S&P Global Ratings believes that in an environment of low growth and low share prices for telecommunications companies worldwide, many face the seemingly discordant challenges of:

• Financing large capital expenditure needs, including accelerated fiber rollouts, overlapping with gradually rising 5G investments (see chart 1);
• Reducing leverage (see chart 1); and,
• Lifting equity prices out of the doldrums for shareholders (see chart 2).

As we described last year, some operators are selling tower infrastructure to help meet these challenges (see "Credit FAQ: Why Telecom Companies Across Europe Are Selling Their Towers," published on Oct. 11, 2019). We expect companies that have fixed-line infrastructure, and fiber in particular, will increasingly join this trend.

In Europe, where fiber coverage was already a priority for the European Commission and domestic governments, we think COVID-19-driven broadband demand will only accelerate investment. Converged operators like Altice and Telecom Italia have sold parts of their fixed-line infrastructure, and we expect more operators to follow.

Stakeholders on both sides of asset sales stand to gain. Telcos can gain balance sheet flexibility from the sale proceeds if the valuation multiple is high enough. It can also introduce new owners to share the cost of additional investment or take it off balance sheet altogether. This last point can be especially compelling for telcos early in the fiber investment cycle or those looking to extend coverage to less dense areas that have high connection costs per household.

From a shareholder perspective, infrastructure sales can benefit from the higher valuation of telco assets and give a welcome boost to telecom share prices. Premiums have ranged from a 20% higher EBITDA multiple for fiber assets with more investment requirements and commercial risk, to 3x higher for more mature tower assets.

While these dynamics are motivating telecoms to sell, loose monetary policy and historically low interest rates are providing ample funding for buyers. This includes private equity and infrastructure developers seeking investments with growth potential and characteristics that could support long-term stable cash flow.

Here, we discuss the key rating drivers of wholesale fiber companies in the first section, and then, in the second section, the credit impact for telecoms that spin off fixed infrastructure assets.

Chart 1

Chart 2

Analytical Approach for Fiber Wholesale Companies

1. How does rating a fiber infrastructure firm compare to rating a traditional utility? Does it have a stronger business profile than an integrated telco, and how important is regulation?

Fiber telecom infrastructure covers a broad range of assets. We think the most relevant comparison to utilities is with regional or national wholesale telecom networks like we find in Europe or parts of Asia-Pacific, rather than local or retail carriers in the U.S. But even among this subset, fiber networks come in various stages of development and with differing competitive landscapes.

For traditional utilities, like natural gas or electricity distribution networks, a common feature that helps address heterogenous assets is a supportive regulatory framework. Regulation leads us to view some traditional utilities as having stronger business risk profiles (that is, excellent) than most telecom operators, and higher debt tolerance for a given rating (that is, via our low or medial volatility tables, see "Corporate Methodology," published July, 1, 2019).

Table 1

However, comparable regulation is currently absent for fiber companies. This is a key difference and one reason why we don't view utilities as an appropriate peer group. When fiber assets are being spun off from integrated operators and were key drivers of a telecom's competitive position, we think the spun-off fiber company's business profile can be within a similar range. Until regulation becomes a more salient factor, the other key considerations will likely include the network's penetration rate, the level of competition, the maturity of the asset, as well as the existing regulations in place.

The reason regulation is so important is that, as with other types of infrastructure, it can play a critical role in whether a fiber company can achieve the following objectives:

Recover costs and earn an economic return with a high degree of certainty.

Maintain a business model shielded from competition and disruption.

In the telecom sector, we often think of regulation as a challenge to these objectives, particularly for a company with a dominant position and market power. Anti-monopolistic regulation typically seeks to promote a level playing field and limits excessive pricing to protect consumers and encourage competition. But in the case of utilities, if a monopoly is accepted as a more efficient infrastructure model than competing, duplicative networks, regulation can be a defining supportive factor in our credit analysis.

Indeed, our assessment of a utility's business risk focuses on the concept of "regulatory advantage," rather than a company's competitive advantage in the market. We base our assessment on how a utility's credit quality and its ability to recover its costs and earn a timely return are protected by:

• Regulatory stability,
• Tariff-setting procedures and design,
• Financial stability, and
• Regulatory independence and insulation.

Instead of a framework to punctually recover all costs and earn a return, the examples of wholesale fiber regulation to date largely consist of setting maximum rates, or co-investment rules. Such regulation doesn't provide a utility-like basis for the full recovery of new and expansion fiber, or maintenance capital costs. Fiber companies are also exposed to material market risk in terms of take up, particularly in cases where there is competing infrastructure.

Chorus is one of the closest examples we have to a regulated fiber monopoly, and we assess it using the same approach that we apply to telecom companies. It is the dominant fixed-line telecommunications wholesale provider in New Zealand, and we consider its business risk profile as strong, and use our standard volatility table for its financial analysis. While we don't currently view Chorus as a utility-like company given that it remains exposed to competition from alternative technologies as well as other fiber companies, we will continue to monitor evolving industry characteristics. In our opinion, New Zealand's proposed regulatory framework is credit positive and should provide Chorus with better transparency and predictability in future price setting for its fiber network and improve its revenue predictability and stability. We recently expanded our leverage thresholds at the 'BBB' level to recognize improved cash flow visibility and will continue to review these thresholds, both upward or downward, as its track record of cash generation is further established. We may employ a similar approach within our telecoms framework for other examples of more mature telecom infrastructure assets.

2. Can fiber companies with a dominant market position be compared with utilities for ratings purposes, even in the absence of a strong regulatory framework?

Absent regulatory support, there are also examples of utilities whose credit quality derives from a strong market position that we consider a natural monopoly. In these situations, we assess the "regulatory advantage" of a utility by considering the stability of its monopoly market structure and track record of full recovery without a regulatory framework for rates, including tariff flexibility to counter market risks.

In theory, we can envision parallels for strong wholesale fiber assets, those that are already built out and well invested, with high penetration rates and low competition. Such infrastructure can be stable, cash generative, and difficult to economically compete against through overbuilding.

But in practice, many fiber wholesalers, particularly in dense markets, are not monopolies at all and face significant market competition from cable, copper, or even other fiber providers. This raises another fundamental problem for consideration of utilities as the appropriate peer group, since we define utilities as having little or no practicable substitute.

In the U.S., most fiber companies are local, non-wholesale players, directly competing for retail or enterprise customers. In large Tier I markets they typically face pricing pressure with at least two competing independent providers, the LEC (local exchange carrier), and cable (either Comcast or Charter). Even Tier II and Tier III cities typically have no less than three players in the market. As a result, we assign a fair or weak business risk profile to most fiber companies in the U.S.

In Europe, most urban markets are similarly competitive. But even in rural regions with less competition, fiber lacks a track record of stable cost recovery and returns. We also question whether the fibercos have rate flexibility to deal with unexpected operating or investment costs if they have set offtake terms with retail telcos.

In addition, there is potential for competitive disruption over time from alternate forms of data transmission. With traditional utilities, the physical need to deliver their commodities is difficult, if not impossible to economically disrupt. But data can already be provided through fixed-line, mobile, and satellite alternatives to fiber today. 5G fixed-wireless access, in particular, has received a lot of attention. While we think the millimeter-wave mobile frequencies needed to truly compete with fiber performance face coverage and reliability challenges, it is nonetheless an example of a potential and evolving technology risk.

Putting technology disruption concerns aside for the moment, we think the natural monopoly argument for fiber may be more compelling in rural environments than urban markets. These non-dense areas would not be economic for multiplayer infrastructure competition because the higher cost to connect each household requires stronger penetration rates to earn a return, acting as a significant disincentive to overbuilding.

But even being the only game in town (or more realistically, the countryside) doesn't ensure cost recovery. Fiber companies will still face commercial take-up risk: Do customers need broadband like they need electricity and water? And even if they do, do they really need fiber, or is copper-based DSL technology sufficient? To be clear, we think the long-term trend tilts to fiber broadband, and this is being accelerated during the COVID-19 pandemic by consumer demand for high speeds to support increased virtualization of work, school, and entertainment. But penetration rates and timing in several years' time are difficult to predict, making timely recovery uncertain. As chart 3 shows, household penetration of fiber compared to traditional utilities in Spain, a market with relatively well-developed fiber, is not yet comparable as an essential service.

Chart 3

3. Could S&P Global Ratings change its view that regulatory support, substitution, and penetration risk for wholesale fiber infrastructure is weaker than for traditional utilities?

Possibly, but only if regulation develops into a more comprehensive framework (see question 1 above) and if fiber assets can address commercial risks by demonstrating both their essential nature to customers and a lack of viable substitutes (see question 2 above).

More supportive regulation may result from the COVID-19 pandemic, which has enhanced the strategic profile of telecom infrastructure in the eyes of many stakeholders. As governments have urged reduced mobility to combat the spread of the virus, virtualization has enabled parts of the economy to avoid a more severe downturn. Broadband has kept numerous businesses running and remote workers employed and has expanded the online sales channel for many brick-and-mortar businesses.

Noticeably absent from the beneficiaries of this virtualization are the telecom operators themselves. With some exceptions like in the U.S., where pay TV cord cutting has been offset by upgraded broadband packages, fixed broadband providers have generally been unable to effectively monetize the spike in data flow caused by increased virtualization, despite providing the enabling infrastructure. Given the societal benefits, government and regulators may reprioritize advanced telecom networks as a more strategic objective and therefore be willing to pursue frameworks that encourage investment. Regulation that gives better visibility and confidence in an economic return could support such goals.

In terms of commercial risk, with a more mature fiber asset, investment risk moderates and recovery of capital begins to recede as a rating concern. If the asset operates in a clear, circumscribed service market with high penetration, low competition, and effective economic barriers to entry, revenue visibility improves. This doesn't eliminate the possibility of customers downgrading to cheaper copper services, but we think the risk of churning away from fiber is more manageable than the risk of migrating to fiber in the first place, especially if copper phaseouts begin. A degree of asset expansion risk may even be tolerable if mitigated by supportive pre-commitments by customers. Although technology disruption risk from alternative services may remain, when the supportive characteristics we describe are reinforced by a track record of stability, they may be sufficient to consider a natural monopoly argument until (and if) substitution risks become more acute.

Italy is a market that may move toward a single fiber provider if OpenFiber and Telecom Italia combine their fiber assets, and it could be a test case for the regulatory framework in Europe about how far it goes toward supporting investment and cost recovery. But to be clear, this is not the case today. Before changing our approach, we would likely need a track record of stable take-up and revenue visibility, and a regulatory regime that ultimately allows fiber wholesale to recover costs and earn a timely return. This could realistically take several years. As with the Chorus example in question 1 above, we are more likely to start reflecting improved wholesale fiber prospects via a stronger business profile or greater financial ratio tolerance within the telecoms sector framework rather than switching to a regulated utility approach.

4. Are there other asset classes that make for a good comparison with wholesale fiber companies?

In some ways, we see a useful comparison with midstream energy infrastructure, typical of North American master limited partnerships (MLPs), and with telecom tower and satellite companies. Among these asset classes, a strong contractual framework, coupled with significant scale, has supported high business risk profiles.

In the case of MLPs, companies like Enbridge, TC Energy, Colonial, and Kinder have excellent business profiles because of their diversity and their scale as USD10-15B EBITDA companies. A key attribute of such pipelines is their highly predictable operating cash flow. Large portions of their cash flow are protected against uncertain economic conditions by contract structures that typically include long-term take-or-pay terms or have fee-based activities with low volume risk. We also evaluate their duration and counterparty risk. And similar to utilities, we consider the support of regulation that allows for tariff increases to offset increased operating and capital expenses. For example, some Canadian midstream assets have a cost-of-service mechanism in pricing. In the U.S., pricing is more similar to what we may initially expect of European fiber companies: the regulatory body (FERC) sets max rates, and the companies then negotiate commercial contracts.

Table 2

By contrast, many of today's wholesale fiber agreements with integrated incumbents carry volumetric risk. As the fiber wholesaler's customer (the fiber retail operator) gains or loses retail subscribers, its number of lines leased can fluctuate, and with it, payment to the fiber wholesaler. This lack of revenue certainty, common among integrated European telcos that also provide wholesale services, weakens the comparison with the stronger midstream companies. The adoption of take-or-pay like commitments would help to address this, though we question whether retail telcos are willing to accept this transfer of take-up risk. Pipelines typically sidestep this risk by fully contracting their capacity before a final investment decision is taken and construction begins.

Well-positioned pipelines also bear little substitution risk from overbuilding, and other modes of transport (trains or trucks) are not economically competitive. An existing long-haul pipeline (that is, natural gas, oil, or refined products) has a significant advantage over new builds because of environmental and regulatory headwinds and the expense of construction. A new pipeline is also likely to be less competitive once put into service than a fully depreciated pipeline since its rates will need to recoup the high construction cost. As a result, there is little economic incentive to develop directly competing substitutes unless justified by material excess volume demand, which can be the case for pipelines that get to a hub or close to a demand center or export market.

In the more highly distributed telecom market, direct as well as disruptive competitors have a broader set of entry points to challenge. Examples are disruptive fixed-wireless or satellite competitors or a local fiber or cable competitors in a selected, perhaps underserved markets. They can also start small and scale up over time, which is more difficult for competitors in the midstream sector. Even DSL technologies can be upgraded to compete with fiber for less demanding customers.

If fiber operators develop stronger contracted revenue structures, they could also bear parallels to telecom tower companies. The strongest towercos, though still evaluated as telecom companies and with a standard volatility table, can merit excellent business risk profiles and support higher leverage than typical telecom operators. This includes those with scale, long-term offtake contracts, and a stable industrial model of increasing efficiency by growing tenancies per tower.

Table 3

But this efficiency point is somewhat unique to towers. Compared with that of fibercos, towerco efficiency is less reliant on penetration of end customers. Towerco customers (mobile retail operators) pay for and need service area coverage on day one regardless of their retail customer penetration, and the towerco faces little marginal cost as the retail base grows. By contrast, fiberco customers (fiber retail operators) don't provide coverage by service area; they have fixed-address retail customers and want to procure wholesale service line by line. This makes penetration rates critical to a fiberco. On top of that, towercos can further increase utilization by rationalizing duplicative towers, and when their telco customers densify their networks. Densification can result from greater network utilization as retail customer penetration increases, much like for fibercos. But unlike fibercos, it can also come from higher utilization per retail customer, peak load profiles driven by events and traffic patterns, and as new technology (5G) is rolled out.

Many fiber companies are also earlier in the asset cycle and still have significant capital investments ahead of them. This is because they are either greenfield or are still in their build phase. In contrast, most tower companies are at the mature end of their investment cycle, with a large asset base. While many engage in M&A-fueled growth, these also tend to be mature assets with low capex requirements and high revenue visibility. To the extent they build out new infrastructure, it is relatively small incremental investment compared to their asset base. And when capex does rise significantly above maintenance levels, as our forecast increase in capex intensity shows (see chart 4 below), the reason is typically that the towerco is undertaking build-to-suit projects for telco clients. Such contracted structures can eliminate or reduce the risk of recovery and an economic return on the investment. By contrast, the absence of such structures to cover the comparatively large investment cost of fibercos would leave them facing market risk, which weakens their relative business profile.

Chart 4

We also see potential parallels to the satellite segment. Long lead times of about three years, launch risk, and all-or-nothing network investment decisions created high barriers to entry, and we assessed many players with a strong business risk profile. But with aggressive forecasts for demand growth, and technology improvements that significantly increased throughput, those barriers weakened, and overcapacity built up in the latter half of the past decade. Competition to satellite services has also come from disruptive alternatives, including fiber itself, resulting in an erosion of business risk profiles to satisfactory in 2018. Depending on the market, we think fiber wholesalers could potentially face similar competition and substitution risks beyond the next three to five years.

5. Could a project finance approach be used to rate greenfield fiber companies?

Possibly, if its key features are consistent with our project finance approach. This is relevant because the construction phase, with its high investment costs and minimal initial revenue, may have depressed or even negative debt to EBITDA and free operating cash flow to debt. A project financing structure, including appropriate terms and conditions, funding certainty, and a strong risk transfer to experienced contractors may help offset weak leverage and cash flow until construction is completed and the asset is operating.

However, we identify significant remaining hurdles for new fiber buildouts to qualify as in-scope for project finance. Most notable are commercial take-up risk (see question 2 above), and the long-term horizon that typically characterizes project finance debt. The longer the tenor of the debt, the more acute the technology substitution risk for fiber. This is significant given the already existing threat of alternatives and is only likely to grow with technology advances, such as 5G fixed wireless access. The evolution of technology to transmit data and customer preferences is extremely hard to predict, especially considering that the prevalent technology only 15 years ago was ADSL and 2G on feature phones. These risks to revenue certainty weaken the prospects for predictable debt service coverage ratios, acting as a drag on the operations phase assessment in the absence of adequate risk allocation and credit worthy counterparties to provide sufficient mitigation.

As with the corporate approaches discussed in question 4 above, a strong contractual framework that supports debt service and repayment over the project period can address operational phase risks. If so, another key consideration is whether the fiber wholesale issuer's purpose and scope is limited to a defined project. This may be challenging given the more corporate-like profile of wholesale fixed-line operators we have observed to date. The risk attributes of such a model, where the fiber company continues to invest in expansion and upgrades to evolve with technology and the competitive environment, are unlikely to fit with our project finance approach.

Analytical Approach For Telecom Companies That Sell Their Fixed Networks

6. What is the ratings impact on integrated telcos that sell their fixed network? Can the remaining retailco/serviceco still have a strong business profile?

When an integrated telecom operator sells its fiber assets, we consider the impact on its business profile. Our only rated example to date is Telcom New Zealand Ltd. (an integrated telecom), which was rated 'A' with a strong business risk profile. It was renamed Spark after splitting off its fixed network, which was named Chorus. The split led to a downgrade of Spark and a reduction of its business risk profile to satisfactory.

We think this is instructive for other operators that sell their networks: business risk profiles are likely to fall. Unless mitigated by retention of other "differentiating" assets, telcos without networks are unlikely to have strong business profiles. Compared to towers sales, we view the impact of selling fiber as having a larger impact. The key consideration is how "differentiating" the asset is.

We view the prospects for mobile differentiation in the relatively mature 4G market to be weaker than for fixed network differentiation where larger gaps in network quality still remain, particularly when it comes to fiber. We therefore see a more benign starting point for the impact of tower sales to a telecom's business profile. Furthermore, operators typically retain ownership of the active RAN (radio access network) equipment, the mobile core network, and spectrum, three key remaining elements to establish service quality.

For fixed networks, we consider what assets are sold and what is retained. The resulting impact on a telco's competitive position depends on whether the sold assets were unique and how extensive they are. For example, we view an incumbent selling off all its fixed network assets as likely to have a more significant negative impact to its business profile than a challenger telco selling off its network, or when the network is overbuilt by competitors and not a unique, differentiating asset. We also see a more modest impact if an incumbent is selling off either its backbone or its distribution network, but not both. In the latter example, if parts are being sold instead of the entire asset, we typically view the last-mile network as more important than the backbone for assessing the competitive position. This is because of its greater monetization and differentiation capability as the "customer facing" part of the network.

We also consider the extent of the sell-off. If only minority portions are sold and control is retained, we typically take a sliding-scale approach. For example, as ownership falls toward 51% but control is retained, the impact on business profile is only incrementally negative.

When a sale goes beyond 50% and the telco loses control, we view the business profile as encountering a potential cliff. In contrast to minority sales, there is a step change upon losing control. Table 1 below details our assessment of the impact on components of our business risk profile from an incumbent telecom selling off its fixed network. While there are some puts and takes, the overall impact is materially negative due to the loss of differentiation. This is typically a greater risk for operators with an incumbent network advantage, and less so for challengers with subscale networks still requiring significant investment.

Table 4

7. Can improvements to a company's financial profile offset the negative business consequences of losing control over its fixed network?

Telecoms are exploring asset sales in part because of their high valuation multiples, which can materially improve financial flexibility, thereby helping to offset the resulting erosion of business profiles. Greenfield, or early-stage fiber assets, are unlikely to see multiples high enough to produce significant deleveraging but are also not likely to seriously damage the telecom's business profile since the network isn't yet developed. At the mature end, fully developed assets like Altice International's Portugal fiber was able to garner a 20x multiple, similar to the high end of tower asset multiples.

Table 5

When there is a large difference in valuation multiples between the integrated telecom and the asset being sold, the potential financial profile benefit grows. This is because of the greater scope for deleveraging--either from sale proceeds and debt paydown or from a transfer of debt to the network company. in our view, at least a turn lower in leverage is likely needed on average to maintain ratings, depending on the extent of business profile erosion as discussed above.

8. What adjustments are important to keep in mind when evaluating the financial impact of fiber sales on the telco?

After the amount of sale proceeds and leverage at the fiberco, the stage of asset development and the scope of consolidation are key factors that determine the financial impact on a telco of selling its fiber.

To begin, outright sales of a mature asset can result in deconsolidation of significant EBITDA, but relatively little capex. Conversely, the sale of a greenfield project is likely to deconsolidate little EBITDA, but significant capex spending. Due to these differences, the scope of consolidation in minority asset sales is important because it can distort financial ratios. For example, if a partly sold, mature fiberco is fully consolidated, the sale proceeds reduce net debt, but EBITDA and the consolidated income statement remain unchanged despite the loss of ownership of the minority stake.

We do not believe such "free lunches" are an accurate reflection of the underlying economic reality. We therefore check for the distortionary impact of full consolidation by running a proportionate consolidation analysis to gauge the difference in financial ratios.

Similarly, full consolidation can eliminate commitments between telcos and fibercos. This is important because if there are minimum commitments, our approach is the same as with tower infrastructure and we capitalize lease-like cash flows, either automatically under IFRS16 or through our operating lease adjustment. This can potentially offset the leverage benefit of sale proceeds. On the other hand, when access is secured through variable arrangements like bitstream contracts, the number of lines can vary from month to month. In these cases, we may forego adjustments and treat the cost as an expense that reduces EBITDA margins and our view of operating efficiency. And to reiterate, more flexible commitments for the telco come at a cost to the fiberco in less revenue certainty as a result.

For both the scope of consolidation and lease capitalization, if the impact is material, we will incorporate proportionate adjustments into a telco's financial ratios or tighten our financial triggers for its ratings by the amount of the distortion.

While the "free lunch" issue is of less concern with majority sales and deconsolidated reporting, we may still consider proportionate consolidation if we think it provides a more accurate view of the economic reality, as in the Altice fiber example above. For example, if a large and strategic capex program is being moved off-book during a fiber buildout, we may proportionately reconsolidate the minority interest. This can especially be the case when the minority owner is still performing the construction services and would thereby recognize the fiber investment as revenue rather than capex. Similarly, if the deconsolidation might be temporary and there is potential to reconsolidate at a future date, a proportionate approach avoids the whipsaw effect of a binary accounting view.