Chapter 1

Figures converted from Indonesian rupiah at historical FX rates — see data/company.json.fx_rates. Ratios, margins, and multiples are unitless and unchanged.

What Saratoga is

PT Saratoga Investama Sedaya is a Jakarta holding company that owns minority and controlling stakes in a concentrated set of Indonesian businesses and reports its own worth as the market value of those stakes less net debt. At the end of 2025 that figure was $3,617 million, or $0.267 per share [1]. The shares closed at $0.097 on 27 July 2026 [2]. The distance between those two numbers is what this report is about.

Saratoga describes itself as an active investment company, founded in 1997 and listed on the Indonesia Stock Exchange under the code SRTG [3]. It came to market on 26 June 2013 at $0.385 a share, split the stock five-for-one in May 2021, and has had 13,564,835,000 shares outstanding since [4]. That IPO price is $0.077 in today's shares, so thirteen years of listed life have produced a 60% price gain plus modest dividends. It employs 67 people [5]. There is no operating business at the holding company: it buys stakes, sits on boards, collects dividends and occasionally sells. Its own annual report says plainly that marketing is a matter for the investee companies, not the parent [6].

The company publishes its net asset value on a fixed arithmetic each year: the sum of investee company values, minus debt, plus cash [7]. For 2025 that was $3,646 million of investee value, less $87 million of debt, plus $58 million of cash, giving $3,617 million — up 12% on 2024's $3,347 million [8].

Sources: NAV per share from the FY2025 Annual Report NAV bridge [9]; closing price on 27 July 2026 [10]; market cap and discount derived from that price and the 13,564,835,000 shares outstanding [11].

What it owns

Four listed holdings account for three-quarters of the portfolio. Tower Bersama, a telecom tower operator with more than 24,300 sites, is the largest position at 30.7% of portfolio value [12]. Merdeka Copper Gold is a gold, copper and nickel producer; Alamtri Resources and Adaro Andalan are the two halves of the former Adaro coal group, separated in late 2024 into a minerals-and-renewables company and a standalone thermal coal producer [13].

Source: FY2025 Annual Report, Net Asset Valuation as of 31 December 2025 and 2024; portfolio shares derived from the same table [14].

Two things about that table matter for everything downstream. The first is concentration: TBIG, MDKA, ADRO and AADI together are $2,789 million, or 76.5% of investee value. The second is what sits underneath the four names — towers, gold and copper, metallurgical coal and minerals, thermal coal. Roughly 46% of the portfolio is priced off commodities, and coal in particular still funds much of the cash that reaches Jakarta.

The rest is a mix of a majority-owned automotive group (Mitra Pinasthika Mustika, 57.7%), industrial gas, construction, and a set of private businesses the company has been scaling: Brawijaya Healthcare, six hospitals and a clinic, which opened Brawijaya Taman Mini in 2025 [15]; ZAP, 118 aesthetic clinics [16]; the solar developer Xurya [17]; and the data-centre platform Digital Realty Bersama [18].

Where the money comes from

Saratoga's income statement is two very different things stapled together, and a reader who treats reported profit as earnings will be badly misled.

Sources: FY2025 Annual Report, Consolidated Statements of Profit or Loss for 2025 and 2024 [19]; FY2023 Annual Report for 2023 and 2022 [20]; FY2022 Annual Report for 2021 [21].

The blue bars are unrealised revaluation of the stakes. They swing from a $1,709 million gain in 2021 to an $898 million loss in 2023 and back to a $248 million gain in 2025 [22][23]. They are a restatement of the same NAV that appears on the balance sheet, not income in any cash sense. Reported profit follows them: a $660 million loss in 2023, a $204 million profit in 2024, a $439 million profit in 2025 [24]. Earnings per share ran from minus $0.049 to $0.015 to $0.032 over the same three years [25]. Any multiple built on those numbers is arithmetic performed on a price index.

The orange bars are the cash business, and it is small, steady and legible. Dividend income was $181 million in 2023, $235 million in 2024 and $162 million in 2025, of which ADRO and AADI supplied $129 million in 2025, TBIG $14 million and MPMX $18 million [26]. Against that, the whole holding company costs $14 million a year to run and $10 million to finance [27]. Operating cost is 0.4% of NAV [28]. On a fee-comparison basis that is cheap for a managed pool of assets, and the cost line has barely moved in three years while NAV rose 23%.

The dependence on coal dividends is the obvious fragility. In 2025 the two Adaro entities were 79% of dividends received, and dividend income fell 29% year on year as thermal coal prices normalised, even as NAV rose 12% [29].

What shareholders have received

For three years Saratoga kept almost all of the cash it collected. That changed with the 2025 distribution.

Sources: dividends income by investee, FY2025 Annual Report [30]; distributions for FY2023 and FY2024, FY2025 Annual Report dividend table [31]; FY2025 distribution derived from the $0.00568 per share reported paid on 12 June 2026 and shares outstanding [32].

The distributions for FY2023 and FY2024 were $0.00136 and $0.00089 a share, $18 million and $12 million in total — 11% and 5% of the dividends the company itself had received [33]. For FY2025 the board deferred the amount to the general meeting of 18 May 2026 [34], and Indonesian financial press reported a payment of $0.00568 per share on 12 June 2026 [35]. That is roughly $77 million, seven times the prior year and about half the dividends received. At $0.097 it is a 5.9% yield. The corpus contains no company document stating the FY2025 per-share amount, so that figure rests on press reporting rather than a filing — worth confirming against the eventual FY2026 annual report.

Price against value

The company's NAV per share has been roughly flat for three years while the share price has not.

Sources: NAV per share from each year's Net Asset Valuation table — FY2022 Annual Report for 2021 and 2022 [36], FY2023 Annual Report for 2023 [37], FY2024 Annual Report for 2024 [38], FY2025 Annual Report for 2025 [39]; year-end closing prices from the FY2022 Annual Report [40], FY2023 Annual Report [41] and the FY2025 Annual Report quarterly closing price table [42].

NAV per share was $0.292 at the end of 2022 and $0.267 at the end of 2025 — down 1% across three years in rupiah terms, having fallen 20% in 2023 and recovered since [43][44]. The share price over the same window went from $0.164 to $0.095, and the intraday high of $0.250 set in the second quarter of 2022 is 54% above the July 2026 price in rupiah terms [45][46]. The discount to published NAV widened from 33% at the end of 2021 to 64% at the end of 2025, with a partial narrowing in 2024.

At $0.097 the market values the whole company at about $1.31 billion. The TBIG and MDKA stakes alone were marked at $1,768 million on 31 December 2025 [47]. On those marks the market capitalisation is 19% below the value of two holdings, with ADRO, AADI, MPMX, the industrial gas and construction stakes, the hospitals, the clinics, the solar platform and the data centres carried at less than nothing. The comparison mixes a July 2026 price with December 2025 marks, and the underlying stakes have moved since — the first-quarter 2026 statement shows investments in shares up 12% to $3,800 million [48]. Applying the company's own NAV arithmetic to that statement gives roughly $4.07 billion, near $0.300 a share, and a discount closer to 65%.

Who owns it

Three parties hold 89% of the shares. Edwin Soeryadjaya, the president commissioner and co-founder, held 35.87% directly at the end of 2025; PT Unitras Pertama held 31.62%; Sandiaga Salahuddin Uno, the other co-founder, held 21.51%. The public float — every holder below 5%, some 22,383 of them — is 10.995% [49]. The three executive directors hold a further 0.15% between them, and each added shares during 2025 [50]; all three bought again on 1 July 2026 at $0.086 under the long-term incentive programme [51].

Concentration of that degree cuts both ways and the report should hold both edges. Owners with 89% of the equity bear the discount alongside the minority, which is the strongest structural argument that management's incentives point at NAV per share rather than at size. The same 89% leaves an 11% float in a market where average daily volume ran at 6 to 13 million shares a quarter through 2025 [52] — roughly $0.6 million to $1.1 million a day at current prices. An institution cannot build a position of any size, and the discount is partly the price of that fact. Coverage is correspondingly thin: two brokers, one published earnings estimate, and no revenue consensus at all [53]. The company holds no earnings calls; the corpus contains no transcript for any period, because none exists.

How solid the balance sheet is

For an investor whose first question about a leveraged-looking holding company is whether it can go bankrupt, the answer here is unusually clean. Borrowings were $87 million at the end of 2025 against $58 million of cash, giving net debt of $29 million and a loan-to-value ratio of 0.8%, down from 3.0% a year earlier [54]. Total liabilities of $216 million sit against $3,751 million of assets, and $123 million of those liabilities are deferred tax rather than anything payable to a lender [55][56]. The loan maturity ladder runs $15 million in 2026 and $46 million as late as 2029, against annual dividend receipts of $151 million to $263 million [57][58]. By the end of March 2026 borrowings had fallen further to $48 million against $62 million of cash and time deposits — a net cash position [59].

The financing risk in this structure is not the holding company's own leverage. It is whatever debt sits inside the investees, which does not appear on this balance sheet at all — a question the report should take up separately.

How much of the NAV is checkable

The NAV is only as good as the marks behind it, and the marks are not uniform. Of $3,405 million of investments in shares at the end of 2025, $1,567 million was Level 1 — quoted prices in active markets — $1,529 million Level 2 and $309 million Level 3 [60].

That looks worse than it is, and the auditor's own language explains why. KPMG's Indonesian member firm flagged the valuation of Level 2 and 3 investments as a key audit matter, noting they totalled $2,073 million, or 55.26% of consolidated assets — but also that $1,529 million of the Level 2 balance consists of investments in entities that hold a direct ownership in publicly traded shares [61]. Those are intermediate holding vehicles whose underlying assets carry observable prices. Netting them out, about $3.10 billion of the $3.65 billion portfolio — roughly 85% — traces to listed quotes either directly or one level down, and about $0.55 billion, some 15%, rests on discounted cash flow, comparable-company multiples or recent acquisition cost [62][63]. The auditor is explicit that the remainder is inherently subjective, uses unobservable inputs, and that holding investments through other investing entities further limits price transparency [64].

Two adjustments are worth making before treating that NAV as a target. The published NAV takes no deduction for the $123 million deferred tax liability that would crystallise on realising the gains; charging it in full takes NAV per share from $0.267 to about $0.258 and the discount from 60% to 59% [65]. And the marks are last traded prices applied to stakes of 15% to 32%; the exit price for a block of that size is not the screen price. Neither adjustment comes close to accounting for a 60% gap.

The question this report is built to answer

Saratoga is a founder-controlled Indonesian holding company whose shares trade around 60% below the market value of the stakes it owns, and this report exists to establish whether that gap is a permanent feature of an 89%-held, commodity-weighted portfolio, or a discount that the company's own cash flows, disclosure and record of monetising assets can narrow.

The evidence assembled here points to a real asset base rather than an accounting one: 85% of the portfolio is anchored to observable listed prices, the holding company runs on 0.4% of NAV a year, net debt is effectively zero, and management owns 89% of what is being discounted. The strongest fact on the other side is that the discount has widened, not narrowed, in three of the last four years while all of those conditions already held — which says the market is pricing something the balance sheet does not capture, whether that is float, governance, the coal concentration in the dividend stream, or simply the absence of a mechanism that converts NAV into cash for a minority holder. The FY2025 distribution of roughly half of dividends received, if repeated, is the first evidence in three years of such a mechanism. A second year at that payout, or a large realisation returned to shareholders, would change the read; a reversion to a 5% payout would confirm the bearish one.

What this chapter does not settle

Forward estimates are outside what the filings support: two brokers cover the stock, one publishes an earnings number, and no revenue consensus exists [66]. For a company whose reported earnings are a mark-to-market artefact, a consensus EPS is close to meaningless in any case; the forward question is dividends and NAV growth, and it needs its own treatment. The debt and prospects of the investee companies, the quality of governance around related-party dealings, management's compensation, and the track record of actual exits — each is a chapter's worth of evidence, and none is settled by what is above.