Chapter 8

Figures converted from Indonesian rupiah at historical FX rates — see data/company.json.fx_rates. Balance-sheet and market figures use each year's own year-end rate (Rp16,667 per US dollar at 31 December 2025, Rp16,129 at 31 December 2024, Rp15,385 at 31 December 2023). Ratios, margins, multiples, percentages and share counts are unitless and unchanged.

Peer Discounts

Five Indonesian holding companies sit alongside Saratoga in this corpus, and at the end of 2025 they were priced between 0.36 and 8.31 times their own net assets. There is no local holding-company discount to appeal to. The one peer keeping its books on the same fair-value basis traded at 1.01 times net assets with a smaller free float and more debt, having just raised $224 million of equity at a premium to book.

Only one peer keeps the same books

Saratoga is a qualifying investment entity under PSAK 110, so its controlled entities, associates and joint ventures are measured at fair value through profit or loss instead of being consolidated [1]. Its reported equity is therefore a marked net asset value, which is why price-to-book and price-to-net-asset-value are nearly the same number here: $1,286 million of market capitalisation at the 31 December 2025 close of $0.0948 [2] against $3,533 million of equity attributable to owners [3] and $3,617 million of published net asset value [4].

Exactly one peer uses that accounting. PT Provident Investasi Bersama (PALM) states the policy in the same words Saratoga does — controlled entities, associates and joint ventures at fair value through profit or loss under PSAK 109, with the PSAK 110 investment-entity exemption [5]. The others do not. PT Indoritel Makmur Internasional (DNET) equity-accounts its three associates, carried at $810 million or 57% of consolidated assets [6]. PT Multipolar (MLPL) consolidates operating subsidiaries [7] and equity-accounts the rest [28]. Astra International (ASII) and Indika Energy (INDY) are operating groups that consolidate at historical cost. For those four, the multiple below is a price-to-book ratio, not a discount to marked value, and it can move for reasons that have nothing to do with what the assets are worth.

Sources: each company's own 2025 annual report at 31 December 2025 — Saratoga market capitalisation of $1,286 million [2] and equity [3]; Provident [8] and [9]; Multipolar [10] and [7]; Indoritel [11] and [12]; Astra [13] and [14]; Indika [15] and [16]. Indika reports in US dollars and its equity is shown as reported; every other figure is converted at the 31 December 2025 rate of Rp16,667 per dollar. Net assets are equity attributable to owners except for Indoritel, which reports the total.

The spread is the point. Two listed Indonesian investment holdings bracket the range [29]: Indoritel at 8.31 times its own equity, Multipolar at 0.36. Whatever a 60% discount is, it is not something the Indonesian market applies to holding companies as a class.

The like-for-like comparison

Provident is the only comparison that survives the accounting test, and it is a close one. It is an IDX-listed Indonesian investment entity whose 2025 gains came from Merdeka Copper Gold and Merdeka Battery Materials — the same Merdeka complex in which Saratoga holds 19.4% [4] — at $119.8 million, alongside $13.3 million from the logistics operator MMLP and $0.5 million from the telecom operator EXCL [17]. It is also a company Saratoga itself owned: a 19.9% stake carried at $57 million at a $0.0406 share price at the end of 2023, gone by the end of 2024 [18] after a $24 million negative fair-value change in the exit year [19].

Sources: Saratoga market capitalisation of $1,758 million (2024) and $1,286 million (2025) [2] against equity attributable to owners of $3,208 million and $3,533 million [3]; Provident closes of $0.0272 and $0.0228 on 15,773,797,158 and 15,732,874,458 shares [8] against equity attributable to owners of $254 million and $356 million [9].

Leverage does not explain the gap; neutralising it widens it. Provident's marked portfolio was $520 million at the end of 2025, against $193 million of bank loans and bonds and $21 million of cash [9]. Market capitalisation plus net debt comes to $530 million, or 102% of the marked portfolio. Saratoga's investments in shares and other securities stood at $3,639 million against $87 million of borrowings and $58 million of cash [3]; the same construction gives $1,315 million, or 36% of the marked portfolio. Adding Saratoga's $129 million of other liabilities, which are mostly deferred tax, takes it to 40%. On the marked dollar, the market pays 102 cents for Provident's portfolio and 36 to 40 cents for Saratoga's.

Float and turnover do not sort the multiples

The most common structural defence is that a closely held, thinly traded holding company cannot be priced properly. The peer set does not support it. Provident is more closely held than Saratoga, not less: four parties own 91.03% and the public holds 8.55% [20], against Saratoga's three principals at 89.005% and a public float of 10.995% [21].

Turnover runs the same way. Saratoga is the third-most-traded name in the set and carries the second-lowest multiple; Indoritel is the least traded by three orders of magnitude and carries the highest.

Sources: Provident, Indoritel and Multipolar report quarterly traded volumes directly [8] [11] [10]; Saratoga [2] and Astra [13] report average daily volume, annualised here at 61 trading days a quarter.

Indoritel traded 132,200 shares in the whole of 2025 — 7,600 in the first quarter — and closed the year at $0.5445 for a market capitalisation of $7,723 million [11]. That number is arithmetic, not a price at which anything could be sold, and it is the reason the 8.31x at the top of the table should not be read as a re-rating Saratoga could hope for. It does, however, dispose of the idea that illiquidity mechanically produces a discount: here it produced the opposite.

What is actually different about Provident

Four differences are visible in the filings, and they point in different directions.

Provident realises whole positions. During 2025 it sold its entire holding in EXCL in April and its entire holding in MMLP in September [17]. Saratoga's own Investment Milestones name no exit at all for 2025, as the Realisation Record sets out. Two complete exits in one year is the behaviour a discount is supposed to reward, and of the four differences it is the one most likely to be doing the work.

Provident can issue equity. Between 20 March and 4 April 2024 it raised $224.3 million through a rights issue of 8,654,256,802 shares at $0.0259 [22]. Its equity was $160 million at the end of 2023 [23] on the 7,119,540,356 shares that preceded the issue, so $0.0215 of book per share was sold at $0.0259 — a 21% premium. A company priced at net asset value can fund itself by issuing shares; one priced at a third of net asset value cannot. Saratoga's share listing chronology records two corporate actions in its whole listed life: the 2013 initial public offering and the 2021 stock split [26].

Provident retires stock. In May 2025 it cancelled 40,922,700 treasury shares, 0.26% of issued capital, through a capital reduction [24]. Saratoga's own treasury activity feeds its incentive plan rather than the share count, as Pay and Alignment sets out.

Provident has no distribution record to speak of. Its unappropriated retained earnings were negative $9.5 million at the end of 2024 and only turned positive, at $101.7 million, during 2025 [9], and Indonesian law permits a distribution only out of positive retained earnings, as its own dividend policy records [25]. Saratoga paid $0.0057 a share for FY2025 on 12 June 2026 [27]. Payout does not sort the two multiples either.

The strongest fact against reading Provident as the answer is what its own shareholders earned. Its shares closed 2023 at $0.0406, 2024 at $0.0272 and 2025 at $0.0228 [18] [8], while total equity ran $160 million, $254 million and $356 million on losses of $215 million and $123 million before 2025's $111 million profit [23]. The book grew because capital was raised and marks recovered; the share price fell 39% across the two years. Full net asset value is a rating, not a return, and Provident's rating sat above 1.0x through a period in which its holders lost money. Saratoga's own exit during 2024, from a position last marked at $0.0406 and counted in the Realisation Record as one of nine in thirteen years, looks better against that price path than against the equity line.

What would change the read

The evidence points one way on the narrow question. A 60% discount is not what the Indonesian market charges for the holding-company format, for concentrated ownership, or for a thin float — the closest-matched comparator has more of all three and trades at net asset value. That leaves conduct and portfolio as the candidate explanations, which is where the difference between two realisations in a year and none actually bites.

The read is bounded by what a single comparator can carry. Provident is a tenth of Saratoga's size, its portfolio is younger and narrower, and its multiple has been anywhere between 1.0x and 1.7x in two years — a range wide enough that the 1.01x at the end of 2025 could be coincidence rather than an equilibrium. A second fair-value-carried Indonesian investment entity trading near book would firm it considerably; none exists in this corpus. What would move the read the other way is Provident's multiple collapsing toward Saratoga's without any change in its conduct, which would suggest that the two ratings were never comparable. What would move it toward the discount being closable is Saratoga selling a whole position at or near its mark — the one behaviour Provident displayed in 2025 and Saratoga has not since 2024.