Scenarios and Watch Items

Scenarios and Watch Items

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

The earlier chapters established the pieces separately — the fragile balance sheet (Cash and Solvency), the aligned but unchecked controller (Alignment and Control), the flat end-market and unproven new line (Demand and Durability), the full-to-premium price (Margin of Safety), and the single government order behind the growth (One Government Customer). Reconciled, they resolve into one correlated risk rather than three independent ones. The bear case and the refinancing wall are one event: a FY2026 in which the Badan Gizi order does not recur is also the year the $2.53m inventory does not convert, even as ~$1.45m of BCA debt matures against $0.17m of cash. [1][2][3][4] The order at risk is about 16.8% of FY2025 revenue; the $2.53m finished-goods position is roughly 23% of equity and carries no obsolescence provision; and cash covers the maturing debt only about 12%. The mitigant sits in the same treatment: the founder personally guarantees the full BCA facility on family land [5], and BCA has rolled and expanded these lines before, so the correlated risk is bounded by the controller's own wealth rather than by net cash. PART's next-year earnings turn mostly on two measurable things: whether the $3.7m government food-tray order recurs, and whether reported profit converts to cash again. Around those pivots the plausible range runs from a low-double-digit multiple on a shrinking base to roughly 7x on a scaling franchise. The binding constraint throughout is a single-lender refinancing wall, not the income statement.

The two pivots that move the outcome

The case is most sensitive to two things the FY2026 filings will answer.

The first is the metal-household line. It went from zero to $3.73m in FY2025 and supplied 61% of the year's revenue growth [6], and Note 24 shows the entire segment is one buyer — Kantor Pusat Badan Gizi Nasional, the National Nutrition Agency behind the Makan Bergizi Gratis free-meals programme — whose $3,726,629 order matches the segment total to the dollar [7]. With automotive customer PT Chemco ($2.49m), the top two accounts were 28% of FY2025 sales [8]. Whether that order was a one-time kitchen fit-out or the first year of a recurring supply relationship is genuinely two-sided, and it swings roughly a fifth of revenue.

The second is cash conversion. FY2025 turned $1.81m of reported profit [9] into only $0.93m of operating cash — a 0.51x conversion [10] — because finished-goods inventory rose to $2.53m [11]. If that stock is staged deliveries, it converts in FY2026 and cash normalises; if it is speculative production for an order that does not repeat, it is where a write-down would land. The two pivots are linked: the same government order drives both the revenue and the inventory.

Base, bull and bear

The three paths below are illustrative, not forecasts — PART has no analyst coverage and publishes no guidance, so these are built from the FY2025 segment structure and stated assumptions, holding the share count and the $0.0059 price fixed. The scenarios differ almost entirely in one input: the household line.

Source: analyst scenarios derived from FY2025 segment disclosures [12] and the profit and loss statement [13]; assumptions stated below.

Source: analyst scenarios; base figures from FY2025 segment note [14] and profit and loss [15]. P/E computed at the 17 Jul 2026 price of $0.0059.

The bear path assumes the fit-out was one-time: the household line all but disappears as the initial kitchen equipment is delivered, automotive grows a modest 8%, and the loss of high-utilisation government volume compresses margin toward the run-rate the first post-order quarter implied. Revenue falls about 9% to roughly $20.2m and profit to around $1.38m, so the same $0.0059 price is a ~12.6x multiple on a shrinking base — the opposite of a value entry. This path is not a tail: an interim data point after the order shipped annualised earnings well below the FY2025 figure, consistent with mix reverting once the burst is gone (one unaudited quarter, outside the filing corpus).

The base path assumes partial recurrence — tray replacement and continued kitchen additions at roughly half the FY2025 rate — with automotive up 10%. Revenue holds near $22.1m and profit near $1.74m, leaving the stock around 10x. Here cash is the story that improves: if the $2.53m finished-goods position converts and capex steps down from the $4.32m FY2025 build toward the $1.08m depreciation run-rate, free cash flow moves toward breakeven for the first time in three years.

The bull path assumes the household line is a franchise, not a fit-out: the free-meals kitchen network keeps scaling nationally, PART wins repeat and expanding volume to roughly $5.4m, and automotive grows 15%. Revenue reaches about $26.6m and profit about $2.40m, putting the stock near 7x with cash conversion normalising as growth capex tapers. This is the path that would retroactively justify treating FY2025 as a base rather than a peak.

The scenarios are most sensitive to the single household input; the automotive assumptions move the answer far less. That is the concentration risk restated as arithmetic.

The constraint that outranks the income statement

For a reader whose first requirement is a near-zero chance of bankruptcy, the scenarios above are secondary to a balance-sheet fact. PART carried $7.40m of bank debt at end-2025, all with one lender, Bank Central Asia, and roughly $1.45m of it — a $0.14m short-term line plus $1.31m of current maturities — falls due within a year [16]. Against that sits $0.17m of cash [17] — about 12% cover — and $11.1m of equity [18].

Source: FY2025 Annual Report — bank loans [19] and financial position [20]; ND/EBITDA computed on full bank debt.

Two facts pull in opposite directions. Against the fragility: the founder personally guarantees the full BCA facility and family land secures it [21], so the controller's own wealth is exposed to a default and strongly motivates timely refinancing, and headline leverage at about 2.0x net-debt-to-EBITDA is not extreme. For the fragility: cash covers only a fraction of what rolls each year, the company has been free-cash-flow negative for three straight years, and a January 2026 shareholder meeting authorised the directors to pledge substantially all company assets and act as guarantor [22]. The honest read is that default risk is low but not negligible, and it is a refinancing risk, not a leverage-ratio risk: solvency depends on BCA continuing to roll roughly $1.45m a year while the business is not self-funding. The refinancing wall and the bear scenario are the same FY2026 outcome, set out at the top of the chapter.

What to watch

Each item below is a specific line in a specific future filing, with the threshold that would move the read. The first two settle the pivots; the rest bound the balance-sheet risk.

Source: watch items defined by the analyst against FY2025 disclosures; each resolves in the FY2026 interim and annual filings.

Reconciling against a value lens

Set against a buyer who wants a wide margin of safety, a founder with real skin in the game, and near-zero bankruptcy risk, the pieces land unevenly. The alignment test passes cleanly: the founder guarantees the debt with personal assets and pay is modest. The margin-of-safety test does not clear as easily — at $0.0059 the stock sits at the top of its listed peer range rather than at a discount, so a buyer is paying a full peer price for earnings that lean on an unproven order. And the bankruptcy test is the one that most resists a confident answer: the business is solvent on ratios but thin on cash and dependent on one bank's willingness to keep lending.

The evidence points to a company whose FY2025 result was flattered by a single government order and whose safety rests on refinancing rather than self-funding; the read that would overturn that is a FY2026 in which the household line recurs at scale and cash conversion returns to 1.0x. Both are checkable within a year, in the filings named above, which is the useful place to leave it.