Chapter 8

Backlog Run-Off

At 31 March 2026 CBDK held Rp9,081.7 billion of customer money against product not yet handed over [1] — about 3.1 years of revenue at the March-quarter rate, against 4.5 years at the end of 2023. New orders replaced 17% of the revenue recognised in 2025, against 94% in 2024 [2]. Stripping net cash and the profit still to come out of that pool leaves a residual that values everything unsold at 0.8 to 1.1 times its historic cost.

Sources: advances [3]; order intake [4]; cash and bank loans [5] [6]; coverage and replacement rate derived.

What the pool has to cover

CBDK sells before it builds. Cash arrives on contract, sits on the balance sheet as advances from customers, and becomes revenue only when the buyer takes handover [7]. That pool stood at Rp9,081.7 billion at 31 March 2026: Rp8,476.6 billion owed to third-party buyers, Rp585.1 billion to related parties, and Rp19.9 billion of rental prepayments and unallocated deposits [8].

The pool is the company's stock of already-funded revenue. Set against the revenue actually recognised, it has been shrinking as a multiple for two years.

Sources: advances from customers at 31 December 2023 [9], at 31 December 2024 and 2025 [10], and at 31 March 2026 [11]; FY2023 revenue [12]; FY2024 and FY2025 revenue [13]; Q1 2026 revenue [14]; order intake [15]. Coverage years are derived.

Two things moved at once. Revenue rose 28% across the three years, from Rp1,953.5 billion to Rp2,503.6 billion [16] [17], and the pool stopped growing: up Rp1,238.0 billion in 2024, down Rp389.7 billion in 2025, down a further Rp465.2 billion in the March 2026 quarter alone [18] [19].

The March quarter shows the mechanism cleanly. CBDK recognised Rp742.6 billion of revenue [20] while collecting Rp278.8 billion from customers [21] — a gap of Rp463.8 billion against a Rp465.2 billion fall in the pool. The two figures agree to within 0.3%, which is what makes the pool usable as an arithmetic quantity rather than a metaphor: at CBDK, revenue is very nearly the pool being drawn down.

The replacement rate

Order intake — what the company calls marketing sales — is the only thing that refills the pool. Against revenue recognised, it collapsed in 2025.

Sources: order intake 2023–2025 [22]; 2026 intake target of Rp563,054 million [23]; revenue [24] [25]. The FY2026 revenue bar is the single published analyst estimate, not company guidance.

In 2023 the company sold Rp1.15 of new product for every Rp1.00 it handed over; in 2024, Rp0.94; in 2025, Rp0.17 [26]. Management's own 2026 target of Rp563,054 million does not aim at replacement either [27]: it is 20% of the only published revenue estimate for the year. On the company's own plan, the pool keeps draining.

CBDK does not disclose quarterly intake, so there is no mark on 2026 between the December year end and the half-year statements. It is the largest gap in this arithmetic, and it closes shortly.

What is already paid for

The useful consequence of a pre-sold model is that a large slice of future profit is not a forecast — it is a delivery obligation against cash already banked. Rp9,061.7 billion of the pool is sales advances on land plots, houses, shop-offices and warehouses [28]. Handing that product over converts it to revenue at whatever margin the mix carries.

Two margins bracket the plausible range. The FY2025 real estate segment earned 67.4% on Rp2,464.2 billion of revenue against Rp803.7 billion of released inventory cost [29]. The March 2026 quarter earned 88.2% on Rp736.8 billion against Rp86.8 billion [30] — a released cost that reads as bare land, the highest-margin product in the book. FY2025's mix is the more representative case; the March quarter is an upper bound.

Sources: advances [31]; segment margins [32] [33]; below-gross costs and the parent share held at FY2025 ratios [34] [35]. Derived arithmetic, undiscounted.

Selling and administrative expense, final tax and income tax together consumed 15.4% of FY2025 revenue — Rp221.8 billion, Rp114.3 billion and Rp49.1 billion respectively [36]. Parent shareholders took 93.0% of group profit in FY2025 [37] — a conservative anchor, since the March quarter's split was 99.1%. On those ratios, delivering what has already been paid for leaves Rp4,383 billion to Rp6,134 billion for CBDK's own shareholders, spread over roughly three to four years and not discounted here.

Setting that against the market value gives a decomposition rather than a valuation:

Sources: shares outstanding of 5,657,686,900 after 11,257,600 held in treasury [38]; cash and bank loans [39]; assets [40]. Share price as reported on the Indonesia Stock Exchange; the rest is derived.

Net cash is cash and equivalents of Rp2,745.5 billion less bank loans of Rp233.8 billion [41] [42]. The remaining book is what survives the run-off: inventories of Rp13,977.6 billion less the cost released, plus investment property of Rp2,529.7 billion, fixed assets of Rp497.2 billion, other investments of Rp1,509.8 billion and prepaid taxes of Rp1,019.7 billion, plus Rp151.8 billion of time deposits, receivables and prepayments, less Rp648 billion of payables, accruals and employee benefit obligations and less Rp2,961.3 billion attributable to non-controlling interests [43] [44] [45].

The residual divided by the remaining book gives 1.10 times at the FY2025 margin and 0.84 times at the March-quarter margin. On the stated assumptions, then, once cash and pre-sold profit are taken out, the market pays roughly one times historic cost for CBDK's undeveloped land, its convention centre and its hotel. That is a narrower premium than the headline price-to-book multiple of 2.24 times, which measures the Rp21,272.9 billion market value against Rp9,506.5 billion of parent equity [46], and it sits alongside rather than replaces the per-square-metre work in Land at Cost. Three caveats belong with it. The run-off profit is undiscounted and takes years to arrive. Below-gross costs are held at FY2025's ratio, when general and administrative expense would not fall proportionately in a genuine wind-down. And historic cost is a cost, not a market value: nothing in the corpus marks this land to a transaction.

Three intake paths

Over a long enough horizon revenue cannot exceed intake. The table below runs CBDK's own FY2025 cost structure at three sustained intake levels: gross margin at 67.4%, selling expense at 2.2% of revenue, general and administrative expense fixed at Rp166.3 billion, associate losses at Rp27.7 billion, net finance income at Rp228.3 billion, final tax at 4.6% of revenue and income tax at Rp49.1 billion [47]. Applied to FY2025's actual revenue the model returns Rp1,503 billion against Rp1,467 billion reported — 2.4% high, which is the accuracy the table carries.

Source: derived from the FY2025 cost structure [48] and FY2025 reported EPS of Rp241.45 [49]; intake levels [50] [51].

The price is most sensitive to which of those paths holds. At intake sustained near where it ran in 2025 and where management has targeted 2026, the earnings currently being capitalised are roughly five times the earnings the order book would support, and the multiple on that steady state is between 70 and 93 times. At intake back near the 2023–24 average, current earnings are close to sustainable and the multiple is 17.5 times. The pool buys three to four years to find out which it is.

Two facts cut hard against reading the low path as the likely one. First, intake at CBDK is lumpy by construction: commercial land plots were Rp1,488.9 billion of 2024's Rp2,114.6 billion, and Rp343.5 billion of 2025's Rp430.2 billion [52]. A single corporate buyer can restore a year's intake — the two customer sales recognised in the March 2026 quarter alone produced Rp517.7 billion of revenue [53], more than the whole of 2025's order book (Customer Concentration). Second, the pool is not the asset. Delivering everything already collected releases only about Rp2,954 billion of the Rp13,977.6 billion inventory book at FY2025 margins — roughly a fifth [54]. Rp6,987.7 billion of undeveloped land and Rp5,494.1 billion of land under development would still be there [55]. A land bank does not have to sell on schedule; it has to sell eventually.

The read the evidence supports is that the low path is a floor on reported earnings power rather than a forecast, and that the residual multiple of 0.84 to 1.10 times cost is the more durable anchor than any near-term P/E. What would move it: an intake print near or above Rp1.0 trillion for the first half of 2026 would put the replacement rate back in a range where current revenue is defensible; a first half near Rp0.2 trillion would confirm the Rp563 billion full-year target as the ceiling rather than the base.

Where the balance sheet can and cannot break

CBDK's liabilities are almost entirely a promise to build, not a promise to repay. Of Rp9,963.5 billion of total liabilities at 31 March 2026, Rp9,081.7 billion — 91.2% — is advances from customers, and bank loans are Rp233.8 billion against Rp2,745.5 billion of cash [56] [57]. Trade payables, other payables, taxes, accruals and employee benefit obligations total about Rp648 billion — a quarter of the cash balance. On the filed record there is no maturity wall and no lender able to force the issue; the solvency detail sits in Financials and Estimates.

The constraint is slower and runs through cash. Free cash flow — operating cash flow less purchases of fixed assets and investment properties — was minus Rp592.0 billion in FY2025, on Rp719.4 billion of operating cash flow against Rp1,311.4 billion of capital spending [58], and minus Rp257.0 billion in the March 2026 quarter, on minus Rp130.9 billion of operating cash flow and Rp126.1 billion of capital spending [59]. At the FY2025 rate the cash balance funds about four and a half more years; at the March-quarter rate, about two and a half. Neither is a near-term insolvency risk, and the repurchase of 11,257,600 shares for Rp61.3 billion in the March quarter, under a programme running from 3 February to 2 May 2026 [60], indicates management does not treat the balance sheet as tight. The burn is nonetheless real, and it runs against a pool that is no longer being refilled.

What the next filings settle

The half-year 2026 statements are due at the Indonesia Stock Exchange at the end of July 2026, with the earnings calendar carrying 29 July 2026. Four line items in that filing are checkable against thresholds this chapter has set.

The fourth of those is the cleanest test in the set, because the amount is exact: Rp268,827,273 thousand is the figure the March 2026 note attributes to PT Global Jet Express [61], and the same figure appears in the nine-month 2025 note but not in the audited FY2025 one (Customer Concentration).

Beyond that filing, the FY2026 annual report — due around April 2027 — carries the target-versus-realisation table for 2026 and the 2027 target, which is where management's own view of the replacement rate becomes visible again.

Three things this chapter could not resolve. There is no external mark on the land: no appraisal, transacted price per square metre or broker valuation for the CBD PIK2 district exists anywhere in the corpus, so the remaining book is priced against its own cost. There is no buyer-level breakdown of the Rp9,081.7 billion pool beyond the Rp585.1 billion owed to related parties [62], so the concentration of the run-off is unknown. And no 2026 broker commentary or half-year press reporting was available to bring against the filed record.