The same fibre network modelled twice, once selling broadband to households and once selling lines to internet providers, with the penetration, wholesale price and exit multiple each model needs.
On the same network, retail earns more unless open access makes it up with penetration, price or exit value. On Oakmere Fiber, the fictional network of The Digital Infrastructure Investor, retail earns an unlevered IRR of 10.34 per cent at 40 per cent penetration, open access 7.55 per cent at 50 per cent. To match retail, the open access model needs 57.2 per cent penetration, a wholesale price of 44.01 dollars instead of 40, or an exit multiple of 18.9 times instead of 14.
Worked in full in The Digital Infrastructure Investor by Julian R. Sterling, with every figure reproduced in a free workbook.See the book on Amazon →
A retail fibre operator sells broadband to the household: it bills, markets, installs and answers the phone. An open access operator sells lines to internet providers, who own the customer. The network, the build cost and the cost of running the fibre are the same. What changes is revenue per connected home, the direct cost of each subscriber, the overhead, the churn and, the open access case argues, the take-up, because several providers sell on one network.
A fund buys Oakmere for 130 million plus 2 per cent costs: 60,000 homes passed, three years old, growing by 20,000 a year to 120,000. It carries no debt and exits after ten years at 14 times year-10 EBITDA less 1.5 per cent. Build at 1,150 dollars per home passed, connections at 700, reconnections after churn at 200, network opex 3 dollars a month per home passed. Each build cohort follows the same take-up curve, reaching terminal penetration at age six. All figures are illustrative.
| Input | Retail | Open access |
|---|---|---|
| Terminal penetration | 40% | 50% |
| Revenue per line, year 1, $ a month | 70 | 40 |
| Opex per subscriber, $ a month | 18 | 5 |
| Contribution per line, $ a month | 52 | 35 |
| Overhead, year 1, $m | 4 | 3 |
| Churn a month (a year) | 1.3% (15.6%) | 0.5% (6.0%) |
Step 1: subscribers. Each cohort's homes passed times terminal penetration times the curve at its age. In year 10 retail has 48,000 subscribers on 120,000 homes and open access 60,000.
Step 2: EBITDA. Revenue is average subscribers times ARPU times 12, ARPU rising 2 per cent a year. Deduct subscriber opex, network opex per home passed and overhead.
Retail, year 10: revenue 48.19, EBITDA 24.62, margin 51.1%
Open access, year 10: revenue 34.42, EBITDA 20.61, margin 59.9%
Step 3: free cash flow. Deduct build (71.09 in total, identical), connection capex (new subscribers at 700, churned ones reconnected at 200) and maintenance. Open access connects more homes but churns fewer, so its connection capex, 35.04, is close to retail's 36.97.
Step 4: exit and IRR. Year-10 EBITDA times 14, less 1.5 per cent: 339.6 for retail, 284.3 for open access.
Cash flows: −132.6 at year 0, free cash flow in years 1 to 10, exit in year 10
Retail IRR = 10.34%, multiple 2.34x
Open access IRR = 7.55%, multiple 1.88x
Excel: =IRR(D50:N50), the cash-flow row of each model in the workbook
| Year | Retail subscribers | Retail EBITDA | Open access lines | Open access EBITDA |
|---|---|---|---|---|
| 1 | 23,400 | 6.4 | 29,250 | 5.3 |
| 3 | 36,800 | 13.1 | 46,000 | 10.8 |
| 5 | 44,600 | 19.6 | 55,750 | 16.2 |
| 10 | 48,000 | 24.6 | 60,000 | 20.6 |
Open access has a quarter more lines and a higher margin, and still less EBITDA in every year. A wholesale line earns 35 dollars of contribution against 52, two-thirds as much, while the network costs the same to build and run. The higher margin is a ratio on a smaller revenue line, and a fund is paid in dollars, not in margin.
| Penetration | Year-10 EBITDA | IRR | Multiple |
|---|---|---|---|
| 40% | 14.63 | 2.77% | 1.27x |
| 45% | 17.62 | 5.33% | 1.57x |
| 50% | 20.61 | 7.55% | 1.88x |
| 55% | 23.61 | 9.52% | 2.19x |
| 60% | 26.60 | 11.30% | 2.51x |
Three thresholds, each solved with the other inputs held:
At equal penetration the gap is wider still: retail at 50 per cent would earn 15.14 per cent. The case for open access rests entirely on its claim to more lines, a lower risk profile or a richer exit, and each of those has to be evidenced, because the base arithmetic runs against it.
The common mistake is to compare margins. Open access's 59.9 per cent EBITDA margin looks better than retail's 51.1, and the lower churn, 6.0 per cent a year against 15.6, looks safer. Neither decides the return: the dollars of EBITDA do, and they are lower. The second mistake is to give the open access case its higher penetration for free. Take-up is the input every return here depends on, and the cohort curve, ARPU and churn should be tested together, as in what penetration an FTTH network needs to earn 10 per cent. Both models, cohort by cohort, are in the free workbook for Oakmere Fiber; lifetime value per subscriber under each model is compared in how to calculate a fibre customer's lifetime value.
A network whose owner sells wholesale access to several internet providers instead of selling broadband to households itself. The owner gives up the retail margin but saves billing, marketing and customer-care costs and usually suffers less churn. On an illustrative network the wholesale line earns 35 dollars a month of contribution against 52 for a retail subscriber.
Because each line earns less. On an illustrative FTTH network, a retail subscriber pays 70 dollars a month against 18 of direct cost, a wholesale line 40 against 5. With the same build, network opex and connection capex, open access reaches 10 per cent only at 56.3 per cent penetration, against 39.4 per cent for retail.
Some buyers argue so, because wholesale cash flows are less exposed to retail competition and churn. On an illustrative network, open access at 50 per cent penetration needs an 18.9x exit multiple on year-10 EBITDA to match retail's 10.34 per cent at 14x. At 16x it earns 8.76 per cent. That is a large premium to assume.
This article is one calculation from The Digital Infrastructure Investor. The book takes the same case from first principles to the decision, chapter by chapter, and every figure it prints is a live formula in the free companion workbooks.
Get the book on Amazon →Free companion files
Also on Amazon UK · Amazon Germany · Amazon France · Amazon Canada
Reading guide: infrastructure, data centres and energy → · All 1340 articles →
If this book helped, or didn’t, a few lines on Amazon are worth more than they look: they are what the next reader goes on. Write a review. The workbook stays free either way.