Open in interactive viewer → charts, metric popovers & call review

XLSMART's Merger Math Finally Works—Until the CapEx Bill Arrives

Record ARPU, returning subscribers, and $294 million of synergies make the integration story credible. A third CapEx raise and a fresh Link Net impairment keep it from being clean.
EXCL.JK · Earnings Call · 2026-11-18

The inflection is real

After years of subscriber cleanup and integration costs, XLSMART's Q3 2026 call finally had a different tone. CEO Rajeev Sethi said network integration is "close to 100% of our completion target" and cumulative gross synergies reached $294 million, ahead of the initial target and about 80% higher year-on-year. The company raised its 2026 gross synergy target to above $300 million. That is not boilerplate; it is the first time the merger's cost story has visibly outpaced its own plan. synergy realization. The customer metrics were the real surprise. Blended mobile ARPU hit an all-time high of IDR 47,500, up 34% from IDR 35,500 in Q2 2025 when the new company began. Subscribers returned to growth, up 1% quarter-on-quarter to 70 million, adding 600,000 net new customers after several quarters of deliberate rationalization. XL Home grew 7% quarter-on-quarter to 1.03 million, with fixed wireless access above 55,000. CFO Antony Susilo: “our mobile subscriber base has returned to growth, increasing by 1% quarter-on-quarter as we are adding 600,000 new customers in the third quarter” — Antony Susilo, CFO · 2026-11-18. That matters because the prior call explicitly warned of further subscriber declines while the company purged low-quality SIMs. The cleanup may be ending. quality subscribers Mobile ARPU. 5G is the monetization lever. Coverage reached 80 cities and 35,800 sites, with a target above 100 cities by end-2026. The analyst question was sharp: why no pricing premium between 4G and 5G? David Oses admitted the 5G portfolio is "evenly priced" but argued that customers who adopt it see double-digit ARPU uplift: “when the subscribers adopt our 5G portfolio, we see a double-digit increase in their ARPU” — David Oses, Executive · 2026-11-18. The passive monetization is better experience lifting consumption; the active monetization is higher-priced 5G packages. It is a plausible bridge from network investment to revenue, but still early.

The quarter was messy by design

Reported numbers look ugly. Revenue fell 4% quarter-on-quarter to IDR 11.7 trillion, driven by the absence of one-off enterprise project revenue booked in Q2. Core data and digital service revenue was stable at IDR 10.6 trillion, with data revenue up 1.8% quarter-on-quarter. Digital services declined because of a gross-to-net accounting change with mobile app revenue aggregators, not because of underlying demand. Reported PAT was a loss of IDR 606 billion, hit by accelerated depreciation from network integration and a IDR 234 billion impairment on the 20% Link Net stake. Link Net asset impairment. Normalized PAT was IDR 856 billion in Q3 and IDR 3.6 trillion for the first nine months, up 94% year-on-year and already above full-year 2025 normalized PAT of about IDR 3 trillion. The gap between reported and normalized is the story. Management said the heavy non-cash charges are done: “we already fully depreciated all the assets in Q3 2026. So Q4 is a home run, no more asset impairment.” — Antony Susilo, CFO · 2026-11-18 If true, Q4 is the first clean quarter in years. But investors have heard "one-time" before.

The CapEx raise is the tension

The headline guidance change was CapEx: capitalized CapEx for 2026 raised to roughly IDR 24 trillion from IDR 20 trillion. The analyst noted this was the second raise, after an initial IDR 5 trillion increase to IDR 20 trillion. Management's explanation is that it is not extra spend but acceleration — pulling 2027 rollout into 2026 because vendors moved faster. Rajeev:

this is not an additional spend. It is finishing off our rollout earlier than what we had planned for.

Rajeev Sethi, CEO · 2026-11-18
He added that 2026 is the peak CapEx year and 2027 will be "much smaller." CapEx guidance Capitalized CapEx. That is a reasonable explanation, but repeated raises invite skepticism. The company has now guided capitalized CapEx up multiple times in a single year, and the cash flow statement will eventually have to prove the pull-forward was real. The balance sheet can handle it: gross debt/EBITDA was 2.88x and net 2.7x at Q3, below the internal 3.5x comfort level and well under the 4.5x covenant. The IDR 6 trillion maturity in 2027 can be funded from operating cash flow, management said. It also shifted the fixed/floating mix from roughly 20/80 to about 35/65, reducing rate exposure. debt to EBITDA loan maturity. Still, the interest cost line is rising quarter-on-quarter because of spectrum-related borrowings, and the company capitalizes no interest. That is conservative accounting, but it means the P&L absorbs the full financing burden. The 700 MHz and 2.6 GHz spectrum will add regulatory fees in 2026; the planned return of 900 MHz in December 2026 should give relief in 2027.

What changed versus the prior calls

The prior February 2026 call set expectations for accelerated depreciation to continue through end-2026 and integration costs below IDR 1 trillion. The current call delivered better: integration costs were only IDR 128 billion for the first nine months, and accelerated depreciation was fully booked by Q3. Prior call: “the accelerated depreciation, I think will not be witnessed by first half of 2026, it will still continue until end of 2026.” — Antony Susilo, Executive / Senior Management · 2026-02-13 That is a positive surprise. Similarly, the prior call expected subscribers to keep declining in Q1 2026: “probably in quarter 1, you will see the total amount declining. But hopefully, internally, we will see the good quality subscribers keep increasing.” — David Oses, Executive / Senior Management · 2026-02-13 Q3 2026 showed the opposite — total subscribers grew. The cleanup phase is maturing into a growth phase. integration phase. The dividend message is consistent but worth flagging. No dividend for 2026; the company will reassess for 2027 or latest 2028, with a policy of around 30% of normalized net profit. That is not new — the prior Q3 2025 call already signaled no dividend in 2026 — but it now has a clearer timeline. dividend policy.

The bottom line

XLSMART is no longer just a cost-cutting merger. It has record ARPU, returning subscribers, and synergies ahead of plan. The integration phase is ending, and the earnings base is shifting from normalized to reported. But the quarter also shows the risks: a Link Net impairment, a reported loss, and a CapEx number that keeps moving up. The stock tape is unavailable in the context, so the market's vote is unknown. The next two quarters need to prove that the CapEx pull-forward was real, that Link Net is truly done, and that 5G monetization can lift ARPU without another price war. If they do, the "most efficient telecom provider by 2027" ambition stops sounding like a slogan.