Welcome to Broadcom Incorporated's third quarter fiscal year 2019 financial results conference call. At this time, for opening remarks and introductions, I would like to turn the call over to Beatrice Russotto, Director of Investor Relations of Broadcom Incorporated. Please go ahead, ma'am.
Thank you, operator, and good afternoon, everyone. Joining me today are Hock Tan, President and CEO, and Tom Krause, Chief Financial Officer of Broadcom. After the market closed today, Broadcom distributed a press release and financial tables describing our financial performance for the third quarter of fiscal year 2019. If you did not receive a copy, you may obtain the information from the Investors section of Broadcom's website at broadcom.com. This conference call is being webcast live and a recording will be available via telephone playback for one week. It will also be archived in the Investors section of our website at broadcom.com. During the prepared comments section of this call, Hock and Tom will be providing details of our third quarter fiscal year 2019 results, guidance for fiscal year 2019, and commentary regarding the business environment. We will take questions after the end of our prepared comments.
Please refer to our press release today and our recent filings with the SEC for information on the specific risk factors that could cause our actual results to differ materially from the forward-looking statements made on this call. In addition to US GAAP reporting, Broadcom reports certain financial measures on a non-GAAP basis. A reconciliation between GAAP and non-GAAP measures is included in the tables attached to today's press release. Comments made during today's call will primarily refer to our non-GAAP financial results. I'll now turn the call over to Hock.
Thank you, B. Good afternoon, everyone, and thank you for joining us today. Looking at the third quarter, consolidated net revenue was $5.5 billion, a 9% increase from a year ago. Semiconductor Solutions revenue was $4.4 billion, down 5% year-on-year and up 6% quarter-over-quarter. Networking continued to perform well, driven by strong demand for our merchant switching and routing platforms. As we also expected, shipments of custom silicon solutions in AI, SmartNICs, and video transcoding to cloud data centers were strong. Wireless is, of course, seeing the beginning of a typical seasonal uptake and the initial positive effects of increased content. These tailwinds were partially offset by weaker demand in storage and broadband. Revenue for Infrastructure Software was $1.1 billion. The CA business is running above our expectation, benefiting from sustained enterprise demand for our mainframe and distributed software.
However, SAN switching demand has paused as our partner OEM supply chain compressed in these uncertain conditions. Now, let me address the current environment and outlook. Enterprise and mainframe software customer demand continues to remain stable, particularly in North America and Western Europe. SAN switching demand will likely continue to be down another quarter while inventory in the OEM channels are being worked down. As it relates to semiconductors, although the U.S.-China trade conflict continues, we have not seen further deterioration in our business, both specific to China as well as globally. Accordingly, we continue to expect to achieve over $22.5 billion of revenue in fiscal 2019, including $17.5 billion from Semiconductor Solutions and $5 billion from Infrastructure Software. Looking into next year, Infrastructure Software is stable as renewals among our core customer base continue to be very solid. However, visibility continues to be very limited on the semiconductor side.
We are managing the business with an expectation that we will continue to operate in a very low growth, uncertain macro environment for the foreseeable future. Fortunately, the fundamentals of our Semiconductor Solutions business remain strong. As you know, our business is all about connectivity, from CPUs to memory in data centers, core to edge in networks, central office to client devices in distributed systems. Here we continue to benefit from the underlying trend in the IT world, an insatiable need for increasing bandwidth to connect things in data centers. Tomahawk switching has gone from 3.2 terabits per second just three years ago to 12.8 terabits per second today. In cloud computing, as the limits to Moore's Law constrains CPU and GPU performance, the pipes linking computing cycles to the network and storage expands. PCI Express Gen4 today at 16G replaces Gen 3 with the likes of next generation Rome CPU.
Legacy network interface controllers, NICs as they say, are becoming really intelligent, what we now call SmartNICs, and take on the task of accelerating workloads offload from non-optimized CPU within cloud computing. Even in SAN, that's storage area networks, Fibre Channel progresses its bandwidth at 32 gigabit per second in generation 6 today to generation 7 at 64 gigabit per second next year to reap the full benefit of all-flash arrays in enterprise storage. To truly connect computer storage, replacing direct attached copper, fiber optics running in 100G channels are moving to 400G as our hypercloud customers scale out data centers with Tomahawk 3 today. Turning to telco networks, call routing has gone from 1.6 terabit per second a few years ago to 9.6 terabits as represented by our Jericho2 router today.
In broadband, cable modem with DOCSIS 3.0 at 1 gigabit today will move to DOCSIS 4.0 at 10 gigabit over the next few years as cable operators need to compete against 5G networks. It is with DSL, digital subscriber line, where at a mere 500 megabit per second of data flow today, it will upshift to over 1 gigabit per second in G.fast, which may seem inadequate, so we need GPON at 2.5 gigabit. Even with that, we're poised today to launch into mass markets with 10G XPON. Of course, wireless connectivity too has seen the most headlines. In enterprise access gateways, the protocol has moved from 802.11ac to the new OFDMA enabled 802.11ax, otherwise called Wi-Fi 6. We are at the cusp of satellite communication migrating from 4G to next generation 5G in radio access networks and smartphones.
We are enabling these fundamental trends in the marketplace. This gives us confidence that we will continue to sustain and grow our Semiconductor Solutions business over the long term. Moving on, let's talk about software. First, with CA. As we mentioned, our model for CA is to focus on the 500 largest enterprises in the world, the biggest users of our Infrastructure Software. Based on experience through fiscal third quarter 2019, we expect our core customer business that's up for renewal in fiscal year 2019 to grow over 20%. Meanwhile, the attrition rate of business from the long tail of customers behind this core group is anticipated to be over 10% for fiscal 2019. We have another two plus years to turn over the CA customer contract.
Based on the trends of renewal growth from this core customer base in excess of the attrition of non-core business over the last nine months, we're confident that we can meet, if not exceed, the long-term revenue and profitability targets that we laid out for CA to you last year when we acquired that business. Our integration activity is largely complete with operating expenses to support CA approaching target levels. Finally, let me take a few more minutes to talk about our planned acquisition of the Symantec enterprise business announced in August. Acquiring Symantec furthers our efforts to build one of the world's leading infrastructure technology platforms. It is the logical next step in Broadcom's Infrastructure Software strategy and adds $160 billion cybersecurity market to Broadcom's addressable TAM. We will gain a portfolio of mission-critical security solutions that are deeply embedded among our core customers.
There will be meaningful cross-selling opportunities with Brocade and CA solutions. We believe this acquisition will enable Broadcom to gain a larger share of the wallet of these core customers. We expect this transaction to add more than $2 billion of sustainable run rate revenue with its leading franchises in cybersecurity. We also expect to achieve in excess of $1 billion in run rate cost synergies within 12 months post-close. Importantly, this transaction gives us the opportunity to achieve our ongoing financial objective of double-digit cash on cash returns. The integration planning process is well underway, and as you likely saw, we cleared HSR last week. We remain on track to close the transaction in the first quarter of fiscal 2020, subject to antitrust approvals in the European Union and Japan, as well as, of course, customary closing conditions.
To sum, our broad and increasingly diversified portfolio of leadership technology franchises has allowed us today to sustain revenue and increase cash flows, even in this challenging market environment. Now, let me turn this call over to Tom.
Thank you, Hock. Consolidating net revenue for the third quarter was $5.5 billion, a 9% increase from a year ago. The Semiconductor Solutions segment revenue was $4.4 billion and represented 79% of our total revenue this quarter. This was down 5% year-on-year on a comparable basis. Our Infrastructure Software segment revenue was $1.1 billion and represented 21% of revenue. Free cash flow was 42% of revenue, or $2.31 billion, and grew 8.5% year-over-year. Let me now provide additional detail on our financial performance. Operating expenses were $1.01 billion, driven by further reductions from CA-related activities. Operating income from continuing operations was $2.91 billion and represented 52.8% of net revenue. Adjusted EBITDA was $3.06 billion and represented 55.6% of net revenue. This figure excludes $141 million of depreciation.
In terms of working capital, a payables increase of $237 million was somewhat offset by receivables increase of $55 million and an inventory increase of $57 million from the prior quarter. I would also note that we accrued $110 million of restructuring and integration expenses and made $164 million of cash restructuring and integration payments in the quarter. Finally, we spent $112 million on capital expenditures. In the third quarter, we returned $2 billion to stockholders, consisting of $1.1 billion in the form of cash dividends and $977 million for the repurchase and elimination of 3.5 million AVGO shares. We ended the quarter with $5.5 billion of cash, $37.6 billion of total debt, 398 million outstanding shares, and had 442 million fully diluted shares for the quarter. Turning to our fiscal year 2019 guidance.
As Hock discussed, we are maintaining our full-year revenue guidance of $22.5 billion, including approximately $17.5 billion from Semiconductor Solutions and approximately $5 billion from Infrastructure Software. IP licensing is not expected to generate a material amount of revenue. On a non-GAAP basis, operating margins are expected to be approximately 52.5%. Net interest expense and other is expected to be approximately $1.3 billion. The tax rate is forecasted to be approximately 11%. Depreciation is expected to be approximately $600 million. CapEx is expected to be approximately $500 million. As a result, free cash flow is expected to be approximately $9 billion, which does take into account projected restructuring and integration charges of approximately $1.1 billion. Note, as of the end of the third quarter, $6.9 billion of free cash flow has been generated and includes $969 million of restructuring and integration charges.
Stock-based compensation expense is expected to be approximately $2.2 billion. Finally, we expect average diluted share count to be 440 million for Q4, and this excludes any impact from share buybacks and eliminations. Now on to capital allocation. As many of you know, Broadcom has a business model that generates a very healthy amount of free cash flow across an increasingly diverse and stable set of mainly infrastructure technology franchises. Over the last few years, we have worked to create a more transparent and balanced capital allocation policy. First and foremost, we have committed to return half of our free cash flow to shareholders each year in the form of cash dividends. In essence, this allows the AVGO stockholder to decide how best to reinvest 50% of the free cash flow that we generate.
In return, we have, in effect, asked the Broadcom stockholder to put trust in management to optimally reinvest the remainder of the free cash flow after the dividend is distributed. We think that we have a unique M&A strategy that allows us to consistently reinvest these excess cash flows that will drive returns well above our free cash flow yields. Over the past few years, we have developed a roadmap primarily around Infrastructure Software, starting with Brocade and then with CA and now Symantec, that will allow us to continue to execute on this strategy for many years to come. Going forward, our plan is to use this excess cash for acquisitions and/or to pay down debt that we borrow to make these acquisitions. Over the last year or so, we have bought back a lot of stock
We had an opportunity to buy stock at depressed prices following the CA announcement. We also wanted to limit the dilution from the one-time multi-year grant we did earlier this year. In all, we have invested $13.1 billion to repurchase or eliminate a total of 54.5 million shares at an average price of approximately $240 per share over the last 16 months through the end of our fiscal Q3 2019. In summary, we think this decision made a lot of sense. That being said, maintaining our core capital allocation strategy of dividends and M&A while pursuing meaningful buybacks in parallel has caused us to increase our leverage and leverage multiples pretty substantially. Especially in light of the weak macro environment we're seeing today, we are conscious of the risks that a more levered balance sheet creates and are very focused on managing those risks.
As a result, we have started to transition our focus to de-leveraging the balance sheet following the recent Symantec acquisition announcement. That concludes my prepared remarks. During the Q&A portion of today's call, please limit yourselves to one question each so we can accommodate as many analysts as possible. Operator, please open up the call for questions.
Yes, sir. As a reminder, to ask a question, you will need to press star one on your telephone. To withdraw your question, press the pound key. Please stand by while we compile the Q&A roster. Our first question comes from the line of Harlan Sur of J.P. Morgan. Your line is open.
Good afternoon. Thanks for taking my question. Good to see the business bottoming here in the second half of the year. Hock, we continue to hear that Tomahawk 3, Jericho2 are seeing strong demand for the 200 and 400 gig cloud networking adoption. Also, your revenue and design win pipeline on compute networking and security offload acceleration ASICs is pretty strong with guys like Google, Facebook, Microsoft, et cetera. Last earnings call, you had anticipated full year double-digit growth in your data center networking compute offload businesses. Question is, are you still tracking to that? Despite the muted growth outlook for the overall business looking into next year, given your design win pipeline, do you expect continued double-digit growth in the data center networking and compute offload looking into next year as well?
I guess the bottom line answer is our outlook, which we shared with you much earlier on, has not changed that materially at all. Yes, we do see continuing improve our ramp or ship deliveries into hypercloud guys on those various networking as well as computing offload silicon. That hasn't changed, and that has, in fact, given quite a lot of buffer to otherwise been a fairly uncertain and difficult market at this point.
Thank you.
Thank you. Our next question comes from Vivek Arya of Bank of America. Your line is open.
Thanks for taking my question. Hock, you started mentioning something about next year, and I just wanted to flesh that out. I know you're not giving next year guidance per se, but on the positive side, you're saying business is bottoming. You outlined a number of product cycles or so. Then you sound a little bit cautious on just the environment. I was curious how you're thinking conceptually about next year, pluses and minuses. Then as part of that, if you could also give us some indication of how Huawei figures into that, both in the near term, Q4 and next year. Thank you.
Well, you actually answered my question for me in many ways, which is you implying, which is correct, that the U.S.-China trade dispute is turning into an extended affair with lots of twists and turns and uncertainty. We're assuming things are now conditions, environment is not going to change from what we're seeing now. If we make that assumption next year, you probably see a very uncertain 2020. As we sit here right now, and we probably have at least another three months before we probably give you a much more clear 2020 guidance. As we sit here right now over the rest of this calendar year, if not fiscal year, what we're seeing is, you're right, it appears we've hit bottom, at least as it relates to the Semiconductor Solutions side.
We have hit bottom, we are looking ahead and seeing that we're staying right here, more or less, with little seasonality that pops up every now and then as we are seeing to some extent, probably in the next few months as we see a ramp up of our large North American OEM customer enhancers. Other than that, broadly, we're staying at the bottom.
Thank you. Our next question comes from John Pitzer of Credit Suisse. Your line is open.
Yeah, good afternoon, guys. Thanks for letting me ask a question. Hock, I realize that you're trying to stay away from giving too specific guidance by business line, but I'm just curious, relative to your comments on the wireless side of the semi business, how should we think about this build cycle for you this year, relative to past years? There is a view that you're gaining content on the RF side. You've got Wi-Fi 6. Your largest customer is also not staggering their phone launch this year. They've got some tariff issues that they might want to pull in some builds. As you look at the results for the July quarter, were they up sequentially, kind of in that mid-teens level? How do we think about October and how long is this build and what's the seasonal look into January?
If you give us any sort of color, that'd be helpful.
Well, it's always in this kind of, even in this kind of program, we have limited visibility as provided by our customers, how far we can go. To be very specific about what you are asking in wireless, as we always done in previous years, we see beginning of uptick seasonality in Q3, which is the July quarter, and we'll see more of it in our last quarter of the fiscal year, which is October quarter. We fully expect to see that. Frankly, that's as far as we see in this typical kind of outlook, because it obviously depend, at the end of the day, on resale of our OEM customers. We do see that this year. We do not see any major departure from that. Year-over-year, things are quite predictable in this respect.
Having said that, as you know, we are a company much more than just wireless today. Even in our Semiconductor Solutions, we have broad areas where we participate in, and there are multiple puts and takes, as I kind of indicated in terms of what happened in Q3. While we see continuing strength in networking, computing offload into data centers, we do see some weaknesses in storage and broadband. Things might probably look better next quarter for wireless, but maybe there are some other areas that will go the opposite way.
Broadly, given our large portfolio and broad diversity, and the fact that we are very fundamentally strong in each of those areas we are in, what we're seeing and what we're saying here is that in the semiconductor macro market, which is where we are pretty well represented on average, we are not as strong as it was same time last year. If you look at our data so far year to date, we're down probably on average about 8% year-on-year, or thereabout. Which is, in my view, what we think the market is on a broad basis, excluding memory, of course. The fact that we were year-on-year down more in Q1, Q2 this fiscal year, and probably less down in the back half of fiscal 2019 should not be taken as a fact that there's a bottom and there's possibly a recovery.
That's why we make the statement. We're pretty confident we are at the bottom. The question is, there is not much clarity or visibility yet or certainty that any sharp recovery is around the corner.
Thank you. Our next question comes from Christopher Danely of Citigroup. Your line is open.
Hey, thanks, guys. Hey, Hock, can you talk about the expected timing of the 5G ramp, especially for your ASIC and ASSP business? When do you think that starts? Has this been pushed out at all? Is it about as good as you thought it would be three or six months ago, or has the forecast changed, and how much?
Well, that's a tough question in terms of that. To be honest with you, we do not participate that much on 5G RAN on radio access in a broad perspective. We touch at the handset level, we touch somewhat at the radio access, we believe we probably touch most at the backhaul, the networks. I think each of them will ramp 5G at different times. Our ability to see how big that ramp is at any particular point in time is also dependent on the fact that different operators in different countries will probably want to ramp up at different points in time, different parts of their network, whether it's the front side ramp or the backhaul.
My guess is probably you won't see much of it until later this year, early part of next year, as far as we are concerned, which is more of the backhaul, where it affects our numbers.
Thank you. Our next question comes from Stacy Rasgon of Bernstein Research. Your line is open.
Hi, guys. Thanks for taking my question. I wanted to know, you've had a growth target for semis of mid-single digits, and obviously we're in an uncertain environment, but you also talked a lot about, I guess, company specific drivers and product cycles that then fundamentally can still drive. I guess, are those fundamental drivers enough to keep the semi business growing at least somewhere in the ballpark of that sort of mid-single digit long-term guidance, even in an environment that's uncertain? I guess maybe even put a different way, could you maybe talk about specifically some of the product cycles and trajectories that are maybe unique to Broadcom that you see kind of ramping next year and maybe compare with what we had this year, across your different businesses?
Well, yeah, when I went on this quantification of the various trends driving the fundamentals of our product lines and our semiconductor business, I really wasn't thinking necessarily, and if I give the wrong impression, I do apologize, one year. This is an ongoing trend. I was using examples because these are examples ongoing that we see today, recent past, and going forward in the future. What I do see is that, yeah, in all these areas we participated, from be it as networking in data centers, routing in core and metro networks, or in cable and broadband, where we all are, one thing is clear, which is the nice thing we have, we see in technology, especially the semiconductor technology. There is a constant evolution of the products we do. There's constant demand for improved performance or, as I put it, increased bandwidth, since we do connectivity.
Depending on the end and the particular applications, some may spend a year, some may spend three years to make a transition, but they do happen, and they continue to happen, and that's what keeps demand sustained for our technology, for our products. That's what underpins, as we put it, a long-term growth forecast in semiconductors for us of mid-single digits. I don't mean a year. I don't mean necessarily even two or three years. I mean over the next five years, 10 years. Because think about it. In 2018 fiscal, our semiconductor business organically, take out any acquisitions, grew 12% year-over-year from 2017. In 2017, we grew over 2016, about 10% organically.
I frankly do not expect that high rate, double-digit rate in this business, where we are pretty well represented across a broad area, to be able to continue at that rate of growth. We see in 2019, a decline from a strong 2018, as I said, of probably mid to high single digits. Not unexpected. Taken over a period of multiple years, we believe we will achieve that mid-single digit compounded annual growth rate. It won't do that every year.
Thank you. Our next question comes from Ross Seymore of Deutsche Bank. Your line is open.
Thanks. Tom, I had one for you on the capital return side of things. In the past, that 50% dividend policy you have, it's very clear, but the free cash flow that generates that at times has been a little more subjective due to one-time charges, finishing the classic Broadcom campus, et cetera. How do restructuring charges fit into the math of free cash flow? Then the investment grade target, where you want to keep it there, what sort of leverage target should we think about with you guys more aggressively paying down debt as opposed to repurchasing shares in the near future?
I think on the cash flow side, Ross, what we've tried to do is create somewhat of a formula, because we do typically have acquisitions, and we do have investments we make to get to the synergy targets. In effect, what we're trying to do is deliver the benefits to the stockholder the year with which we are able to achieve those synergies. What we've done is we've said, look, let's look at the free cash flow from operations, which obviously includes those restructuring charges, and let's add back those for the purposes of calculating the dividend. When we get to the end of the year, we'll take the cash flow from operations. We'll add back the restructuring integration charges, and in effect, we'll divide by two to get to the number that we use to calculate the dividend.
This year, we don't have any specific one-time large campus initiatives or things like that that we would also look to back out. That should be a fairly good proxy for how we think about the dividend and the recommendation of the board at the end of this year. In terms of investment grade, I think we've talked about this before, but maintaining investment grade similar to the dividend policy is a core principle here, and I think the reason for that, and Hock and I have discussed it many times internally, is it provides maximum flexibility for us to continue to pursue our strategy, that's always been very important. Over time, we had this one-time event. We bought back a bunch of stock, as I talked about in the prepared remarks. That's pushed leverage up more so than we normally would.
We think based on the economics and the results so far, that was the right decision. We're really looking forward to going back to the playbook we pursued in the past, which is following the distribution of the dividend. We still have a lot of cash flow, and that cash flow, we think based on our strategy, is best used for M&A because that's where the returns are most optimal. As part of that, we often borrow money to finance those acquisitions, and then we go and pay that debt down. I think we're going back to a more traditional playbook. As part of that, given the size and the scale and the increasing diversity of the business and the sheer profitability and the cash flows, we're very much an investment-grade company, and we're going to be very focused on maintaining that investment-grade status going forward.
Thank you. Our next question comes from Blayne Curtis of Barclays. Your line is open.
Hey, guys. Thanks for taking my question. Tom, for you, just on the annual guidance for op margin would suggest a tick down into October. Just want to make sure how literally to take that, and if so, if you could talk about any drivers as to why profitability will be coming down in October? Thanks.
Yeah, it's a good point. Maybe we're a bit conservative, to be honest, Blayne. I think the reality is you see the mix shift in the second half of the year. Got a couple things going on. One, with wireless ramping seasonally, as Hock discussed, those margins are not as high as the rest of the portfolio. Then on top of that, we're giving back some on the SAN switching side in the back half, which is fairly high margin product. We felt it made sense given the mix to stick to that number, and that's why we're staying there.
Gotcha, thanks.
Thank you. Our next question comes from Toshiya Hari of Goldman Sachs. Your line is open.
Yeah. Hi, guys. Thanks very much for taking the question. Hock, I was hoping you could talk a little bit more about what you're seeing in the hyperscale cloud environment. Quite a few of your peers have talked about signs of a recovery into the back half and potentially into 2020 following a pretty extended period of weakness. I appreciate you guys have drivers that are very specific to the company, but in terms of customer sentiment on spending and capacity expansion, what are your thoughts there? Thank you.
Okay. Well, it's a very interesting question, and I guess I'd put it two ways. One is touching on cloud spending. We pretty much see some of what a lot of people out there are seeing in comments lately, which is an improvement in cloud spending in multiple areas, especially in scaling out new generation data centers as they expand their business. We obviously benefit from that in our switching, routing, and some of our specific custom offload programs in AI or just SmartNICs. We see that. As we all know, these days, cloud spending is turning to be a bit like telco spending, very lumpy. What we see on the other half of our business in infrastructure is with enterprises, end users.
Here, from the viewpoint of an opportunity to see end users directly, we see enterprises or end users, who are the end users, especially the largest bunch, our core customers, their spending continues. Their IT budget is pretty strong, and their spending continues because they need that digital transformation as their business continues fairly decently. That hasn't impacted, which is why our Infrastructure Software business renewals continue to grow very nicely. That's not an issue. It's when we sell components to OEMs who then make it into systems to sell to these same end users that we see a difference. The difference is simply that I think, as far as we could gather, given the uncertainty in this environment, that everybody is constraining, all our partners are constraining their supply chain and compressing their supply chain, so to speak.
That action of compressing is really a reduction in orders, a reduction in what appears to be demand. It's really interesting as we see this, that on a software side, things are direct sales to large enterprises continue fairly unabated. When it comes to selling components through partners to these same end users, we see a different perspective, a slowdown, a constraint, and it's reflected, obviously, in the numbers we are showing here. It also is reflected in the outlook that we're giving to you guys.
Thank you. Our next question comes from Craig Hettenbach of Morgan Stanley. Your line is open.
Yes. Thank you. Hock, switching gears to CA and the software side, can you talk about just some of the efforts around portfolio license agreements, specifically what kind of sales efforts you have in place there and what this could mean for the business over the next couple of years as you try to gear it that way?
Yes. Well, it is. That portfolio licensing arrangements as a sales model or business model we apply to our Infrastructure Software business, which is CA at this point, has been, in my mind, and to be fair, it's less than a year since we launched it. Since we took over CA only last November. It's been, I would call it very successful. It's been very encouraging. A big part of why we think it is encouraging and why it works is because we just focus on the largest five, 600 customers of CA.
These are guys where we have decent size and large footprints to begin with, and as we go towards renewals, our ability to offer a portfolio-wide licensing arrangement, unlimited capacity on the full range of broad products that we have that extend from mainframe to distributed software, is extremely attractive and cost-effective to those large enterprises who buys a lot of those Infrastructure Software. We have done multiple deals as we went through multiple renewals, and we continue to be very engaged with those companies. I guess my conclusion is by focusing on these largest customers where we have large footprint that exists and where the ability to consume more of those software we have, it makes a lot of sense.
As I put in some of my comments earlier on, this ability to focus on these largest guys and put the sales motion technically and commercially, very effectively on these large customers and possibly not focus on the long tail of some non-core customers is a key part of the whole model and what makes the whole model work. Conceptually, we are renewing those contracts with larger footprints on portfolio licensing agreements, as I mentioned, at a rate of over 20% annually, while we are, probably because of lack of focus and reducing or attriting smaller non-core customers at 10%, and those core customers represent 80% of our overall revenue, while the long tail represent only 20%, adds up in a situation after one year, almost one year, where we show a net gain of over 10%.
Get that going, three years later, which is the average term of every contract we put in place, we would see a clear step up in our revenues from this business of, we like to believe, double-digit growth.
Thank you. Ladies and gentlemen, this concludes today's conference call. Thank you for participating. You may now disconnect.