Sumber Alfaria Trijaya (Alfamart)

1. 01Open a small store very close to the customer

   The unit is a minimarket of a few hundred square metres placed within walking distance of housing, not a destination format. Alfamart went from 20,120 stores at end-2024 to 20,925 by the third quarter of 2025, adding about 805 outlets or 4.0 percent in a year.

   EconomicsEach store is cheap enough that capital spending is only 2.1 percent of revenue, so the growth engine is replication rather than a large project. Density is the moat: at this outlet count the competitive advantage is walking distance and replenishment logistics, not price.

2. 02Take delivery of stock you have not paid for

   Suppliers ship on credit averaging 53.3 days while the store holds inventory for 49.2 days and customers pay in cash. That ordering is the entire financial model.

   EconomicsBecause payables exceed inventory days, the supply chain funds the shelf. The cash conversion cycle was 2.2, 1.6, 4.9, 3.1 and 5.2 days across FY2021 to FY2025, which is effectively zero working capital.

3. 03Sell it for cash, at a very thin margin

   Gross margin is 21.9 percent and operating margin only 3.6 percent, because rent, wages and logistics consume most of the gross spread. Net margin was 2.7 percent in FY2025.

   EconomicsDo not read a thin margin as a weak business. Cash collection is immediate and stock turns roughly seven times a year, so the same rupiah of capital earns that thin margin many times over.

4. 04Recycle the float into the next store

   Because growth releases supplier credit rather than absorbing cash, new stores are largely self-funding. Free cash flow was positive in every one of the five years even while roughly 800 outlets a year were added.

   EconomicsThat is the compounding loop, and it is why debt to equity fell from 32.5 percent to 14.0 percent while the store count kept rising. Expansion here strengthens the balance sheet instead of straining it.

Cost structureAlmost entirely cost of goods, then almost entirely fixed store costs. Cost of revenue was IDR 99.0tn of IDR 126.7tn in FY2025, giving a 21.9 percent gross margin that has been remarkably steady between 20.7 and 21.9 percent for five years. Operating margin of 3.6 percent means about 18 points of that gross spread goes to rent, wages, utilities and distribution. That leaves very little tolerance for error: a one point move in gross margin swings operating profit by roughly a quarter. Interest is negligible at IDR 137.3bn against IDR 4.5tn of operating profit, cover of 33 times.

Cash cycleThe shortest cycle in the roster and the reason the model works: 2.2, 1.6, 4.9, 3.1 and 5.2 days across FY2021 to FY2025. The composition is the point. Inventory sits 49.2 days, receivables are only 9.3 days because most sales are cash, and payables run 53.3 days, so suppliers finance more than the entire inventory position. Compare that with HMSP on this site, which funds 85 days of leaf itself at a 74 day cycle. Same country, same consumer, opposite working-capital direction, and it explains why one business can grow without capital and the other cannot.

Unit economicsThis page is the mirror image of BREN, and reading the two together teaches the whole identity. BREN earned a 71 percent operating margin and produced single-digit returns because assets turned only 0.16 times. AMRT earns a 2.7 percent NET margin and produces an 18.9 percent return on equity because assets turn 2.98 times and the cash cycle is five days. Return on invested capital of 18.1 percent sits close to return on equity, so leverage is not doing the work either. Margin alone tells you nothing about a business. Margin multiplied by turnover, adjusted for how much capital the cycle ties up, tells you almost everything, and it is why a shop selling instant noodles at a two percent margin can compound faster than a power plant selling electricity at seventy.

Among the cleanest records on the site, and the cash proves it rather than the commentary. Operating cash flow covered net income 3.29, 2.47, 2.00, 2.56 and 2.25 times across FY2021 to FY2025, never once below twice, which is the signature of a business collecting in cash and paying suppliers later. Free cash flow was POSITIVE in all five years at IDR 4.5tn, 4.9tn, 4.3tn, 5.3tn and 5.0tn while revenue grew from IDR 84.9tn to IDR 126.7tn, so nearly fifty percent revenue growth was delivered without consuming cash. There are no acquisition gains, no revaluations, and minority interests are small at 7.0 percent of equity, so reported profit is close to owners profit. The one honest caution is thinness rather than quality: at a 3.6 percent operating margin, small movements in gross margin or store-level costs move profit a long way, and returns have already drifted down from 25.4 percent return on equity in FY2022 to 18.9 percent in FY2025 even as the store count rose. That drift, not any accounting question, is what to monitor.

Cash conversionOne of the strongest structural conversions in the roster. Operating cash flow exceeded net income by at least two times in every year, cumulative five-year free cash flow was about IDR 24.0tn, and the group moved to a net cash position while opening roughly 800 stores a year. The mechanism is not efficiency, it is sequencing: cash arrives at the till before the supplier invoice falls due, so scale generates float. Judge this business on store additions, same-store sales and gross margin stability, because the cash statement will look healthy almost regardless.

A builder that does not need outside money, which is rare enough to be the point. Revenue grew 49.3 percent over five years and the store base expanded every year, yet debt to equity FELL from 32.5 percent to 14.0 percent and the group finished FY2025 holding net cash of about IDR 2.0tn. Equity attributable to owners nearly doubled from IDR 9.2tn to IDR 18.0tn on retained earnings rather than issuance. Compare PANI on this site, where roughly IDR 29tn of equity growth came from shareholders rather than profit: AMRT is the opposite case, and the contrast is the cleanest illustration of what self-funded growth means.

DeploymentCapital goes almost entirely into new stores and the distribution centres that serve them, at 2.1 to 2.4 percent of revenue every year, which is IDR 1.8tn rising to IDR 2.7tn. Working capital absorbed almost nothing because the cycle is five days. Everything else went into de-levering: debt fell from IDR 3.1tn to IDR 1.9tn by FY2024 before rising modestly to IDR 2.7tn, while cash grew from IDR 3.3tn to IDR 4.7tn. Inventory rose from IDR 8.8tn to IDR 13.3tn, which is the real cost of adding outlets, and it was funded by supplier credit rather than borrowing. The allocation record is therefore simple and consistent: reinvest at roughly 2 percent of sales, hold no meaningful debt, and let the float pay for growth.

Returns trendReturns are high but trending the wrong way, and that tension is the investment case. Return on equity ran 20.9, 25.4, 23.5, 19.2 and 18.9 percent, with return on invested capital close behind at 23.2, 26.2, 22.5, 18.4 and 18.1 percent, so the fade is operational rather than financial. Operating margin peaked at 4.1 percent in FY2023 and sits at 3.6 percent, which on a base this thin is a meaningful compression. The market is nonetheless paying up. At IDR 1,325.14 the market capitalisation of about IDR 54.5tn puts AMRT on an EV/EBITDA of 8.27 times against a 6.25 times retail peer median, a 32 percent PREMIUM, while the forward model produces IDR 974.92 per share, 26 percent BELOW the traded price. Note the contrast with INDF on this site, which trades at a 29 percent discount to its peers on the same consumer economy: the market is paying a premium for a compounding, self-funding store network and a discount for a leveraged holding structure. Whether the premium survives depends on whether new outlets keep earning the old returns, and the drift from 25.4 to 18.9 percent is the early evidence that they may not. Both figures are model outputs, not targets.