Description

Note: Net debt, and therefore TEV, shown above in financial information are on a proportionate basis. P/FCF is based on FCF after estimates for minority payments.

Introduction  

Millicom is an attractive long investment at current prices.  The company has a long runway to grow its cable (and “converged”) business in underpenetrated emerging markets.  The opportunity exists because mobile businesses and emerging markets are generally out of favor with investors, sentiment is mixed and the company has a volatile past.  However, new executive leadership has been communicating and executing on a clear strategy to improve the business and its cash generation.  Investors can earn mid-teens to low-twenties IRRs for the foreseeable future, with further upside on potential acquisitions, timely divestitures of non-core assets or via a sale or merger scenario.

Business @ 30k Feet

Millicom is a mobile and fixed communications provider in Central and South America (together LatAm) and Africa that offers residential and business broadband, video, voice and related services.  This write-up focuses almost entirely on LatAm because the region 1) accounts for ~80% of proportionate revenue, 2) is management’s investment focus and 3) should drive returns for shareholders over the next several years.  Conversely, in Africa management is targeting cash flow break-even, limiting investment, “getting a lot of inbound calls” and says that decision making will be driven by whether or not a sale yields a better return for shareholders than continuing to operate the assets (explicit allocator speak for “we will sell at the right price”).  My model includes Africa for now, but in my opinion a divestiture of these assets is likely.  The various competitive and regulatory dynamics, product mixes and ownership levels across Millicom’s footprint should be considered in evaluating an investment; but the opportunity is interesting even when viewed from a broader perspective.  For this reason, my commentary addresses the most important investment factors with what I view as an appropriate level of detail, while avoiding lengthy detours to address the intricacies of each and every market.

As an introduction to the business, the geographic LTM proportionate revenue mix is 80%/20% between LatAm and Africa.  However, the LTM proportionate EBITDA contribution before corporate expenses skews even more towards LatAm at 85%/15%.  Central and South America are 37% and 44% of LTM proportionate revenue, respectively.  LatAm countries listed by proportionate revenue contribution are Colombia 19% (50% ownership), Guatemala 15% (55% ownership), Paraguay 14%, Bolivia 12%, El Salvador 9%, Honduras 9% (67% ownership) and Costa Rica 3%.  

Management does not provide proportionate information for product revenue, so the product mix detail that follows is all on a consolidated basis.  Mobile is still the largest contributor today at 64% of LTM consolidated revenue, followed by cable at 25%, other at 8% (includes handsets) and mobile financial services at 2%.  Mobile’s contribution is down dramatically from 81% at the end of Q2 2014 to 63% at the end of Q4 2016 due to the merger of Millicom’s Columbian mobile operations with scaled Columbian cable operator UNE in the second half of 2014, the tuck-in purchase of TV Cable Parana in Paraguay in early 2016 and most importantly organic growth in cable homes passed, RGUs and ARPUs (on a local-currency basis) as management prioritizes fixed networks.  Several divestitures in Africa have also impacted mix.  In the same time, consolidated mobile data revenue has gone from 18% of total mobile service revenue to 35% (i.e. voice and SMS are down from 82% to 65% of mobile service revenue).  UNE was first consolidated for a full quarter in Q4 2014, and since then cable’s share of total revenue has increased ~40 bps per quarter on average from 25% to 28% as of Q4 2016 (Africa is all mobile, so cable’s contribution is slightly higher in LatAm than for the consolidated entity).  As I discuss in more detail below, and as many telecom followers will appreciate, this shift improves the quality of Millicom’s cash flows.

Historical Context / Current Management

Millicom gets credit for recognizing the inevitability of mobile early in its evolution – Swedish investor Jan Stenbeck of Kinnevik (Millicom’s largest shareholder today) bid on and won several mobile licenses in the U.S. as far back as 1990.  However, the company’s history appears to be anything but a glide path.  In 1994, the company set up Celcaribe to cover the Colombian-Caribbean region and its market cap reached $1.25 billion.  By 1996 Millicom had interests in 30 operations in 20 countries around the world.  By 1998 the company’s rapid expansion may have stretched it thin – it appears to have been forced to sell a few assets for the first time in its history.  Shares proceeded to go sideways for a couple of years as the geographic expansion resumed.  After the dotcom boom/bust the market cap traded down to tens of millions of dollars (ouch!), and by 2003 the company had slimmed down to 15 markets.  Over the next couple of years Millicom rebounded as mobile penetration rates increased dramatically across its markets – the market cap grew to over $12 billion.  The company announced its first cable acquisition, the $510m purchase of Central American operator Amnet, in July 2008.  Deal timing proved to be unfortunate as the GFC hit.  From the announcement to February 2009, the company lost ~$6 billion of market cap (~60% of its value).  Since the crisis, there have been several more questionable managerial decisions (like investments in Rocket Internet affiliated entities) and more volatility in the stock.  Fast-forward to today and Millicom shares have compounded at just above 4% vs. nearly 7% for the S&P in the 24 years since the company’s 1993 IPO.  You may be confused at this point…“wait I thought this was a long pitch?” It is.

So why bring this history up?  Cable is an attractive business because it provides must-have high-speed data services on a subscription basis, many times in markets with favorable competitive dynamics.  Furthermore, since cable companies frequently generate strong and stable cash flows that can be borrowed against, managers often have access to significant amounts of capital that can be reinvested, used for acquisitions or returned to shareholders (not news to many investors).  However, managers need to balance financial leverage with a keen awareness of business trends and the market environment.  In light of these dynamics, my view is that recognizing strategic assets is not enough to earn superior returns in the telecom sector.  Investors also need to identify managers with a clear view of the future and a solid grasp of ROI and capital allocation.  Millicom’s history made it difficult (impossible?) to check the ROI and capital allocation box until recently.  What changed?  The UNE merger won the company some points; and then, in March 2015, Millicom hired long-time Liberty Global executive and John Malone lieutenant Mauricio Ramos to be its CEO.  Ramos’ experience in the region dates back to Liberty Global predecessor entity UnitedGlobalCom, where he was VP of Business Development for the Latin American Division from March 2000 to June 2005.  He then became CEO of VTR, Liberty’s Chilean business, from 2006 to 2011 before being named President of Liberty Global’s Latin American Division (Chile and Puerto Rico) through the beginning of 2015.  The historical (and current) results of the Chilean business, where he was involved for a decade and a half, are impressive (not worth discussing in this write up, but happy to share that mini case study if readers are interested).  As a bonus, in a region where government relationships are probably more important than elsewhere, his Colombian heritage and local business experience should also be an asset vs. the previous string of European CEOs.

Company Strategic Direction

Ramos’ resume is what you would look for if your goal was to grow a cable business in LatAm – a combination of organic expansion and M&A experience for a John Malone controlled company in the region.  He has said and done the “right things” since taking over as Millicom’s CEO, his strategy is becoming evident in results and he is positioning the company better for the future.  Several interviews and Millicom’s transcripts since he assumed his current post explain his perspective and strategy, but I’ll outline the key points here.  

First, Ramos is a believer in driving data consumption and building “a significant fixed cable footprint under [Millicom’s] mobile footprint in LatAm” to support growth.  He has said that the two key issues communication providers are dealing with are the explosion in data traffic and the convergence of fixed and mobile networks.  In an interview in January 2016 he pointed out that the best way to deliver ubiquitous connectivity at high capacities to meet the needs of consumers is through fixed networks due to their substantially lower costs per bit and flexibility to add capacity relative to mobile networks.  Malone’s comments at LMCA’s 2016 AGM are the best (and most entertaining) summary I have seen on the mobile/fixed market situation Ramos was referring to:

“…and frankly, if using our terrestrial network, we can provide a wireless component that is more cost-effective than the cellular guys because we wouldn't have that massive spectrum investment, right? is a very interesting proposition.  And I think it's one of the reasons you see the big wireless companies outside of the U.S. attempting to acquire terrestrial networks.  So if they believe that there was no future to terrestrial network, then why would Vodafone be acquiring? So they see the limits of their wireless.  As they've -- some of those guys have told me, because I say, ‘Guys, your capital structure is terrible.  You don't use much leverage. You pay a lot of taxes.’ And they say, ‘Yes, but John, don't forget about every 5 years, government makes us buy our business back, the spectrum market, very expensive.’  And those are almost calculated backwards. In other words, how much can we squeeze out of the cellular company by making them compete to stay in business, right?  So you have to look at what we might bring to bear, which is essentially cheap spectrum”

In summary on the first strategic focus, Millicom has a long runway to invest in fixed footprint expansion at attractive rates of return, which is discussed in more detail below in the “Investment Thesis” section.  Millicom is a growth story that will look more and more like a cable business over the next five to ten years.

Second, Ramos is focused on operational efficiencies, which for him come from the business mix transition towards fixed, cost optimization and a better focus on converting topline into cash flow.  Trailing LTM EBITDA margins have improved nearly 300 bps since he joined the company, from 33% in Q1 2015 to 36% in Q4 2016. In Q2 2016 management announced initiatives to eliminate >$200 million in run-rate operating, capex and working capital costs (amounts to ~3% of revenue), with a big focus on IT savings and to a lesser extent efficiencies in sales and marketing.  The results suggest that Ramos is taking care of the simple blocking and tackling better than Millicom’s previous management teams.  Over time, his knowledge and relationships should improve areas like equipment procurement as well.

Lastly, Liberty surely ingrained a focus on total shareholder returns in Ramos.  This mindset manifests itself in several ways.  First, the points mentioned above (prioritizing investment in higher ROI fixed projects over mobile and driving FCF through operational discipline) show a focus on economic profits rather than growth.  Second, Ramos is clearly willing to part with non-core assets at fair prices.  His communication in this department has been consistent, and corporate events since he joined the company should give investors confidence in the decision making process on divestitures.  Millicom exchanged stakes in country-level Helios Towers Africa (“HTA”) assets for a 24% ownership stake in the parent in late 2015, saying that this made the investment more liquid.  Its ownership has been diluted to 22% since but was on the books at $189m at 12/31/16, or about SEK 18 per share.  On Millicom’s Q4 2016 call, management disclosed that a process has been launched to sell this investment.  Also on the Q4 call, management disclosed that it is divesting its mobile business in Senegal for $129 million, or a 6.3x multiple!  For reference, the company trades slightly above 5x on a proportionate basis, and its valuation includes a cable component that deserves a higher multiple.  In early 2016 Millicom announced the sale of its Democratic Republic of Congo mobile business for $160 million.  DRC’s transaction multiple was not disclosed, but Africa capex was greater than EBITDA at the time and back of the envelope math suggests that the price was more than fair.  Third, compensation packages for the senior management team are being linked to total shareholder returns.  All executives’ shareholding requirements relative to base salary are set to increase over the next couple years.  Ramos himself was granted shares at SEK 640 (~30% above today’s price) at the time he joined, and has made open market purchases since assuming his position as CEO.

Why the Opportunity Exists

Quite simply, the company is unloved.  Millicom “checks the box” for a number of easy excuses investors may have for avoiding a company.  Mobile is a challenged business – check.  I’m not touching EM – check.  Accounting is complicated – check.  Ownership levels vary across a hodge podge of assets – check.  It’s a controlled company – check.  It has leverage – check.  

Additionally, questionable decisions in the past resulted in a historically volatile ride for investors; and the primary listing is in Sweden where Kinnevik is based, which may limit exposure.  Potential investors may be unaware of the changes taking place at the company, management’s clear strategy and its growing track record of execution.  Investors could be hung up on legacy voice and SMS declines; but with the business shifting to cable, each passing quarter offers investors a better business mix.  Sell-side sentiment isn’t great – 40% of analysts have a “Hold” rating or lower.  Lastly, maybe investors have taken enough pain in the region with LiLAC (I certainly did).

Everyone is trying to figure out quad play / converged offerings in the developed world, so finding an unloved emerging markets player positioned to capitalize on convergence in underpenetrated and growing markets trading at a discount seems like a formula for a successful investment.

Investment Thesis

Several key factors support my investment thesis.  Some were introduced above (e.g. the quality of the leadership team) so the points below focus almost entirely on highlights that are incremental to what has already been discussed.

1. Organic Growth Opportunity.  The population across Millicom’s LatAm markets is north of 100 million, and the number of households is between 25 and 30 million.  Colombia is the biggest country with a population approaching 50 million and management says there are 13 to 14 million homes.  According to the International Telecommunications Union (“ITU”), home broadband penetration (which by their definition includes lower speed DSL) across Millicom’s footprint is ~25%.  Penetration for pay TV is higher at around 35% but still has room to grow as Millicom drives bundling to maximize asset turnover on its infrastructure.  I am more focused on and excited about the broadband opportunity given the nearly 100% incremental margins it carries.

    

   When Ramos joined Millicom the company passed 7.2 million homes, 5.8 million of which were hybrid-fiber-coaxial (“HFC”) homes.  Since he took the reins, Millicom has updated legacy infrastructure and accelerated its cable footprint expansion.  Today it passes 8.1 million homes, 7.2 million of which are HFC.  On the Q2 2015 earnings call, management communicated a goal to pass 10 million homes, a 40% footprint expansion of its network at that time (shares were at SEK 658 that day).   A year later on the Q2 2016 call, after Ramos had spent more time as CEO, management increased its expectation to 12 million homes.  Shares were at SEK 456 (close to today), or 30% lower than a year prior despite a 20% increase in its long-term home passing target!  Management could conceivably increase this number again in the future but for the medium term feels that one million new passings per year is reasonable given labor availability, permitting and other non-financial constraints.  Regardless, going from 8.1 million to 12 million homes in four years is a 50% increase.  Whether management stops here or decides to pass more homes, Millicom has a strong organic growth opportunity to deliver more fixed services in LatAm.  However, growth for growth’s sake can be dangerous, which brings me to the next point…

2. Unit Economics.  Ramos has walked through the economics of new builds over the course of a couple of earnings calls.  I’ll cite the relevant bits below:

• “The cost of a build in Latin America, we've often said, is about $100 per home passed. So that's a sliver of the cost in a developed market. And the reason for that is threefold….One is the density and these are multi-dwelling units in highly populated urban areas that we're building….The second element is simply the fact of the plant being aerial versus underground…a fraction of the cost of building an underground plant….And simply, lastly, the cost of labor…that is relatively cheaper in our markets….very manageable CapEx spend within our cash flow profile….” – Q3 2016

• “And of course, there is CapEx and installation cost, some of which will recoup when you do install a customer. And that typically is in our markets, if I could just give you a ballpark number, around $150. But again, that's not spent on a home passed. It's spent on a connected customer that is giving you revenue.” – Q3 2015

• “On filling the network, we mean converting homes passed into homes connected or subscribers. Our current penetration for the entire company about 35%. So think of that as a long-term industry average or a target. In the new builds, we're achieving 19% penetration within just 10 months of the new homes being released for sales.” – Q3 2015

• “As we build the homes 18 months out, we already carry penetrations in those homes that are about 25%” – Q1 2016

• “Our ARPU per household is around, I could be off by a dollar or 2 here, so I'm speaking out of memory, but $25-ish [currently $27], and I think we've shown these numbers before.” – Q3 2016

Presumably the $150 in installation costs mentioned above includes customer premise equipment capex and capitalized labor associated with new connects, but some of this total is probably captured in opex as well.  Worrying about these details doesn’t change things too much, so I just calculate the returns including 100% of installation costs for reference (this approach is conservative).  Assuming 40% EBITDA margins, D&A at 20% of revenue and a 30% tax rate, unlevered cash on cash returns are approximately 16%, 20% and 25% at ten months (19% penetration), 18 months (25% penetration) and maturity (35% penetration), respectively.  Lastly, before Millicom started accelerating its build, it was actually 40% penetrated in its footprint, which provides upside to the 35% penetration assumption I made for the business at maturity.  Obviously, net customer additions on fixed plant carry very high incremental margins so any improvement in penetration would be very beneficial to returns.  

3. Improving Quality of Earnings / FCF Focus.  Mobile voice and SMS in LatAm will continue to be a headwind – these products declined 15% in 2016.  However, their contribution was down to 33% of consolidated revenue during 2016 from nearly 50% in 2014.  Cable and mobile data combined for ~50% of consolidated revenue in 2016 and grew 7.4% and 22.7%, respectively.  With each quarter that passes the “growth products” are a bigger piece of the pie and the “legacy products” are a smaller portion, which makes year-over-year comparisons easier and a return to growth imminent – management is targeting low-single digit topline growth in 2017.  More importantly, this growth is coming with higher EBITDA and cash flow margins due to the more favorable competitive and customer dynamics of fixed networks (fewer competitors overall and customers cannot simply switch out a SIM card for home broadband).  The margin performance is already evident in results, which means we are observing rather than predicting this trend.  Management thinks that EBITDA can growth at a mid-to-high-single digit rate in 2017 and EBITDA less capex can grow ~10%.  The trade of mobile voice and SMS revenues for mobile data and fixed cable services is a good one for the company and investors.

4. Monetization of Non-core Assets.  As discussed above, Millicom has been divesting its African portfolio at attractive valuations.  Additional exits at EV / EBITDA multiples at or above the current trading multiple of ~5.3x (like the Senegal sale) would be very favorable to shareholders.  The company is looking to sell its stake in the African tower business HTA, which it carries on the books at $189 million.  Millicom also holds about $140 million in book value of other miscellaneous investments (online marketplaces, telco carrier services, etc.) that it has accumulated over the years.  The HTA stake and the other investments are a small portion of value relative to the African assets but are still > SEK 25 per share at book value.

What are We Playing For?

Unlike LiLAC post-closing of its CWC acquisition, Millicom generates FCF today and has a much less demanding valuation.  On a proportionate basis it has traded between ~4.5x-5.5x EBITDA for most of the past year.  Proportionate revenue should grow 4%-5% annually from 2016 levels.  By reporting region, I think Central America can grow mid-single digits, and South America will likely be mid-single digit “plus” given the runway Millicom has talked about in Colombia and Bolivia.  I hold Africa flat minus $100 million for the Senegal business that was just sold (revenue is a ~11% reduction from 2016 levels due to the divestiture).  This assumption could prove to be conservative, but I don’t have a strong view on Africa.  I give the company 200 bps of proportionate EBITDA margin improvement based on cost saving plans, but think that there is upside potential on profitability given the high incremental margins on fixed broadband growth, which are an increasing contributor to earnings.  These assumptions get you to slightly above $2 billion in proportionate EBITDA in 2020E.  If you give Millicom a blended forward EBITDA multiple of 5.5x, at the high end of its recent range but still below comps like LiLAC (>7x), you get a proportionate EV of ~$11.1 billion by the end of 2019.  With new home builds, dividend payments, minority payments, etc. I assume proportionate net debt expands slightly in dollar terms to ~$3.5 billion, which is actually a 0.2x reduction in proportionate net leverage given underlying EBITDA growth.  The last piece is ~$330 million in book value of non-core holdings.  This analysis implies a ~SEK 700 stock in two to three years’ time, plus ~SEK 70 in interim dividends gets you to a high-teens compounder.  The company seems like it will stick with the dividend rather than shrinking shares so I don’t consider any change in capital allocation strategy.  I think that the 5.5x multiple is conservative but am hesitant to give credit for meaningful multiple expansion in a base case valuation.

Pre-Mortem / Risks

• Key Man Risk:  Mauricio Ramos leaves…

• Competition:  Claro (America Movil) and Telefonica cut off their noses to spite their faces.  Both have large debt loads that probably preclude them from being too aggressive for too long.  Also, both have other larger markets to worry about.

• Controlling Shareholder:  Kinnevik proves to be too stubborn and inflexible.  Its main goal is likely to protect its dividends, which could limit strategic decisions and balance sheet flexibility to accelerate growth or M&A.

• Regulatory / Political: The countries Millicom operates in general have more unstable legal and political systems and institutions than mature markets.  The company had a bribery incident in Guatemala and has been forced to shut down portions of its mobile network adjacent to prisons in El Salvador as the government combats crime.  Needless to say, these types of issues are generally not a concern for investors in developed markets.

• Country-Specific Risk:  Parts of Central and South America are highly dependent on tourism and commodities.  Changes in these markets / industries can have a negative impact on FX, which affects Millicom since much of its programming, capex and debt are in USD / Swedish Krona.

I do not hold a position with the issuer such as employment, directorship or consultancy.

I and/or others I advise hold a material investment in the issuer's securities.

Catalyst

• Stronger than Expected Results:  I don’t play for quarters, but I think that Ramos has the experience to manage expectations and deliver against plans.  As more of the mix shifts to high incremental margin fixed broadband, and legacy voice and SMS contribute less of revenue and profits, the Street could be underestimating growth and margins.  My consolidated EBITDA is 10%-15% ahead of the Street in 2019E.  If this trajectory holds then Millicom will have “beats” along the way.

• Market Perception:  Clear, consistent communication from management and more stable results could shift sell-side sentiment and garner greater appreciation from investors.  Millicom is lining up to be a “good and getting better story.”  A change in perception could result in a re-rating given Millicom’s low valuation multiples (at the same time as earnings begin to improve).

• M&A / Consolidation:  I didn’t discuss M&A above, because I don’t think investors need deals to make money given the organic runway to expand the company’s HFC plant.  However, Millicom has several opportunities to be an acquirer including ETB in Bogota, Colombia, where the government built a state of the art FTTH network that passes 600-700 thousand homes but is looking to sell to fund local social programs (there is also some legacy telco in addition to the fiber homes).  Claro is “too dominant” in Columbia for the government to approve a deal according to a former ETB executive, and Telefonica is not really in a financial position to get a deal done.  The former ETB executive that I spoke to said that he thinks that the government is selling too early since it hasn’t yet filled the network with customers.  These dynamics make Millicom the lone strategic buyer and put it in a strong negotiating position for a good asset.  There is also the potential for broader market consolidation / more transformative deals – yes, I am referring to the combination with LiLAC.  Millicom and LiLAC have essentially no overlap in their footprints.

• Non-core Divestitures:  Favorable terms on exits from non-core assets in Africa would also benefit the company and improve investor perception.