CoBank Touting Edge Computing

A recent article from CoBank is titled, Partnerships are Key for Rural Telecom Operators in Burgeoning Edge Computing Market. The article points out that there are potential opportunities for ISPs to grab a small piece of the edge computing market.

The article defines edge computing as a network architecture where data is stored and/or processed at locations close to where applications are being used. The growth of edge computing is an interesting phenomenon to watch because it reverse the trend of the last decade, where the goal was to move as much data as possible to large data centers and not process or store at the edge.

However, as the volume of data being generated by companies has increased exponentially, the tasks of moving data back and forth from data centers has added cost and time to the equation. Companies are looking deeper at the data they generate and are realizing that a lot of the data doesn’t need to be permanently cached at data centers. Companies also want to avoid the added latency from moving and processing things in a data center.

The article cites the following potential opportunities.

  • C-RAN. The trend identified is for cellular companies to process customer connection functions locally at cell sites instead of in the cloud. The opportunity for rural ISPs is to cell more connectivity to the expanding number of cell sites. However, is cellular data is processed locally, that would imply smaller transport bandwidth needed at each cell site.
  • Private Wireless Networks. The cellular carriers and companies like Microsoft and Amazon are likely to tackle this market. While there may be a few large customers in rural markets that want to participate in a private wireless network, the big opportunity is in selling the service to farms. Local ISPs can partner with one of the big application developers that will provide a communications suite for farms. The ISP opportunity will be selling transport to farms, but also possibly being hired to maintain farm wireless devices and monitors.
  • Internet of Things (IoT). This is the trend to develop smart sensors that can handle data locally without sending everything to a data center. Like with C-RAN, it seems like a stretch to see a role for a small rural ISP in the market other than perhaps being the local agent for the sensor devices.
  • Self-Driving Cars. The article relies on a prediction that a self-driving car will need to offload as much as 5,000 gigabits per hour of driving. I find it impossible to believe that anybody is going to invest in the network in rural areas that will ever serve this market. Most of the auto industry is chasing a future where vehicles will possess the needed computing power onboard rather than rely on somebody building a fiber network and billions of sensors along every mile of US highways. I was surprised to see this in the CoBank article since the chances of this happening seem slim.

The only opportunity on this list that might realistically materialize in the next decade and be a revenue opportunity for rural ISPs is private wireless networks for farms. It’s not hard to imagine a business relationship where rural ISPs become the local agent for smart farming connectivity and devices, in much the same way that many local ISPs were the agent for products like DirecTV. It’s not hard to imagine the rural ISP industry associations negotiating a contract for such services on behalf of members, making it easy to participate.

I was intrigued to see CoBank writing this article because bankers generally concentrate on opportunities that are either here today or on the immediate horizon. This article talks about pretty futuristic stuff. The question any rural ISP will ask is if any of these applications will ever become tangible and actionable. I remember a decade ago when the rage in the industry was telling rural ISPs that there was a lot of money to be made in fostering cellular offload to WiFi. I can’t think of anybody I know that ever made a nickel on the idea, but you couldn’t go to an industry meeting without somebody promoting the idea. There is a whole lot of steps that have to happen  before any of these edge-computing ideas turn into something that the average rural ISP can profitably participate in. But I have no doubt that some of the ideas in this article, or applications we haven’t thought of, will become real eventually. The one thing that rural ISPs have that is hard to duplicate is a local presence and local technical expertise.

Is it Time to Sell?

A lot of ISPs hope to someday cash in on their sweat equity by selling the business. There have been some surprisingly high recent valuations in parts of the industry which raises the question if this is a good time to sell an ISP?

Anybody that has considered selling in the last decade years knows that valuation multiples have been stagnant and somewhat low by historic standards. A lot of properties have changed hands during that time with multiples in the range of 4.5 to 6.5 times EBITDA (earnings before interest, taxes, depreciation, and amortization). Some ISP properties have sold outside of that range based upon the unique factors of a given sale.

In November, Jeff Johnston of CoBank posted a long blog talking about how valuations might be on the rise – particularly for companies with a lot of fiber or with other upsides. He pointed to three transactions that had valuations higher than historic multiples for the sector.

  • Zayo sold their network of 130,000 route miles of fiber transport for a multiple of 11.1 times EBITDA.
  • Bluebird Network in Missouri and nearby states sold a 6.500-mile fiber transport network for a multiple of 10.4 times EBITDA.
  • Fidelity Communications of Missouri sold an ISP with nearly 135,000 customers for a multiple of 11.7 times EBITDA.

Johnston doesn’t say that these high multiples are the new standard for other ISPs. However, he does surmise that the high multiples probably indicate an uptick in valuation for the whole sector. That’s something that’s only proven over time by seeing higher valuations coming from multiple and smaller transactions – but the cited transactions raise the possibility that we’re seeing an increase in valuation for fiber-based businesses.

It’s important to ask why any buyer would pay 10 or 11 times EBITDA. A buyer paying that much will take a decade to recoup their investment if the purchased business continues to perform at historic levels. Nobody would pay that much for a business unless they expect the margins of the acquired business to improve after acquisition – that’s the key to higher valuations. The buyers of these three businesses are likely expecting significant upsides from the purchased properties.

Buyers often see a one-time bump in margin from the increased efficiency of adding an acquisition to their existing business. This is often referred to as an economy of scale improvement – overheads generally become more affordable as a business gets larger. However, buyers rarely will reward a seller for the economy of scale improvements, so this is rarely built into valuation multiples.

A buyer is usually only willing to pay a high multiple if they foresee the possibility of significant growth from the purchased entity. The purchased company needs to be operating in a footprint with upside potential, or else the purchased company needs to demonstrate that they know how to grow. A buyer must believe they can grow the acquired business enough to recoup their purchase price and also make a good return. For a fiber ISP to get a high valuation they have to be able to convince a buyer that the business has huge upside potential. An ISP needs to already be growing and they need to be able to demonstrate that the growth can be ongoing into the future.

One of the more interesting aspects of getting a high valuation multiple is that a buyer might expect the core management team to remain intact after a sale. That often means that part of the compensation from the sale might be incentive-based and paid in the future based upon post-sale performance.

To summarize, an ISP can get a higher valuation if they can convince a buyer that there is future upside to the business. ISPs that don’t have growth potential will not see the higher valuation multiples cited above – although many potential sellers will think these multiples apply to them. The bottom line is that if your ISP is growing and can keep growing, and you can paint that picture to a buyer, your business might be worth more than you expected.

CoBank Supports Telemedicine

For those who don’t know CoBank, it’s a bank that specializes in loans to telecom and electric cooperative but which also has funded numerous rural fiber projects for other borrowers over the years. In August CoBank filed comments In FCC Docket 18-213 in support of expanded use of the Universal Service Fund for rural telemedicine. CoBank is a huge supporter of telemedicine and has made substantial grants to telemedicine projects dealing with diabetes management, opioid abuse, prenatal maternity care, and veteran care.

As part of that filing, CoBank discussed a telemedicine trial they had sponsored in rural Georgia. The trial was conducted in conjunction with Perry Health, a software provider and Navicent Health, a healthcare provider in Macon, Georgia.  The trial was for 100 low-income patients with uncontrolled Type 2 diabetes. These patients were on a path towards kidney failure, amputation, loss of vision, and numerous other major related health problems. These are patients who would normally be making numerous emergency room visits and needing other costly medical procedures.

In the trial, the patients were provided with tablets containing Perry Health software that provided for daily interaction between patients and Navicent. Patients were asked to provide daily feedback on how they were sticking to the treatment regimen and provided information like the results of blood sugar tests, the food they ate each day, the amount of daily exercise, etc. The tablet portal also provided for communication from Navicent asking patients how they generally felt and providing recommendations when there was a perceived need.

The results of the trial were hugely positive. In the trial of 100 patents, 75% of the patients in the trial showed a marked improvement in their condition compared to the average diabetes patient. The improvements for these patients equated to reduced health care costs of $3,855 per patient per year through reduced doctor visits and reduced needs to make emergency room visits. The American Diabetes Association says that patients with Type 2 diabetes have 2-3 times the normally expected medical costs, which they estimate totals to $327 billion per year.

Patients in the trial liked the daily interaction which forced them to concentrate on following treatment plans. They felt like their health care provider cared about how they were doing, and that led them to do better. After the trial, Navicent Health expanded the telemedicine plan to hundreds of other patients with Type 2 diabetes, heart failure, and Chronic Obstructive Pulmonary Disease (COPD).

One of the interesting outcomes of the trial was that patents preferred to use cellphones rather than the special tablets. The trial also showed the need for better broadband. One of the challenges of the trial was the effort required by Navicent Health to make sure that a patient had the needed access to broadband. To some degree using cellphones gives patients easier access to broadband. However, there are plenty of rural areas with poor cellular data coverage, and even where patients can use cellular data, the cost of cellular data can be prohibitive if heavily used. Landline broadband is still the preferred connection to take advantage of unlimited WiFi connections to the healthcare portal.

One thing that struck me about this study is that this sounds like it would be equally useful in urban areas. I’ve read that a lot of healthcare costs are due to patients who don’t follow through on a treatment plan after they go home after a procedure. The Navicent Health process could be applied to patients anywhere since the biggest benefit of the trial looks to be due to the daily interface between patient and doctor.

The FCC has already pledged to increase funding for the rural medicine component of the Universal Service Fund. However, that funding is restricted. For example, funding can only be granted to rural non-profit health care providers.

Telemedicine has been picking up steam and is seeing exponential growth. But telemedicine still only represents just a few percentages of rural healthcare visits. The primary barrier seems to be acceptance of the process and the willingness of health care providers to tackle telemedicine.

Banking Challenges for Fiber Builders

I’ve often mentioned in this blog that it’s gotten harder to finance fiber infrastructure. Today I want to discuss a few of the specific issues that fiber builders face when trying to find bank financing. There are two traditional sources of funding for the industry – the Rural Utility Service (RUS) and CoBank.  However, many fiber builders don’t qualify for this funding since both institutions favor established mature companies. Any company that doesn’t fit the profile of these two lenders must turn to the only other source of funding – local and regional banks. Following are some of the issues I see when trying to borrow from banks.

Familiarity with the Industry. Local banks often are leery about lending to telecom companies because they are not familiar with the business and they fear lending into an unknown industry. Local banks are much more comfortable lending to businesses they understand and make loans to car dealers, retail stores and the other kinds of local businesses that have been their long-term core borrowers.

Amount of Borrowing. Every bank has some pre-determined maximum amount they are willing to lend to any one borrower and it’s easy for a fiber overbuilder to quickly hit this limit. I’ve rarely met a fiber overbuilder who doesn’t see endless opportunities for expansion and it’s not hard to hit a bank’s maximum lending limit.

Loan Terms. Local banks are often uncomfortable with the longer-term loans needed to finance fiber.  Banks prefer to make loans for relatively short periods of time, with their preference being short loans of 2 – 5 years. Fiber builders are often forced to only chase projects that fit the short loan terms – which means cherry picking only the best opportunities. In doing so they will be passing up opportunities that would thrive and produce good returns with a longer loan terms of 5 – 15 years.

Collateral. Banks are often uncomfortable with a fiber network as collateral. It’s not hard to blame them for this. A fiber network, once in the ground or on the pole does not automatically have a liquidation valley equal to the cost of the construction. The real value of a fiber network is the revenues from customers who are added to the network – and banks have a hard time accepting this concept. A little research will show bankers that failed fiber ventures have often liquidated the physical fiber network for pennies on the dollar, and that rightfully frightens them.

Quantifying Risk. It can be difficult for a bank to understand the downside risks of building a fiber project. One of the key steps to making a loan is to understand the likelihood of the borrower not meeting the proposed business plan, and bankers have a hard time quantifying and getting comfortable with the potential downsides of the proposed business.

Meeting Metrics. Many banks are driven by metrics – meaning that they look for key financial performance metrics from a borrower. It’s hard to meet the typical metrics for a new fiber network. When a network is first built it boosts the balance sheet – but revenues then lag a few years behind until the new network has enough customers to meet expected metrics. This cycle of early losses followed by eventual gains does not fit easily into the expectations of a metric-driven bank.

Unfortunately, any one of these issues can convince a bank that the fiber loan is too risky or doesn’t fit their comfort zone. Many banks are comfortable with infrastructure loans, but there are infrastructure loans that better meet their expectations. Consider a loan to build an apartment complex. There is the same period of zero revenues while the buildings are constructed, but the expectation is that the borrower will then quickly reach full revenues within a relatively short period after the end of construction. An apartment building also provides comfortable collateral because there is an established market for selling repossessed buildings. Bankers in general understand the apartment complex operating model and are comfortable with the variables of operating an apartment building.

Fiber overbuilders need to be prepared to tell a story that can get a banker comfortable with each one of these concerns. I always advise fiber builders that they must put themselves into the banker’s shoes and look at their own business plan as a skeptic. I’ve often seen fiber builders who point to a business plan that eventually makes a lot of money and who can’t understand why a banker doesn’t see their plan the same way they do. Many of the misgivings that a banker might have about funding a fiber project are legitimate and the borrower must convince the banker that the overall level of risk is small – a tall task.

Maybe Coops are the Answer

I’ve been talking with a lot of rural counties lately and also with rural service providers. For the vast majority of rural broadband projects the biggest roadblock to getting started is almost always funding. Building fiber-to-the-home or even fiber backbones to extend fiber deeper into rural communities is expensive and there are not a lot of funding sources ready to support fiber projects. But there is one business structure that can sometimes make financing a little easier and perhaps it is time for more communities to consider forming a cooperative as a way to get a broadband solution.

Cooperatives are governed under federal law by the Capper Volstead Act. There are also state laws governing coops that differ a bit from state to state, but are mostly the same everywhere. A cooperative is a legal entity owned and controlled by its members and members generally are also the consumers of its products or services. Cooperatives are typically based on the cooperative values of self-help, self-responsibility, community concern, and caring for others. Cooperatives generally aim to provide their goods or services at close to cost and any excess earnings are generally required by law to be reinvested in the enterprise or returned to individual patrons based on patronage of the cooperative.

There are several advantages of coops that make them worth considering:

  • Coops are corporations and not municipal entities. Coops ought to be exempt from all of the many state laws that prohibit or discourage municipal ownership of broadband networks. If you’re in a place that makes it hard to create a municipal broadband solution then a cooperative might be a great alternative.
  • Cooperatives don’t have the same profit-motive as privately-owned entities. From a financing perspective this makes them look more like a municipal venture in that a coop is happy with cash flows that cover costs rather than having to also make a profit.
  • Cooperatives often have some tax advantages over other kinds of corporations. For example, ‘profits’ from serving their customers is often income-tax free. This can vary by state, but for the most part cooperatives pay little income taxes as long as they focus only on serving their own members.
  • The typical financing sources for broadband are used to working with cooperatives. The RUS, part of the Department of Agriculture has a long history of lending to cooperatives. CoBank, a bank that is part of the US Farm Credit System was established specifically to loan to agricultural, electric and telecom cooperatives. While the RUS was tasked a number of years ago to include municipalities under their umbrella, the nuances of that program make it nearly impossible for a municipality to borrow from them.
  • Cooperatives have a unique funding source that is not available to anybody else. Coops are allowed to loan excess cash to each other and I’ve seen new coops get low, or even zero interest loans from other cooperatives to help them get started. Many older electric and agriculture coops sit on big cash reserves that they might consider lending – particularly when the new telecom cooperative covers the same member territory.

But as you might expect, there are other issues that present challenges for new cooperatives:

  • Any lender to a new fiber venture is going to want to see some equity put into a new venture so that it is not 100% financed. It can be more of a challenge for a cooperative to raise equity compared to a commercial company because there is no way to guarantee that such equity will earn a good return or that it can be returned in any reasonable time frame. So this basically means that an equity drive means asking prospective members in the community for money. I’ve seen a few cooperatives get started and it can be done – but it’s not easy.
  • Cooperatives are governed by Boards elected from the membership base. Existing coops hire employees to operate the business and these employees provide the technical expertise that makes lenders trust lending to the business. But until a new cooperative is funded and can hire those employees there is a classic chicken-and-egg dilemma in that a lender can’t be positive that the cooperative knows how to operate their business.
  • The local acceptance of the cooperative idea varies by region. In some places in the Midwest a majority of local businesses are cooperatives, but there are other places where there are few if any cooperatives.

There are many situations where a cooperative might be the only reasonable operating structure for a rural area to get the broadband they want. If a community is not finding any solutions from a commercial provider and is unable to provide municipal funding, then a cooperative is well worth considering.