NVR is a homebuilder that has been written up twice before on VIC: In 2006 as a short candidate and in 2001 as a long.  We believe that for investors with a reasonably long time horizon, NVR presently offers compelling value as a long and asymmetric risk/reward.

We would encourage readers to check out cmn3d's detailed piece on US housing which lays out a pretty good argument for a housing recovery.

Ok, so now for the case on NVR. To start with, let us say from the get-go that NVR is not the juiciest way to play a recovery in US housing as there are many other names out there with greater upside. However, we believe that NVR offers a very safe way to play this recovery while mitigating the chance of permanent impairment of capital. As evidence of the safety, we point out that NVR was FCF-negative during only 1 quarter of the most recent recession, a testament to the defensibility of its business model. The company has over $800mm of net cash on balance sheet.  NVR currently trades at ~2x book, which may sounds high (it is much higher than all other homebuilders) but: A) NVR's long-term average multiple is about 4.3x, and B) NVR's unique business model and huge ROIC relative to other builders warrants a much higher book multiple.

Background on NVR

NVR builds and sells homes (mostly single-family detached) in 22 MSAs within 12 states in the eastern US. NVR also has a mortgage banking and title services company which primarily services its homebuilding operation. NVR is among the 5 largest homebuilders in the US.

NVR emerged from bankruptcy in 1992. The BK was the result of an overleveraged balance sheet that deteriorated in the housing recession of the late 1980s. Prior to BK, NVR was a full scale land developer, manufactured housing matierals, had a more extensive finance arm, and was geographically diversified across the country. Post-BK, NVR refocused its business on the Mid Atlantic and Rust Belt, shedding the housing materials and most of the finance business, and ceasing to develop land and build on spec. In 1997, NVR bought Fox Ridge Homes for $20mm, giving it entry to the Nashville market. In 2005, NVR bought Marc Homebuilders for $8mm, in the Columbia SC market.

NVR is run by CEO Paul Saville, who has been with Ryan Homes (precursor to NVR) since 1981 and has been CEO for about 15 years. He owns 2.5% of the company, largely the result of LT incentive comp. Chairman Dwight Schar was CEO before Saville, having joined Ryan Homes in 1969. He owns 1% of the co.

Markets

65% of NVR's gross profit comes from the Mid-Atlantic (Wash DC, Baltimore, etc). 16% is Mid-East (Cinci, Cleveland, Pittsburgh), 11% is South East (Charlotte, Nashville) and 8% is North East (Philly). 

What is important is that NVR's markets did not get nearly as overbuilt as some others in the country during the RE bubble. Peak-to-peak unit growth overall from 1986 to 2005 was only 0.3% CAGR. However, whereas in some very overbuilt markets there has been very little new construction in the past 5 years, in NVR's markets construction has continued at a slightly higher pace. The result is that current months of new home inventory is about 6 months for both NVR and for the country as a whole.

NVR has been a consistent market share gainer in its footprint. Estimated current mkt share is about 18%, about a double in share in the last 5 years (downturn). Prior to the downturn, NVR gained on average 60 bps of share per year since emerging from BK in the early 90s.

Business Model

NVR has a unique business model within the homebuilders that is predicated on two main points: 1) nearly 100% of land controlled is optioned rather than purchased on B/S, and 2) NVR is geographically dense (ie local economies of scale). The result of these aspects are that due to #1, NVR operates much more capital-light than other builders (as evidenced by its 50%+ after-tax ROIC through the cycle) and #2 allows for NVR to be far more efficient and low-cost in the markets in which it operates.

NVR considers itself more of an assembly line manufacturer than a custom home builder (per mgmt, "We like to think of ourselves as the In-N-Out Burger of Homebuilders".) NVR has 6 manufacturing and distribution facilities where they pre-assemble exterior and interior walls, stairs, railings, etc. Prefab packages are delivered to building sites where they can be assembled quickly. From day 1 of laying foundation, NVR can usually deliver a home in <90 days, often 60-70 (compared to ~100-120+ industry avg).

Consistent with its focus on efficiency, NVR offers fewer home models (4-5) per community and fewer customization options per model. If <20% of buyers select a given option, it is kicked out of the option pool for future communities. Importantly, NVR is NOT a spec builder; rather homes are constructed only after the receipt of a signed contract and deposit from a buyer.

NVR has an in-house mortgage banking business designed to underwrite loans for buyers of their homes. Loans are underwritten per the standards of the eventual loan owner (Wells Fargo, BofA, etc) and are held on B/S for <30 days on average. 

NVR rarely buys land for development. Instead, NVR enters into options on land, paying 3-10% of aggregate purchase price as a premium. It can walk away at any time. It takes down finished lots on an as-needed basis as homes are contracted to be sold.

Mgmt is compensated with a focus on shareholoder value creation and a LT outlook. Most comp comes in RSUs and Options that vest over 4-5 years. Site managers are incentivized based on delivering homes quickly.

Cash is primarily used to buyback stock. NVR has bought back about 2/3 of co since 1994.

Investment Thesis in a Nutshell

NVR is a high quality, best-in-class homebuilder with a proven business model that was resilient to the worst resi real estate market in the last 80 years (profitable every quarter but 1). Current consensus sentiment on resi real estate is extremely bearish, presenting an intriguing entry point to own a high quality mean-reverter at a reasonable price.

Our normal year assumptions call for NVR to deliver about 17k homes, which implies flat market growth average-to-average from the last cycle to the next (recall that NVR's markets did NOT get overbuilt that much during the last cycle) and 13.5% market share for NVR (versus 18% today). We assume that NVR ASPs are flat for the next 2 years and then grow with inflation thereafter. Our model also calls for an Operating Margin about 50 basis points below NVR's long-term average despite the fact that they are more efficient operators today. We also assume that this "normal" year does not occur until 2016 (10 years after the beginning of the downturn). Thus, we view this case as a reasonably conservative case.

Assuming a recovery to this conservative "normal" in 2016 and prudent capital allocation (as NVR's history suggest should be the case), we should generate a high teens IRR on this investment from the current level. We get an option on a quicker recovery and/or a higher actual normal year, both of which are totally reasonable (and arguably more likely) outcomes. Downside is protected by the quality of the business, its proven resiliency during downturns, and the entry multiple.