the historic numbers with what you're reporting going forward.

I think that's what I did, right. I mean the 255 and the 100 million are product and services for each of the respective product lines, so I'll be happy to, you know, give you some more information after the call. When you break out infrastructure, that's, that's all-- that's product? And services.

Can you break that 255 into product and separately service?

Right. The 255 includes the portion of service that is related to the infrastructure products, right. As I mentioned on the call.

How much of the 255 was service ?

I would need to go back and check on that number. It was 35.4 million.

Service, and how much of the security number was service?

I don't have that breakdown easily available because it's coming in those different pieces that we have talked about and that's why I would suggest I would help you a little bit later.

Okay.

Next question, please?

Our next question comes from the line of Raj Srikanth from Deutsche Bank Securities. Please proceed with your question.

Thank you. One question I have of, recently we have seen some of the software window support were having tough times in terms of meeting the numbers. Is this-- do you see it generally, Scott or Marcel, is it slowing down in terms of IP spending taking place and how does it play into NetScreen's business going forward? Thank you.

Raj, couple thoughts there. It's a little bit outside of our expertise here, but this is not so much a comment on software specifically, but we've also been seeing some of the shortfalls and challenges that others have talked about and so I'll just give you my theory here, which is sort of worth what it costs.

But anyway, I think that some of the difference that exists outside of Juniper's business and maybe the networking industry in general is there's more discretion in some of the capital purchase decisions like software that is available to be practiced by the buyers, whereas a lot of our business is a function of, of network usage, number one.

And number two, in the markets that we serve, whether it be security or build out of broadband services, there is a strategic imperative that success depends on investment in those areas, so there is no stand-still strategy that will succeed when it comes to securing the network or building out broadband service to grow revenues in the face of potentially falling legacy revenue in other network services and things like that.

So it isn't to say that software and other things aren't strategic, but there is a clear, many times compelling strategic imperative behind direct investment in the network and in the securing of that network and a lot of it's usage driven, whereas I think there is-- there may or may not be quite the same imperative or direct connection to the strategic priority, but there is clearly more control over timing that is not as dependent on usage for some of the capital purchases that may drive other sectors. So I can't tell you that's true, frankly, because I don't know, but if you ask my guess, that's what we see.

Is NetScreen continuing to gain market share?

Remains to be seen on how people report these things, but we're pleased with the quarter and certainly, you know, seeing the kind of opportunities that, that produce the quarter available to us as we go forward. So we're going to continue to grow the business, share counters will count share, but I'm pretty confident in our ability to continue to execute on the plan and to enjoy the momentum that drove the attractiveness of the acquisition in the first place.

Thanks, Scott.

Your welcome Raj, thank you.

Next question, please?

Our next question comes from the line of Brant Thompson from Goldman Sachs. Please proceed with your question.

Hi, Scott. I was wondering if you could give us an indication of how we should be thinking about the, you know, the longer term revenue synergies as we move to the back of the year and into '05 with NetScreen. Any way of quantifying that or, you know, average deal size? Just any color on that would be useful. Thanks.

It's a little-- I mean certainly the synergies are what drives the improved guidance that we offered here as we look across the second half. We're seeing opportunities begin to materialize, but actually think it's still a bit early to be completely convinced of the magnitude of those opportunities.

There is clearly -- there is not a network that doesn't have a, in many cases, a S.W.A.T. team of people reporting to a senior executive whose specific dedicated responsibility is security, whether it's a corporate network, a large distributed private environment or obviously the public infrastructure, security permeates and in fact other markets as well, government being another great example where the security issue and concerns over it are significant.

The thing that we see happening in the industry, and that meaning security as a subject, the threat is escalating at least as fast as the ability of technology as well as operations and policies to keep up with it. The threat is becoming more and more malicious in its intent.

It's not any longer simply kids in college trying to have fun with, with users. There are increasingly there's evidence of very focused efforts sponsored by various parties to, to impact the security of networks of all types, and so as long as we're in the condition where the threats escalating at least as fast as the response, I think we're going to continue to see a market that's, I guess it's good and bad, good for people providing the solutions to those things, bad for all of us until we can get a better handle on this.

And it's very hard to distinguish between security as a stand-alone proposition and the requirements of the infrastructure, because security doesn't exist on an island where everything that surrounds it is vulnerable and infrastructure doesn't exist without security. So there is clearly synergy out there. We're seeing some of it. It's early, and, and the threat and therefore the demand remains pervasive.

So, again, it's a lot of what drove our, our assessment of a better outlook as we look forward to the second half.

So should we assume that the incremental better half outlook is really owed primarily to synergies from the deal?

I don't know that I could be that precise. I think in fact, it would be false precision for me to try to attribute that exactly.

Our infrastructure business is a strong business right now. People are largely, if not almost unanimously consented around IP and MPLS and there's significant security capability within our infrastructure product line independent of our security products, which also address the issue of security without being a function of synergies from the acquisition.

So, you know, I think the trend is there and our ability to-- the real thing that we're doing, which is what creates the opportunity, is we're making smart networks that are fast and safe, and you can competitively if you looked at other approaches to this, you can relax any one of those three and the portfolios of others look a lot more attractive but nobody's willing to do that. So a network that's smart and fast and not safe or fast and safe, but not smart or whatever, doesn't really win the market competition. So, you know, being able to do all three is really what's driving our upturn.

Next question, please.

Our next question comes from the line of Subu Subrahmanyan from Sanders Morris Harris. Please proceed with your question.

Thank you. I had a couple clarifications on the numbers. Marcel, your guidance for the combined company not accounted for GAAP purchase accounting was around 340 million. The 355 million which seems like the actual number you reported, 50 million upside -- primarily comes from the infrastructure routing side, which is about 14% sequentially, NetScreen side up 7%. Can you talk about what the underlying trends are in those two businesses is it kind of just a cycle point where infrastructure is that much stronger, or is the IP budget a little bit weak, not as strong as infrastructure, which is why security is not quite as strong and also, just to follow-up, is there an EPS number related to that $355 million number that we should think about?

Well, Subu, I think actually we are very pleased with the performance of both the security business and the infrastructure business, so considering the effect of being in transition and the fact that we had to take, you know, as I mentioned, focus in certain areas, et cetera, re-brand the product and go through a significant amount of conversion, I think that the performance achieved in security was great.

And certainly it was as good, if not better, than what we expected and, you know, as you noted, I think the performance in infrastructure was also very good.

So the-- you know, the effect is that obviously we're coming in above the combined guidance we have given and the way we're looking at the business is as one combined entity. We're going to continue to report the numbers for security and infrastructure, but in terms of kind of how we manage the business, what is important to us is the overall top line and then how we're doing in terms of relative performance, in terms of market share in each of those specific markets.

All right. Thank you. Good quarter.

Next question, please?

Our next question comes from the line of Ehud Geldblum from J.P. Morgan. Please proceed with your question.

Hi, thanks, Marcel and Scott. I have a quick clarification also and a more kind of global question. Marcel, when you gave guidance last quarter, you gave guidance for NetScreen to do 35-37 million in recognized revenue. You said at the time that excluding 65 million of revenue to purchase accounting. So it was kind of-- if purchase accounting didn't exist that would give you about 100 million.

Right.

Now what happened was you actually recognized 55 million and you didn't recognize 49, so it gets you back to the same 100 million, I want to make sure I'm understanding that.

Is the fact that you recognized more, is that just a-- I guess did the sales force do pretty much the same thing you had expected, so that there was no outperformance and it really was a difference in what you could recognize or was it really outperformance in the sales force and at the same time there was for some strange reason that 65 million was squashed into a 49. That took me a lot longer to say, so actually I have a global question for...

I think overall, the performance came close to what we had expected, so, again, I think that shows significant growth over the 93 million that was done the quarter before. So I feel really good about the number. I think the-- just to clarify the numbers, I mean we recognized 52 million in our own numbers and then the sum of the other numbers that were not recognized is 49 million and hopefully next quarter we can do without all this complication.

Right, so I'm not-- it's correct to say that-- first of all, why could you recognize more than you thought you could recognize?

Well, the difficulty in estimating what would be recognized and what would not be recognized is really the amount that would come out of the channel and really, you know, as the model works, as I explained the revenue recognition, we reserved everything that moved into the channel, so the focus is really on what is selling out of the channel and what is going into the end user, so the question is really the balance between what is shipping into the channel that gets deferred and the amount that comes out in sales out and in the case of this acquisition, you know, we had a lot of uncertainties in terms of what was the amount of deferred revenue that would be lost at the time of the close of the transaction.

Okay. That is definitely helpful.

Scott, if I could ask you a global question. On the Q4 conference call in February actually, you had referred to what you thought was seeing a budget flush. This was before you bought NetScreen, so this is on the Juniper side. And the content of the budget flush sounds as though Q4 is stronger than normal than you expect it to back off in Q1 and Q2, although it's not really a flush, more of a secular thing-- now that we know about Q1 and Q2, we've moved up from the strong Q4 to an even stronger Q1 to a still even stronger Q2.

Do you still think it was a budget flush in Q4 and how has your thought changed in terms of what really is going on?

You know, I think there is still-- again, it's a bit of a guess about Q4 because it's not really identifiable, but I do think there was some end of year activity that presented a strong Q4 and then there is, you know, there is-- the cause and effect here is a little hard to link, deterministically, but I think there are actually separate things going on, there was I think some of that kind of activity.

We guesstimated a 5-10% maybe budget flush consequence in Q4, and I still-- maybe it was at the low end of that given your characterization and, you know, some cause and effect behind just better business being done, but the other thing that happened, or is happening I think, unrelated is we just continue to see a very potent impact in the market that's driven by a growing consensus around the importance of a couple of things.

IP infrastructure and the need for that infrastructure to have all of these characteristics we've talked about and security, and those are separate observations around demand and importance of each and then a collective observation about the importance of the combination of the two.

So for me, that's another way of agreeing with you, that I think there is an underlying trend here that is stronger than just a Q4 phenomenon, because certainly both the information that translated into our results in the first half, as well as the outlook that we have going forward suggests that that is true.

I mean some of the guidance that we gave here as we look out to Q3 incorporates the reality that Q3 is usually seasonal and we still see growth.

So, you know, I do think there was a Q4 phenomenon, but I also think there is an underlying business here that continues to produce evidence that these are not-- you know, these are not moments in time or passing fads, and this is why I think some of the commentary we offered about the notion of the industry being shaped here, it really is, you know, an industry realignment and a redefinition that is every bit as fundamental as the evolution of microprocessors or relational databases or PC's in times past and there is clearly going to be the kind of company built here, whether it's us if we continue to execute, or somebody else, but there is going be a company built here of similar magnitude because these are fundamental redistributions of, you know, tens of billions of dollars along new definitions of success and failure and we feel like we're right in the middle of it.

Okay. Thanks.

You're welcome, thank you.

Operator, we'll take our last question now.

Our next question will come from the line of Tim Long from Banc of America Securities. Please proceed with your question.

Thank you. If I could just touch on the routing side here, you mentioned there was a release earlier in the month about British Telecom trial.

Could you just give us an update on what you expect the timing of that might be to turn into revenues and also if you could just give us a sense on the scale, could this be something that is as big as say a Deutsche Telekom, could this be a real major revenue driver over the next 6-18 months? Thank you.

Tim, I'm not sure actually what BT have said specifically about timeframes in all of this. Clearly, the trial activity will take place over the balance of '04, and I don't know if they have been specific about, about the rollout subsequently in precise terms of impact, but what I would say is this.

It has every bit the potential to be as significant as anything happening in frankly any network in the world. BTT's in Europe or networks in the U.S., because what BT are really doing with this concept of 21st century network is saying we're not going to, you know, incrementally implement our strategy here on this evolution to the next gen of networking. We're going to aggressively fundamentally do it. As we do that, we're certainly tackling-- the focus is voice on the front end, but they're talking about building a single unified infrastructure that is going to be responsible for the bulk of, let's say, if not the entire revenue of BT, as soon as possible.

And the line and company, there are very focused on what it's going to take to do that and there is an aggressive plan for it, and an assembly of partners that they are putting together in a trial process to confirm the viability and the logistics around rolling this thing out at scale and then from everything they have said, there is an intention to be very aggressive,e because they believe a rapid deployment is a strategic advantage.

It should also be said here, and this is, you know, important to know and we certainly know it, is that there is a lot of work yet to be done there and the trials that are under way, while they certainly indicate a set of chosen partners and in the role that we play there with our partner, Siemens, we are very committed to proving the points that we have to offer with our product, but there is also another cycle in all of this, which will be the actual selection process and the commercial follow-on to the conclusion of the trial once it's proven to be successful, and that is yet to be determined.

So there is a lot of work to be done. Obviously it's better to be in the trial than not, because that certainly makes a statement, but it doesn't make the statement.

And your view, similar type of competitive environments for this bid compared to the others, despite some noise from some of your competitors?

I'm not sure of the noise you referred, to but it's always out there and I'm sure there will be plenty of crying and complaining from anybody who is not involved, but, yeah, that's the kind of thing that's going on everywhere others are not involved and, and we're still in the same kind of competitive marketplace I think that we have been in for a few years here, which is a very short list of legitimate candidates for these things.

The only things that changing, I think, is it's getting increasingly clear and hopefully easier for customers to separate the distinctions in the choices that they have to make, and for us, that's a good thing.

Okay. Thank you.

You're welcome.

Operator, we'll now give you our final remarks. We would like to thank you for your participation today.

There will also be an audio replay available of this call in the Investor Relations section of our website at www.Juniper.net/company/investor/conference call.html. In addition, you can call 800-633-8284 and enter reservation number 21200631 through July 20th. The numbers are 800-633-8284 with reservation number 21200631.

We currently plan to report Q3 results on October 13th.

If you have any additional questions, please feel free to call the Investor Relations department.

Again, thank you for your participation on the call today. Have a nice evening.

Ladies and gentlemen, that does conclude the conference call for today. We thank you for your participation. And we ask that you please disconnect your lines.