ISAT How Do You Lose Half Your Subscribers and Call It a Cleaner Customer Base?

The SIM Card Purge, and What It Actually Cost

Nine years after the last quarter this site covered, Indosat looks recognizably like the same company - Ooredoo Asia Pte. Ltd. still holds 65.00% (the "Qtel" that took control back in 2009 rebranded to Ooredoo), the Government of Indonesia still holds 14.29% and a golden veto share, and the same three-segment structure (Cellular, MIDI^», and Fixed Telecommunications) still runs the P&L. What's different is that 2018 was the year the Indonesian government forced every cellular operator to re-register prepaid SIM cards against a citizen's national ID and family card number, to combat SIM-based fraud and enable lawful-intercept compliance. Indosat says compliance alone cut its customer base 22% in the first half of 2018 alone compared with a year earlier. By year-end, prepaid subscribers had fallen from 109.0 million to 56.4 million - a 48.2% drop - and total cellular subscribers from 110.2 million to 58.0 million, down 47.3%.

That regulatory purge landed on top of a deliberate strategy shift Indosat's own management describes as moving from a "push" go-to-market model - the entire Indonesian industry's norm, built on pre-activated SIM cards handed out in bulk through distribution channels - to a "pull" model where the company stopped pre-activating SIM cards altogether, accepting a known short-term revenue hit in exchange for a cleaner, more monetizable customer base. Both effects show up in the same numbers, and the filing doesn't cleanly separate how much of the subscriber loss was the regulation and how much was the company's own choice to stop juicing gross additions - which matters, because one is a one-time compliance cost and the other is a management decision that could recur.

The result: full-year revenue fell 22.7% to Rp23,139.6 billion, Cellular revenue alone fell 26.4%, and Adjusted EBITDA^» nearly halved. Indosat posted a net loss for the year - the first quarter of this company's history covered on this site where the bottom line was actually negative. But the headline loss undersells how bad the underlying business got, not the other way around: buried inside "other income" is a Rp924.9 billion gain that has nothing to do with phones, voice, or data (see Beyond the Usual). Management's own framing - that the network finished re-optimizing by mid-year and commercial momentum "reignited" in Q3 and finished 2018 on "a strong fourth quarter" - is the case for reading this as a one-year shock rather than a structural decline. The numbers below are the test of that claim.

The Prescription

Indosat should keep leaning into the MIDI segment and enterprise cellular, the two places where this year's damage was smallest and the underlying demand story is real. B2B initiatives already contribute nearly 23% of cellular revenue by management's own account, and MIDI - even with margin cut nearly in half this year - is still the only one of the three segments that stayed solidly profitable while cellular swung to an outright segment loss (see Segment Comparison). A subscriber base that just shed half its names is also, by definition, a much cleaner one - the SIM re-registration purge, however painful, removed exactly the kind of low-value, possibly-fraudulent prepaid lines that never generated real ARPU^» to begin with. Building the next phase of growth on quality data subscribers and enterprise connectivity, rather than chasing gross prepaid additions back to 110 million, is the only version of this business that survives the next regulatory or price-war shock.

What it should stop doing: treating vendor-financed capex as a free lunch, in the same year its covenant cushion nearly evaporated - see the companion post on this year's leverage and governance findings for the full picture.

Key Financial Metrics

FY2018 (year ended Dec 31, 2018) vs FY2017, consolidated

FX: Rp14,481 = US$1 (Dec 31, 2018) and Rp13,548 = US$1 (Dec 31, 2017) - Indosat's own period-end Bank Indonesia reference rates, as disclosed in the filing. Converting each year at its own period-end rate means the USD year-over-year change below reflects both the underlying rupiah result and the rupiah's depreciation against the dollar over the year.

Metric	FY2018 (Rp)	FY2018 (US$)	FY2017 (Rp)	YoY (Rp)	YoY (US$)
Revenue	Rp23,139.6B	~$1,598.0M	Rp29,926.1B	⚠️ -22.7%	⚠️ -27.7%
Adjusted EBITDA» (operating income + D&A)	Rp6,500.1B	~$448.9M	Rp12,762.8B	⚠️ -49.1%	⚠️ -52.4%
Operating Income»	-Rp1,748.9B	~-$120.8M	Rp3,909.9B	⚠️ swung to loss	⚠️ swung to loss
Net Income (attributable to owners)	-Rp2,403.8B	~-$166.0M	Rp1,135.8B	⚠️ swung to loss	⚠️ swung to loss
Free Cash Flow» (operating cash flow - capex)	-Rp1,296.4B	~-$89.5M	Rp3,129.7B	⚠️ swung to cash burn	⚠️ swung to cash burn
Total Cash	Rp1,045.0B	~$72.2M	Rp1,674.7B	⚠️ -37.6%	⚠️ -37.6%*

*Cash held up slightly better in USD terms than the raw percentage suggests once each year's own period-end rate is applied, but the direction is the same either way.

Basic loss per share was Rp(442.38), against Rp209.02 of earnings per share in 2017 - the same 5,433,933,499 Series B shares (plus the single golden Series A share) outstanding as in the 2009 quarter this site last covered, so no dilution is muddying the comparison. Including non-controlling interests, the Group's total net loss for the year was Rp2,085.1 billion (against Rp1,301.9 billion of total profit in 2017); the larger loss attributable to owners of the parent reflects that non-controlling interests actually posted a gain this year (see Beyond the Usual on why).

Revenue, EBITDA, operating income, and free cash flow all moved the same direction this year - down, sharply, together. Unlike the 2009 quarter this site covered, where a currency loss masked an otherwise healthy operating quarter, this year's headline loss is real and operational, not a one-line accounting artifact - if anything, one accounting item is making it look better than the underlying business actually performed (see Beyond the Usual).

Key Operational Metrics

Metric	FY2018	FY2017	YoY
Total cellular subscribers	58.0 million	110.2 million	⚠️ -47.3%
- Prepaid	56.4 million	109.0 million	⚠️ -48.2%
- Postpaid	1.6 million	1.2 million	✅ +32.3%
Blended ARPU»	Rp18.7 thousand/month	Rp20.3 thousand/month	⚠️ -7.9%
Prepaid ARPU	Rp17.3 thousand/month	Rp19.0 thousand/month	⚠️ -8.9%
Postpaid ARPU	Rp92.4 thousand/month	Rp129.1 thousand/month	⚠️ -28.4%
Data usage	1,870,428 terabytes	1,082,942 terabytes	✅ +72.7%
Data share of cellular revenue	69.4%	59.0%	✅
Employees (incl. subsidiaries)	3,700	4,392	✅ -15.8%

The postpaid base actually grew (+32.3%) even as the overall subscriber count nearly halved - it was the low-value, unregistered prepaid lines that the government's rule and the company's own "pull strategy" pivot both targeted. Data usage growing 72.7% while data yield stayed among the lowest in the world globally (per management's own characterization) explains why data's rising share of revenue (59.0% to 69.4%) couldn't offset the collapse in voice and SMS: voice revenue fell from Rp7,086.1 billion to Rp4,521.4 billion (-36.2%) and SMS revenue fell from Rp3,829.4 billion to Rp1,551.6 billion (-59.5%) as both were substituted by cheaper, lower-yielding data and OTT alternatives - a structural industry shift Indosat calls out explicitly, not something unique to this year's regulatory shock.

Segment Comparison

Indosat reports three segments: Cellular, MIDI (fixed connectivity, IT services, fixed internet), and Fixed Telecommunications (international calls and fixed-line telephony). Segment profit below is calculated as segment revenue less depreciation/amortization and other segment expenses, consistent with how the company's own segment footnote breaks out the numbers; it ties closely, but not exactly, to consolidated Operating Income above because of intersegment eliminations.

Segment	Revenue FY2018	Revenue FY2017	YoY	Segment Profit (Loss) FY2018	Margin 2018	Margin 2017
Cellular	Rp18,064.9B	Rp24,526.7B	⚠️ -26.4%	-Rp2,113.7B	⚠️ -11.7%	✅ +12.2%
Fixed Telecommunications	Rp729.3B	Rp913.0B	⚠️ -20.1%	-Rp203.5B	⚠️ -27.9%	⚠️ -13.4%
MIDI	Rp4,383.3B	Rp4,517.5B	⚠️ -3.0%	Rp606.3B	✅ +13.8%	✅ +23.9%

Every segment shrank this year, but only two of the three flipped, or stayed, in the red. Cellular - still 78% of total revenue - is the one that matters most, and it went from Indosat's most profitable segment (12.2% margin) to a loss-making one (-11.7% margin) in a single year, a nearly 24-point swing. MIDI is the only segment that held a real profit margin, though even it was cut nearly in half (23.9% to 13.8%) as its own revenue slipped 3.0% - largely IT services and electronic payment revenue falling from Rp1,224.5 billion to Rp857.1 billion, partly a function of Indosat's Lintasarta subsidiary losing consolidated control over its payments unit mid-year (see Beyond the Usual).

Cellular

The flagship segment absorbed almost the entire year's damage: a Rp6,461.8 billion revenue decline and a swing from +Rp2,983.3 billion segment profit to -Rp2,113.7 billion segment loss. Capital expenditure allocated to Cellular still rose 58.8% year-over-year (Rp5,174.9 billion to Rp8,218.5 billion, on the segment note's accrual basis) even as its revenue cratered - Indosat kept building network capacity through the subscriber purge rather than pulling back, a bet that the post-purge, higher-quality base can eventually be monetized at a healthier ARPU.

MIDI

MIDI held onto its position as the one profitable, resilient segment, even as its own numbers softened. Fixed connectivity revenue was essentially flat (Rp2,530.2 billion to Rp2,506.5 billion), fixed internet grew (Rp762.8 billion to Rp1,019.7 billion, +33.7%), and only IT services/electronic payment revenue declined meaningfully - the segment where the Lintasarta/APE ownership change (see Beyond the Usual) shows up most directly.

Fixed Telecommunications

The smallest segment (3.2% of revenue) was already loss-making in 2017 and got worse: revenue fell 20.1% on declining international call volumes (incoming traffic minutes down 26.1%, outgoing down 65.8%), and segment margin worsened from -13.4% to -27.9%. This is a shrinking legacy business by any measure, and unlike the 2009 quarter this site covered - when Fixed Telecommunications was Indosat's highest-margin segment - it's now the only one of the three that's structurally unprofitable in both years shown here.

Beyond the Usual

The reported loss includes a nine-figure gain that has nothing to do with running a telecom

Buried inside "other income" is a Rp924.9 billion "gain associated with the loss of control of a subsidiary" - and it isn't from selling towers or divesting a business unit for cash. Indosat's subsidiary PT Aplikanusa Lintasarta owned 55% of Artajasa Pembayaran Elektronis ("APE"), an electronic payment company. To comply with a Bank Indonesia regulation restricting foreign ownership of payment-system providers, Lintasarta's shareholders restructured APE's shares in April 2018 so that Lintasarta kept its 55% economic ownership but only 20% of voting rights - losing accounting control without giving up a rupiah of its stake. Under accounting rules, that loss of control triggers a one-time fair-value remeasurement of the remaining investment, and in this case it produced a Rp924.9 billion non-cash gain, recognized entirely because of a regulatory ownership-structure change, not because the business it relates to did anything differently. Strip that gain back out and the year's operating story - a segment loss in Cellular, a near-halving of EBITDA, a genuine cash burn - looks worse than the headline net loss already shows, not better.

This year's covenant-cushion, vendor-financed-capex, and related-party leverage findings - including a decade-old corruption case now booked as a debt to the Government of Indonesia - are covered in the companion post rather than repeated here.

An accounting policy change quietly trimmed the reported revenue decline. Indosat early-adopted two new Indonesian accounting standards - PSAK 71 (financial instruments) and PSAK 72 (revenue from contracts with customers), the local equivalents of IFRS 9 and IFRS 15 - two years ahead of their mandatory effective date. One consequence: distribution-channel incentives that used to be booked as marketing expense are now treated as a reduction to transaction price (contra-revenue) under PSAK 72. The company's own disclosure shows this reclassification alone reduced reported Cellular revenue by Rp365.9 billion and MIDI revenue by a further Rp9.2 billion in 2018, with a matching reduction to marketing expense - a wash on the bottom line, but it means part of this year's headline revenue decline is a bookkeeping reclassification against an unrestated 2017 comparative, not pure subscriber-driven decline.

Nine figures of vendor commitments, still open. As of December 31, 2018, Indosat had signed but not fully executed procurement commitments to PT Huawei Tech Investment (USD27 thousand and Rp810.0 billion), PT Ericsson Indonesia and Ericsson AB combined (USD26,290 thousand and Rp519.8 billion), PT Nokia Siemens Networks (Rp1,199.3 billion), and smaller IT-services commitments to NEC and ZTE - the paper trail behind the network buildout that continued straight through this year's subscriber collapse.

A legacy tower deal keeps generating contracted income years later. Following the 2012 sale-and-leaseback of 2,500 telecommunication towers to Tower Bersama Infrastructure, Indosat still holds future minimum lease payments receivable of Rp4,081.6 billion as the lessor on space it subleases within those same towers to other operators (including PT Hutchison 3 Indonesia, PT Dayamitra Telekomunikasi, and PT Smartfren Telecom), up from Rp3,932.8 billion in 2017 - a reminder that a sale-and-leaseback from six years earlier is still an active, growing revenue stream, not a closed transaction.

Target Valuation Range

Bottom line: the stock is priced for a business in genuine distress, not just a one-year regulatory shock - and on the numbers actually available this year, the market's skepticism is hard to argue with.

Indosat's shares closed at a split-adjusted Rp421.25 on December 31, 2018 on a historical price basis - Indosat completed a 1-for-4 forward stock split in October 2024, so that historical close is stated on a post-split basis. Multiplying by 4 recovers the actual nominal price quoted on the Indonesia Stock Exchange that day: Rp1,685.

Market cap → enterprise value	FY2018 (period-end)
Share price (period-end, nominal)	Rp1,685
Shares outstanding (Series B)	5,433,933,500
Market capitalization	~Rp9,156.2 billion (~$632.3 million)
Plus: gross debt (incl. finance leases)	Rp24,945.5 billion
Less: cash	Rp1,045.0 billion
Enterprise value	~Rp33,056.6 billion

Peer-multiple sanity check	Q1 2009	FY2018	Change
P/E	~54x (not meaningful)	not meaningful (loss-making)	-
Book value per share	~Rp3,226	~Rp2,056	⚠️ down
P/B	~1.46x	~0.82x	⚠️ down
EBITDA	Rp8,726.7 billion (annualized)	Rp6,500.1 billion (full-year)	⚠️ down
EV/EBITDA	~5.06x	~5.09x	➖ roughly flat

This is the first quarter of profitability data available for this company since Q1 2009, nearly a decade earlier with no history in between, so the "prior quarter" column above is really a prior-decade reference point, not a genuine sequential comparison. P/E is not meaningful this year - Indosat posted a loss, so there's no positive earnings figure to divide the price by. The stock trades below book value for the first time in this company's coverage here - a level that either prices in further value destruction or represents a genuine dislocation, depending on whether 2018 turns out to be a one-year shock or the start of a trend. The EV/EBITDA multiple lands almost exactly where it did back in Q1 2009 - but that similarity is deceptive, not reassuring: the 2009 multiple was calculated off a genuinely stable operating quarter, while this year's EBITDA is itself down 49.1%. A "normal-looking" multiple applied to a severely depressed earnings base isn't cheap - it's a coin flip on whether that earnings base recovers.

No DCF is included this quarter. A discounted cash flow needs a normalized view of forward free cash flow, and this year's actual FCF (-Rp1,296.4 billion) is dominated by a one-time regulatory subscriber shock that may or may not repeat - projecting it forward either assumes a full recovery (optimistic) or assumes this year's collapse is the new baseline (probably too pessimistic, given the Q3/Q4 recovery management describes). The peer-multiple read above is the more honest signal until at least one more quarter of post-purge numbers is available to show which way the business is actually trending. Over the two years to December 31, 2018, the stock fell roughly 74% on a nominal basis (from Rp6,450 in January 2017 to Rp1,685 at year-end 2018) - a decline that tracks the subscriber and earnings collapse closely enough that this doesn't read as a market overreaction so much as a fairly literal repricing to the new numbers.

---