AGB 2020.10 - Broadcom (AVGO)
AGB 2020.10 - Broadcom (AVGO)
Hock Tan is an Outsider CEO
Subscribe to AGB - One analysis of a good business every two weeks.
Broadcom/Avago
To limit confusion, we’ll call the company Avago and only refer to Broadcom as the company that Avago acquired in 2016. Avago acquired Broadcom and then took on the name.
Avago is the fourth largest semiconductor company by revenues after Intel, SK Hynix and Qualcomm. The company was formed in 2005 when the semiconductor division of Agilent was sold to private equity firms, KKR and Silver Lake Partners. Shortly after in early 2006, Hock Tan was brought in as CEO. The company went public in 2009 and at the time was much smaller in size and scope as compared to today.
Avago has grown mainly through a series of acquisitions starting in late 2014 with the acquisition of LSI. Subsequent to that acquisition, Avago has made one major acquisition per year, integrating these new companies with speed and efficiency. The second most impactful acquisition was Broadcom in 2016, which at the time was actually 30% larger than Avago. And so, Avago took on the Broadcom name and new Broadcom was formed. Other meaningful acquisitions were Brocade in 2017, CA Technologies in 2018 and Symantec Security in 2019.
The company started off with semiconductor franchises within wireless, industrial and wired applications. After the LSI acquisition, the company also had a meaningful presence within storage. In 2018, Avago acquired CA Technologies, a software company that services mainframe customers. This shocked the market and the sell-side because running an enterprise software company is much different from running a semiconductor business. The integration and synergy capture for that deal were great and Avago went on to acquire another software company (Symantec) the next year.
The company now has leading semiconductor franchises in switching/routing, set-top boxes and modems, optical components, ASICs, RF filters, Wi-Fi and bluetooth receivers, HDD storage controllers, industrial fiber optics, etc. Avago also has leading software franchises in end-point and web proxy security, data loss prevention, mainframe application software, and fiber channel networks.
What’s impressive about the company is that the management team has been able to efficiently use capital to grow in a short period of time. Since Hock Tan became CEO of Avago in 2006, the company has made 10 meaningful acquisitions, totaling over $47B paid in cash and stock. Revenues have increased from $1.5B in FY06 to $22.6B in FY19 at a CAGR of 23%.
In that time frame, the margin profile of the business has improved dramatically as well, going from 26% gross margins and -6% operating margins in FY06 to 55% gross margins and 15% operating margins in FY19 (it would be higher at 25% if not for acquisition related expenses).
So what’s the catch and why is the stock trading at a below market multiple? It’s mainly because organic growth is not part of the story. The business model of the company is to have sustainable semiconductor and software franchises, not to invest for high growth emerging franchises. There are other semiconductor companies with end market exposure in high growth areas like artificial intelligence, 5G wireless, IoT, or that have significant share gain opportunities ahead of them (e.g. AMD and TSM). But Avago is generally exposed to steady, low growth areas other than 5G wireless through its RF franchise. The company’s target revenue growth rate is 3% and its adjusted EBITDA margin target is 55%, which has improved materially since acquiring CA and Symantec.
Why is it a good business?
Similar to other large semiconductor companies, Avago benefits from scale economies. Fabrication of semiconductors typically requires high levels of fixed costs and leverage is mainly achieved on the gross margin line. Even though most semiconductor companies have adopted the fabless model (outsourcing production to foundries such as TSMC, Samsung, Global Foundries, UMC, etc.), the leverage in the model still applies. Avago mostly runs a fabless model but the company does have few of their own foundries as well.
This is partly the reason that many semiconductor mergers have been successful. According to Goldman Sachs, synergies ranged between 5%-20% of revenues for semiconductor mergers from 2014-2016 with an average savings of 11%. And that’s just counting the announced synergies at the time of acquisition. Many mergers have resulted in better than announced synergies and margin accretion.
Couple that with an operator that is conscious about acquiring under-optimized businesses with the intention of serious cost cuts and margin increases and you have a good business. What’s unique about Avago’s acquisition strategy is that the company is willing to make the acquisition targets smaller (through divestitures) if it means better returns on capital deployed. Most companies and CEOs want most forms of revenues because it’s frankly really hard to generate. But sometimes the cost of attaining that revenue could mean a lower margin profile for the company. And ultimately, management teams are attracted to revenue because, as we know, executive compensation for larger companies are skewed higher than that of smaller companies.
As we mentioned earlier, the two most impactful acquisitions to date are LSI and Broadcom because of the size of the targets’ revenues relative to Avago at the time. When acquiring LSI, Avago’s revenues were about on par with LSI but the company’s operating margins were far superior. Avago absorbed a 16% EBIT margin company in LSI and somehow increased the combined company’s operating margin to 41% the next year from 32% the year prior.
Similarly, Avago acquired Broadcom whose revenues were 30% larger than Avago’s was at the time. The company then took a 26% EBIT margin company in Broadcom and improved the overall company’s operating margin to over 44% the next year, back to where it was prior to the acquisition.
If any of you have read William Thorndike’s book, “The Outsiders: Eight Unconventional CEOs and Their Radically Rational Blueprint for Success”, you may start to recognize that Hock Tan has some of the characteristics of an “outsider CEO”. These are the reasons why we make this claim:
He is an efficient capital allocator, weighing the return profiles of acquisition, buybacks, dividends and debt repayments
He quickly and efficiently identifies which technologies/business lines are core (reinvests capital) vs. financial assets (milks cash flows)
He doesn’t shy away from making bold moves, often times going after fish bigger than him
He is ruthless when it comes to efficiently running his businesses
These advantages have allowed Avago to increase FCF from under $1B in FY14 to over $9B in FY19. That’s an 11x increase in 5 years, or 65% CAGR. Furthermore the FCF margin has increased from the 20%s to over 40% in FY19.
Returns on capital?
Over the past 10 years, Avago has spent most of its capital on acquisitions and R&D. The company has spent 60% of its capital on acquisitions, 31% on R&D and 8% on capex. Notable acquisitions are:
LSI (2013) – Enterprise storage and networking semiconductors
Broadcom (2015) – Wireless and wired communications semiconductors
Brocade (2016) – Fiber channel storage software & hardware
CA Technologies (2018) – Mainframe and enterprise management software
Symantec Enterprise Security (2019) – Endpoint, web proxy and data loss prevention enterprise software
Analyzing the large deals really exemplifies how the high returns are made on these acquisitions. Broadcom was acquired for $37.3B ($17B in cash and $20B in shares). Projected synergies was $750M within 18 months of closing. To put that into context, Broadcom was doing $2.1B in operating income and was spending almost $2B in R&D and $600M in SG&A annually at the time. $750M is 36% of operating income and 29% of SG&A + R&D. Aside from the cost cutting efforts, Avago also pruned the product portfolio and divested assets worth $830M post acquisition. Netting cash of $6.9B acquired at close, Broadcom was purchased for 3.5x revenues or $10.4x EBITDA.
The CA acquisition may be an even better case study of a high return on capital acquisition. CA was purchased for $18.8B + $2.3B in assumed debt for a total of $21.1B. The company divested assets worth $950M when it sold Veracode to Thoma Bravo. CA also came with $2.75B in cash, which resulted in a net purchase price of $17.4B. Prior to the acquisition CA was doing $4.3B in revenue of which $2.2 was Mainframe, $1.8B Enterprise and $300M Services. EBITDA was $1B or 23% margin. Post the acquisition, Avago outsourced its services division and shutdown $500M of low-end enterprise business, resulting in just $3.5B in revenue. However, EBITDA increased by $1.3B to $2.3B or 66% margin. After netting all this out, CA was purchased for 5x revenues or 7.6x EBITDA.
A similar situation unfolded with the Symantec Enterprise Security acquisition. Since it was an asset purchase there was no excess cash or debt involved. The purchase price was $10.7B. The company was doing $2.7B in revenue and $350M in EBITDA (13% margin). R&D expense was 24% of revenue or $650M annually. Post the deal close the 12 month synergy target was over $1B, and with revenue optimization of $700M, EBITDA jumped to $1.3B or 65% margin. Symantec was effectively purchased for 5.4x revenues or 8.2x EBITDA.
With all this cost cutting and running a lean operation, Avago has a reputation for being ruthless cost cutters at the expense of innovation. The company argues that it’s misunderstood and that they are just better allocators of R&D capital. One area that they point to is the lead in ASIC and wireless FBAR technology, but they’ve had that lead since the early days of Avago.
We estimate that over the past 5 years, Avago has generated between 20%-35% return on its capital deployed as a company. When we make this calculation, we usually only include the cash on cash return but given that the company issued 50% of the shares outstanding to acquire Broadcom and $1B of convertible bonds/preferred stock to acquire LSI, we included the stock values into the return calculation. The company wide returns are still high but we see them dip into the high-teens/low 20%s after the Broadcom acquisition given that the multiple paid was 10.4x EBITDA, which implies a ~10% return on capital.
We’d also like to point out that debt has juiced returns over the past 10 years. In FY13, the company had no debt, but starting with the LSI acquisition, the company started to utilize debt to make acquisitions. The company ended FY19 with over $33B in debt, but has maintained its debt/EBITDA ratio steady around 2x-4x. As of F20Q3, the company has over $44B in debt.
After thinking it through for many days, we’ve come to the conclusion that calculating returns on stock issued for acquisitions is a difficult exercise. Just as a caveat, we typically don’t analyze returns on capital on a per share basis but rather on a company basis. And issuing stock for an acquisition is not that different from a company issuing large amounts of stock based compensation every year. Typically we just try to avoid companies like that, unless the growth and return profile are that compelling like with Workday and Roku.
Reinvestment potential?
Avago doesn’t have enough opportunities to deploy for all of its free cash flow. The company has a policy of paying 50% of the prior year’s free cash flow in the form of dividends. The other 50% is used for acquisitions and share buybacks/debt payments. The bulk of the remaining free cash flow will be used for acquisitions when opportunities arise.
The company has stated that it is looking to acquire technology companies with established and sustainable end markets. This means that the runway for revenues is generally low to medium growth and the company has the ability to maintain its market leadership. Typically an IP portfolio comes with the acquisition, which directly impacts the company’s ability to sustain its market share. Another important attribute is that the asset is undermanaged or under-optimized. And the valuation has to be low enough that the cash returns are attractive.
The company’s last three acquisitions in software (Brocade, CA and Symantec) were all mismanaged prior to being acquired. The problem was that the core businesses were stable but not showing much growth and these companies tried to diversify to achieve growth. The returns on capital to achieve that growth were very low or even negative. Avago is fine with the low growth profile of these core businesses and instead cut investment that went into low-returning growth areas.
We screened for potential semi-conductor technology companies that fit Avago’s acquisition criteria. The parameters we used were:
Revenues greater than $5B
Negative to 20% revenue growth
Lower than 20x forward P/E or 20x trailing LTM EV/EBITDA
SG&A + R&D margin > 35%
This only resulted in a handful of potential semiconductor targets, none of which were that compelling. Avago has stated before that the company doesn’t want to be in memory and microprocessors and while we’re not sure about semi-cap companies, it doesn’t leave many options to choose from within our screen. It’s no wonder that the company has stated that there aren’t many opportunities within semiconductors, given where valuations are and where we are in the semi-cycle.
We also ran a similar screen for software companies and there was more promise there. The parameters we used were:
Revenues greater than $2B
Negative to 20% revenue growth
Lower than 20x forward P/E or 20x trailing LTM EV/EBITDA
SG&A + R&D margin > 50%
This also resulted in a few companies, but a few were interesting acquisition targets. For example, NortonLifeLock could be interesting if Avago would be interested in increasing consumer exposure within security. Furthermore, Citrix Systems seems like an interesting enterprise software/networking play that is also grossly mismanaged.
We estimate that the company has reinvested between 70%-110% of its capital over the past 5 years, with many years above that range due to the large acquisition of Broadcom. With the high reinvestment rate, we estimate that the growth in intrinsic value has ranged between 20%-30% for the company. We will point out again that this was only achievable because the company put on lots of leverage with low cost debt since 2014.
What else is important?
Failed Qualcomm acquisition
Something to consider with respect to the growth engine of the company is regulatory approval in the U.S. and abroad. As Avago gets larger, the acquisition targets also need to get larger to actually move the needle. When thinking about how regulatory approval could impact future acquisitions of large targets, we can look back at the failed Qualcomm acquisition from 2017-2018.
On November 2, 2017 Avago made an offer to acquire Qualcomm for $70/share in the form of $60/share cash and $10/share stock. The 28% premium was large, given Qualcomm’s disputes with Apple and Huawei within its licensing division and its loss of modem market share to Intel. Qualcomm swiftly rejected the offer and never really negotiated in good faith but Avago increased the offer to $82/share on February 5, 2018. If you want to see a masterclass presentation to showcase how poorly a target company has performed, look no further than Avago’s presentation.
Qualcomm had underperformed the semiconductor industry by 12% annually from FY14-FY17 while putting out long-term growth targets that were 4%-6% greater. Sure there were new wireless protocols coming down the pike like 5G cellular, mmWave and Wi-Fi 6, but Qualcomm’s prospects were looking bleak during this time. Luckily for Qualcomm (and its shareholders) the U.S. government stepped in. CFIUS blocked the deal citing 5G as important to national and economic security. Avago had filed to move its headquarters from Singapore to California, but it didn’t matter to the U.S. government.
Acquisitions need to get larger
Avago has actually already run into this problem. Ever since the failed acquisition of Qualcomm, the company has pivoted away from acquiring semiconductor companies. Just looking back at the basic semiconductor screen that we ran, it doesn’t seem like there’s a viable target for Avago to acquire that would move the needle, be complementary with their core product franchises, and that is trading at a low enough valuation.
So the company pivots into software. We believe that future acquisitions will most likely be within software or services given the switching costs that exist with software companies and the relatively low valuations for some of the larger more established software companies.
Wireless business is non-core
On the company’s FY19Q4 earnings call, Avago’s CEO said that the wireless business was not a core asset but rather a financial asset. Of course the market took that as a negative, implying that Avago thought that there was potential of losing market share in RF as handset OEMs move from 4G to 5G. The company did actually lose some share during the 3G to 4G transition, so the concern was warranted. But Avago’s FBAR technology is very valuable for the high-end smartphone, even more so with 5G, because of all the different frequencies that the smartphone needs to filter. On the following earnings call, Avago’s CEO retracted his statement and said that wireless was now a core franchise, given Apple’s decision to re-up its contract with Avago for another three years.
The market may look at the wireless business as a melting ice cube, given the competitive pressures and Apple’s well-known desire to eventually put out their own wireless chip. RF will certainly be another area in which Apple may seek to develop internally or go with a lower-cost competitor at some point.
We view Broadcom’s ability to cut bait and its willingness to sell a franchise that it sees as having the potential to lose its sustainability as a positive. This goes back to Hock Tan thinking like an outsider CEO. He is continually weighing the potential for each product line and franchise based on future returns on capital.
Optionality
Meaningful Acquisition
We’ve discussed this in detail in the “Reinvestment potential?” section.
If you made it this far, I hope you received some value from reading our analysis. Please subscribe to the free newsletter (if you haven’t already) and share with anyone that would find it valuable. Thank you for your support!