NVR: Controlling Finished Lots Through the Housing Cycle
Download case PDF (3 pages)Request teacher’s pack
Decision date and information cutoff: March 1, 2024
Decision scope: How to sequence finished-lot control, land development and homebuilding activity across markets
Case convention: This is a hypothetical strategy exercise. NVR’s disclosures are historical; no private deliberations are represented.
The decision
On March 1, 2024, NVR’s strategy team is preparing to review the company’s land and community approach. NVR generally buys finished lots from third-party developers under fixed-price lot purchase agreements (LPAs), rather than directly owning and developing land. It also has limited joint-venture (JV) and directly owned land positions. As mortgage rates and affordability pressure housing demand, the team must consider whether to keep the LPA-centered model, selectively expand JV or direct development, or pace lot takedowns and home starts community by community.
This decision is hypothetical, not a disclosed management mandate. The public record does not show a fixed capital pool, a target land mix, or a return hurdle. Students should not assume NVR must change its strategy or that one land-control form has a higher return.
NVR’s land-control model
NVR says its land strategy is designed to avoid the financial requirements and risks of direct land ownership and development, support inventory turnover and use less capital. It typically acquires finished lots from third-party developers through fixed-price LPAs. Deposits, often up to 10% of the aggregate finished-lot purchase price, may be forfeited if NVR elects not to perform. The company states that it expects substantially all finished-lot inventory to remain LPA-based, while preserving limited strategic exceptions for JVs or direct acquisition of already-zoned land. These are management’s stated rationale and policy, not independent evidence of causal returns. [P1]
At December 31, 2023, NVR reported 141,500 total lots controlled. This total included 134,900 lots under third-party LPAs, about 4,850 controlled through four JVs and about 1,750 lots from directly owned land under development. The LPA count included roughly 10,700 lots associated with a $53.4 million impairment reserve. Cash deposits and letters of credit associated with LPA lots were $617.0 million and $7.7 million, respectively. These categories differ: a controlled lot is not necessarily owned, finished, or ready for a home start. [P1]
Separately, NVR reported 22,700 expected lots under raw-land contracts, excluded from the 141,500 controlled-lot total. Some parcels could require rezoning or approvals. NVR generally expected to assign feasible raw-land contracts to developers and enter an LPA. [P1]
Future deposit exposure and community decisions
Under existing LPAs, NVR disclosed $391.3 million of additional deposits that would become due if specified developer milestones were met and NVR exercised its options; most were expected within three years. This is conditional on milestones and exercise. It is not a new approved budget, the total price of the associated lots, or an unconditional obligation to exercise. [P1]
NVR says it evaluates potential deposit losses by community using sales absorption, recent home-sale margins, the gap between contracted and current lot prices, developer performance and financial capacity, and contract-default status. When homes cannot be sold at an acceptable margin and pace at the contracted lot price, the company considers performing, restructuring or defaulting and forfeiting deposits. The filing does not state the numerical thresholds behind “acceptable.” [P1]
A mixed operating picture
For 2023, NVR reported homebuilding revenue of $9.315 billion, down 10%; settlements declined 9% to 20,662; average settlement price fell 1% to $450,700; and gross margin declined to 24.3% from 25.8%. New orders, net of cancellations, increased 13% to 21,729, even as the average new-order price fell 3% to $448,400. The year-end backlog increased 12% to 10,229 homes. Orders and settlements are different measures over different periods. [P1]
Management attributed order strength partly to limited resale-home supply and 3% growth in average active communities. It attributed lower new-order prices partly to affordability adjustments as mortgage rates remained high and home prices had appreciated. NVR said it expected continued 2024 margin pressure from rates, economic uncertainty, and building, labor and land costs. These factors may affect demand and margins independently of land tenure. [P1]
Evidence at a glance
| Evidence at the cutoff | What it shows | What remains unknown |
|---|---|---|
| 134,900 LPA lots; 4,850 JV-controlled lots; 1,750 lots from owned land under development. [P1] | NVR has a predominantly LPA-based finished-lot position with smaller alternatives. | Lot-level prices, timing, community availability and comparable economics by land-control type. |
| $391.3m in potential additional deposits, conditional on milestones and option exercise. [P1] | Existing contracts can require further deposits if conditions are met and NVR exercises. | Timing and exercise decisions by lot, project-level cost to complete, and expected margin/absorption. |
| Orders +13%; settlements −9%; backlog +12%. [P1] | Order and delivery trends were moving differently in 2023. | Future cancellation, start and settlement rates, and how much change is due to supply, price, mix or demand. |
| Company’s community-level review considers absorption, margin, lot price and developer status. [P1] | NVR already describes a decision process for evaluating certain contracts. | Actual thresholds, lot-level outcomes and whether different control structures perform differently. |
The choice
Should NVR maintain its LPA-centered approach, use more selective JV or direct-development exceptions, or make the primary change at the community level by pacing takedowns and starts as conditions change? Identify the evidence that would govern the choice, the strongest counterargument, and how management should distinguish finished-lot supply from raw-land pipelines. Do not assume a fixed budget, an optimal land mix, or numerical returns unsupported by the public record.
Sources and conventions
All information is limited to public material available by March 1, 2024. NVR’s FY2023 Form 10-K was filed February 14, 2024. LPA, JV, owned/developing land and raw-land-contract lots are separate categories; the 22,700 raw-land-contract lots are not included in the 141,500 controlled-lot total. Dollar figures are reported in thousands in most filing tables unless stated otherwise. No post-cutoff outcomes are included.
- P1: NVR FY2023 Form 10-K, filed Feb. 14, 2024; Item 1 p. 1; Item 7 pp. 13–15, 22 and 24; Notes 1, 3, 4 and 5, as detailed in the accepted source register.
- A1: Murray (2020), “Time is Money: How Landbanking Constrains Housing Supply,” Journal of Housing Economics 49, article 101708.
- A2: Kim and Choi (2015), “A Study on Operation Mechanism and Feasibility of Real Estate Development Utilizing the Land Purchase Option,” Journal of Korea Planning Association 50(5), 107–117.