Chapter 5

Control and Capital

TOWR is controlled by Indonesia's Hartono family — the owners of Djarum and Bank Central Asia — through two vehicles that together hold about 65% of the shares [1]. In 2025 that family funded almost the entire Rp5.5 trillion rights issue that cut leverage, subscribing new shares at Rp680 against a price now near Rp414 [2]. This chapter audits the alignment behind the "strong-promoter" side of the thesis: who controls TOWR, what they have done with capital, and the related-party arrangements a skeptic should price in.

Who controls the company

The register lists two controlling vehicles. As of the end of 2025, PT Sapta Adhikari Investama held 45.29% and PT Dwimuria Investama Andalan held 19.95% — a combined 65.24% [3]. Both trace to the Hartono family: the audited related-party note groups PT Bank Central Asia, PT Djarum, PT Grand Indonesia and Blibli's parent PT Global Digital Niaga under "family relationship with the ultimate shareholders" [4], and the list extends to Polytron's PT Hartono Istana Teknologi and other family holdings [5]. This is a founder-controlled asset inside one of Indonesia's largest private groups, not a professionally-diffused public company.

The two vehicles are not static. Before the July 2025 rights issue, Sapta held 54.42% and Dwimuria only 5.59% — a combined 60.01% [6]. The rights issue reshuffled the holding: Sapta's stake fell as it left rights unexercised, while Dwimuria — the standby buyer — more than tripled its position. The family's combined stake rose by roughly five points even as the split between its two vehicles inverted.

Sources: end-2025 register per the Q1 FY2026 financial statements [7]; pre-issue register per the Q3 FY2025 financial statements [8].

The rights issue was a family-funded deleveraging

The single most important capital-allocation event of the period was the July 2025 rights issue. TOWR offered 8,083,478,731 new shares — 13.91% of the enlarged capital — at Rp680 each, raising about Rp5.50 trillion, with Dwimuria (then holding 8.33%) committing to take up its full entitlement and to act as standby buyer for any shares other holders left behind [9]. Every rupiah went into subsidiary Protelindo, which used it "to repay all and/or part of its principal debt" [10]. This is the raise that took net debt/EBITDA from about 4.6 times to 3.74 [11].

The mechanics matter for alignment. The family's combined holding grew from roughly 30.6 billion shares (60.01% of 51.0 billion) to about 38.6 billion (65.24% of 59.1 billion) — an increase of close to 7.9 billion shares. At Rp680, that is on the order of Rp5.4 trillion of fresh cash from the family, roughly 98% of the entire raise. Minority holders who did not follow their rights were diluted from about 40% of the company to under 35%; those who subscribed pro-rata were not. Either way, the deleveraging that reset the balance sheet was paid for almost entirely by the controlling family — and at Rp680, well above today's Rp414. On that injection the family currently sits at a paper loss, which is the clearest available evidence that its subscription was a conviction commitment rather than an opportunistic mop-up at a distressed price.

Capital allocation: balance sheet first

The rights issue is one move in a consistent pattern: through 2025 management prioritised the balance sheet over near-term shareholder payout, while continuing to make small, adjacent acquisitions.

The dividend was cut, not defended. Cash dividends fell from Rp24.1 per share for FY2023 to about Rp800 billion (near Rp15.9 per share) for FY2024 [12], and the only FY2025 distribution declared so far is an interim of Rp6.87 per share resolved in December 2025 [13]. For an investor who fears bankruptcy above all, a controller that trims the dividend and injects equity to cut leverage — rather than sustaining a payout on borrowed money — is behaving the right way round.

Sources: FY2023 and FY2024 dividends per the FY2025 Annual Report dividend note [14]; FY2025 interim per the corporate site disclosures [15]. FY2025 is an interim figure only; a final dividend for FY2025 had not been declared at the report date.

M&A stayed small and on-strategy. In April 2025 subsidiary iForte bought 40% of listed fibre operator PT Remala Abadi (DATA) at Rp974 per share, triggering a mandatory tender offer and a call option toward 51% [16]; in September it completed a 51% purchase of connectivity provider PT Media Antar Nusa (MAN) [17]. Both extend the fibre-and-connectivity build that the earlier chapters identified as the group's only real growth engine, and both are modest relative to a company carrying tens of trillions of rupiah in assets.

The buyback was opportunistic and small. Between 4 September and 3 December 2025, with the shares falling, TOWR repurchased about 122.4 million shares using internal funds, explicitly to support price stability [18] — roughly 0.2% of the share count, a signal rather than a return-of-capital programme. The sequencing is telling: raise Rp5.5 trillion of equity to cut debt, then buy back a token amount of stock at a lower price. It reads as balance-sheet discipline first, with the buyback as a modest vote of confidence rather than the main event.

What a skeptic checks: the related-party web

Family control cuts both ways, and the reasonable worry is self-dealing through the group. The audited accounts let us size it. Transactions with related parties are stated to be "based on terms and conditions agreed among the parties" [19], and in the profit-and-loss they are small: related-party revenue was 1% of total revenue and related-party cost of revenue about 5% in FY2025 [20]. Key management compensation was Rp114.3 billion, about 3% of net profit and lower than the prior year [21].

Sources: related-party revenue and cost percentages and key-management compensation, FY2025 Annual Report [22] [23]; BCA loan and total debt per the related-party balances and financial highlights [24] [25].

Two arrangements deserve to be named rather than buried. First, the largest single related-party item is a bank loan of Rp4.00 trillion from PT Bank Central Asia — the family's own bank — down from Rp4.83 trillion a year earlier [26]. That is roughly 9% of TOWR's Rp44.6 trillion of bank loans and bonds [27]: the controller both owns the company and is a meaningful lender to it. In benign conditions that is a stable, captive source of credit; in a stress scenario it is a related party sitting on both sides of the table. Second, the remuneration of the Board of Directors is determined by PT Sapta Adhikari as controlling shareholder, which holds "the authority and power to determine the amount" [28]. Pay-setting sits with the controller rather than an independent committee — a governance concentration that the modest, declining compensation figure mitigates but does not remove.

The read

On the evidence, the alignment leg of the thesis holds up. The Hartono family controls 65% of TOWR, and when the balance sheet needed repair it wrote almost the entire cheque itself — at a price now a third above the market — rather than diluting minorities to strangers or leaning harder on debt. Capital allocation over the period was conservative in the way this reader wants: equity in to cut leverage, dividend trimmed to match, acquisitions kept small and adjacent, buyback token. The related-party dealings that come with a controlling family are real but, in the numbers, immaterial to the income statement — the exceptions being a Rp4 trillion loan from the family bank and controller-set director pay, both worth watching, neither large enough today to override the picture.

What would change this read: related-party balances or cost lines climbing from the current low-single-digit percentages; the BCA loan growing rather than shrinking as a share of debt; a return to a dividend funded by leverage rather than cash flow; or acquisitions that stop being small and adjacent. None of those is visible in the current filings. For an investor drawn to strong-promoter companies the market has left for dead, the controller here has behaved like an owner, not an extractor — which is the more reassuring answer to the question this report is built around.