Description

NVR is a homebuilder. Their operating model, which is unique (and which is
described later), allows them to assume the least risk in the industry and
produce returns that are the largest.

Homebuilders are generally dismissed because they're cyclical and
interest-rate sensitive (really, though, which industry isn't?) and
downturns inevitably leave homebuilders holding large inventories of unsold
properties -- the unlevered builders then suffer large inventory writedowns
while the levered builders go into bankruptcy. However, NVR's model will
prevent it from suffering the same fate and, indeed, NVR will prosper in a
downturn at the expense of the weaker builders.

Two of the most important facets to its operating model are:

(1) NVR acquires control of land inventory through options contracts. These contracts give NVR the right to buy finished lots from developers. NVR secures a supply of land for its homebuilding operations through the use of these options whereas other homebuilders purchase land outright and engage in land development. By avoiding that speculative practice of land
purchase/development, and instead using options, NVR is able to control
large blocks of land (years' worth) in its markets while employing less
capital to do so. The lower capital requirements of this method translate
into lower inventory risk and greater returns on capital.

(2) NVR pre-sells nearly all of its homes. Other homebuilders typically
participate in some speculative construction. NVR does not. Before NVR
begins construction, an order must be placed and a deposit made. This
practice reduces risk and working capital requirements, which further
enhance returns on capital.

In addition to NVR's superior model, consider the following:

-- Low valuation: NVR trades at a P/E of 8.6x trailing (7.1x 2001E EPS) and a TEV / EBITDA of 4.7x (trailing). TEV / (EBITDA - Capex) is 4.8x
(trailing). TEV / FCF is 7.8x (trailing). I am defining FCF as Net income plus D&A minus Capex.

-- Backlog: NVR has a backlog of 5,765 ordered homes. These homes
represent $1.49 billion of revenue. To put this into perspective, this is
nearly three fiscal quarters of revenue. In addition, the homes in backlog
carry higher gross margins than the ones in the historical results. All of
this should translate into higher EPS. (Management says 2001 EPS should be
just under $20 per share. In the short history that the company has
provided guidance (previously they refused to) they have consistently been
ridiculously conservative. Their 1Q results and the backlog indicate to me
that the $20 EPS estimate continues to be the case).

-- High ROIC: The low capex nature of its business ($301 mil LTM
homebuilding EBITDA versus consolidated LTM Capex of $5 mil) and the low
working capital requirements of its model allow NVR to produce superior
returns on invested capital: 45.3% in 2000, and 5-year average ROIC of 25%. Bonus fact: In 2000, NVR sold $325 mil more homes than it did in 1999, yet inventory (the bulk of a homebuilder's working capital requirement) increased only $11 million.

-- Intelligent allocation of excess capital: High returns on capital and
excess cash flows are only useful if you have a management that is smart
about deploying it. In NVR's case, management has chosen thus far to deploy that capital to buy back its own stock. Between 12/31/93 and 12/31/00 the company reacquired 13.5 mil shares. In the first quarter of 2001, NVR purchased another 0.85 mil shares For perspective, there are only 8.1 mil primary shares out today (I'm using primary shares to illustrate this but I use diluted shares for enterprise value calculations).

-- Homes a basic necessity: People will always need homes to live in. The
process of building a home has not changed materially in decades. Neither
of these statements is likely to change in the next year, the next 5 years,
or even the next 20 years. There is minimal technological or obsolescence
risk.

-- Dominant in its markets: NVR competes in 18 geographic markets. It is
the #1 player in 10 of them. As for the remaining 8, it is usually #2 or #3 (always at least in the top 5). The rest are markets that NVR has just
recently entered and will dominate with time.

-- Tax factors: The industry has indirectly enjoyed the benefits of a
government subsidy in the form of tax deductible mortgage interest.
Additionally, in the last few years, homebuyers no longer have to pay tax on the first $500k of capital gains on a home. This lowers the effective
purchase price of a home for a consumer, increases the relative
attractiveness of a home as an investment, and adds a little boost to demand for NVR's product.

NVR's profits and market dominance are all the more amazing when you
remember that the results have been achieved without land development. NVR
has margins better than its competitors despite the fact that other
homebuilders benefit from the gross margin boost of speculative development
in an inflationary environment.

Catalyst

The small number of shares outstanding occasionally creates large downward
gaps. NVR's recent 25% drop is one such opportunity.

Also, share repurchases will continue to drive the stock. It's hard to
overemphasize the magnitude of the repurchases or the wonderful track record
of buybacks:

12/31/95: 15.21 (millions of shares outstanding)
12/31/96: 13.57
12/31/97: 11.09
12/31/98: 10.39
12/31/99: 9.17
12/31/00: 8.86
04/18/01: 8.14