One of the things I love most about investing is that wonderful investments can be found in the most random places. Places you would never have assumed were possible, in industries that you normally wouldn’t spend 10 seconds on.

And one industry that hits the nail on the head for being boring is “homebuilders.” Just the mention of that one word makes my eyes glaze over while I let out a large yawn.

But what if I told you one of the best compounders of the last few decades was found in the homebuilding category? I’m not talking about a business that doubles and then fades into obscurity. Rather, a business that, over the last three decades, has returned returned 62,000%!

Got your attention?

The business is NVR. It has the rare combination of being a business with an exceptional history of intelligent capital allocation, a fortress-like balance sheet, and a model that has thrived through every downturn since the mid-90s, including the Great Financial Crisis. During the GFC, NVR was reportedly the only homebuilder to turn a profit. That alone shows you how durable the business is and how superior its business model is to others in the homebuilding industry!

Let’s get into it.

— Kyle

In 1992, NVR faced a severe downturn in the home-building market and filed for bankruptcy. But unlike the fates of most homebuilders, this crisis became a turning point, and didn’t signal the end. When the company came out of bankruptcy, management learned a crucial lesson: the legacy homebuilding model, where you loaded up on debt to carry a highly speculative land inventory, was fragile and not a business model worth maintaining.

The two biggest existential risks that were identified were:

Rather than continuing the same mistakes that led to bankruptcy in the first place, NVR decided to reinvent the wheel. They would no longer speculate on land. Instead, to reduce risk, they would purchase land only when a guaranteed buyer was lined up, wanting to build a home on the lot. (In hindsight, this seems obvious to do, doesn’t it?)

What NVR did uniquely was structure this new approach by buying options on lots in desirable neighborhoods, and, with the option (but not the obligation) to purchase the land for development, would seek a buyer for the property. If found, NVR exercises the option to purchase the land, develop it, and generates revenue from the difference between the sale price and the cost of the land + development costs.

If no buyer is found, the modest premium paid for the land options expires, and that money is lost, but that’s vastly preferable to spending 10x more to buy the land and then being unable to develop it, with the property just sitting idly on the balance sheet indefinitely.

This revamped strategy allowed NVR to avoid entirely taking on price risk for land value. It also dramatically reduced their capital requirements, allowing them to generate exceptional cash flow. By the early 2000s, NVR had industry-leading capital efficiency metrics and still does.

Shortly after NVR changed its business model, a young, up-and-coming hedge fund manager named Norbert Lou began buying its stock. Lou started buying NVR at only $23; today it’s valued at $6,100. He has made a number of outstanding bets that he’s held onto for the long term while compounding capital in the mid-teens.

But his reasoning for buying NVR back in 2001 was straightforward. He liked two areas of the business:

Lou liked these attributes of NVR because they reduced risk. His early write-up on NVR on VIC has been studied in depth by investing legend Joel Greenblatt. Greenblatt liked the write-up so much that he featured it in his teaching lectures. Greenblatt once said regarding Nobert Lou: “To this day, I hand out the first 3 write-ups he wrote on the Value Investors Club to my students at Columbia, to show them what a brilliant, concise, straightforward and clear investment thesis looks like."

As Nobert Lou correctly pointed out, traditional homebuilders do not utilize this options-based contract approach to develop their land. Instead, they purchase the land outright and store it on their balance sheet. Once they believe the market will demand their inventory, they develop it and, hopefully, sell as much of it as possible.

The problem with this, which NVR avoids, is that NVR doesn’t have to keep expensive inventory on its balance sheet when land purchase agreements (LPAs) are in place. The LPA means they must deposit 10% of the land cost. If they don’t foresee themselves developing the land in the future, they can choose to forfeit the deposit to the developer and move on to other lots they can develop and sell for the right price.

Just to give you an idea of how many LPA’s NVR has. As of March 31st, 2026, they currently have 172,100 lots under LPAs. They have increased their contract land deposits to about $944m, up from $856m a year ago. This shows they are still aggressively buying these lots. It also confirms that the average LPA costs between $5,000 and $10,000.

This business model keeps capex at a fraction of that of other homebuilders. This advantage is a major boost to cash flow, as NVR doesn’t have to participate in the development of raw lots; the original developer handles that.

This means they don’t have to tie capital up in inventory that might lose value over time. And since they don’t need to own the entire lots from start to finish, they don’t have to carry many assets on their balance sheet. Check out their assets vs a few other public competitors below:

This strategy has been highly beneficial for NVR shareholders. Not only does NVR not have to carry many assets on its balance sheet, but the assets it does have are very likely to sell due to the LPA strategy. This means in a downturn, NVR doesn’t have unsold land collecting dust on its balance sheet. And in an upswing, NVR doesn’t need to “pay up” for overpriced land to develop if it already owns the option to develop it in a strengthening market.

This allows NVR to maintain lower leverage ratios than its competitors.

All of NVR’s business segments generate profits. There are two segments to NVR:

Both business lines are profitable, but the mortgage segment carries significantly higher margins. I think NVR has done a good job here by sticking with loans on its own properties and not trying to originate loans on behalf of others. This gives them a clear sense of what is happening in their specific markets and provides additional data points they can use to determine where and when to develop future lots.

NVR is no monopoly. It has a ton of competition in many of its markets, and that’s unlikely to change anytime soon. And yet they have clearly done something right over the last three decades to generate so much shareholder value.

When I thought long and hard about what exactly their advantage was, the best thing I could come up with was a type of counter-positioning moat. For those unfamiliar with counter-positioning, the idea is that a company’s competitor (like Lennar) would harm itself by cloning a competitor's business model (in this case, NVR). It’s similar to the innovator’s dilemma.

This can be broken up into a few parts:

Also, NVR doesn’t have to shun the down cycle. If traditional homebuilders know they must wait to sell into the upcycle, they will sit on their assets without earning a return. Meanwhile, NVR can often buy options on land that developers don’t want to develop at discounted prices, while still being able to find buyers, even in weaker markets, to sell to before exercising the option.

NVR has a few things working for it, helping to create enormous shareholder value from compounding earnings per share by around 15% since the year 2000.

Firstly, I attribute this to management’s focus on treating shareholder capital as if they were owners. Since insiders own 8.6% of the shares and the business has been well structured to align management and shareholders, they have kept close tabs on management's ability to allocate capital effectively. Management is incentivized on this after all.

Over the last 1, 5, and 10 years, they have ranked #1 in average annual return on capital and return on equity versus peers, and they are 2nd in average annual operating margins.

But that’s just the start. When you look at the mechanics of why they have done such a good job allocating capital, despite operating in an intensely competitive industry, the biggest lever has probably been their incredible use of share buybacks.

When NVR first went public, the company had 14,300,000 shares outstanding. Today, that same figure stands at just 2,800,000! That’s an 80% reduction in share count, a number that I challenge you to find me a match for (besides Autozone, which is a famous share cannibal in its own right). The best thing about NVR’s buyback program is that, since it operates in a cyclical industry that rarely trades at high multiples, the buybacks have been incredibly value-accretive for shareholders.

The point of a good buyback program is to buy undervalued shares so that existing shareholders, by doing nothing, get a larger piece of the ownership pie by reducing the overall share count. A simple chart will suffice in showing how this is done — see below.

Intelligent buybacks mean that EPS should grow at a faster rate than net income. While net income grows at one rate, earnings per share (net income divided by share count) grow faster thanks to a declining denominator from buybacks.

The problem that most companies have with buybacks is that they buy their shares back at any price. This, unfortunately, shows a lack of understanding of capital allocation. If you buy your shares back when they are above intrinsic value, you are destroying shareholder value.

On that note, NVR has made some fantastic share-repurchase decisions over the years:

Long story short, they have shown they can time these buybacks well!

NVR has intentionally refrained from issuing a dividend because buybacks have worked so well and offer better tax consequences for NVR shareholders. Many public companies cater to shareholders who want to be paid dividends, but NVR has stayed away from that because it understands that, from a capital allocation and tax perspective, share buybacks will deliver the most value to shareholders. (Dividends are taxed as income, creating a tax bill that shareholders didn’t consent to, whereas buybacks are a form of capital return that generates nearly zero tax consequences for shareholders.)

And lastly, the board has authorized a perpetual buyback program. So whenever NVR has excess capital, it doesn’t have to reinvest it; it can opportunistically buy back shares at its discretion. And unlike many businesses that authorize 5%-10% buybacks and then underdeliver, NVR has been very aggressive in buying back meaningful amounts of shares.

NVR has a strong track record of margin expansion, peaking in 2022 at the height of the post-COVID building boom. Both gross margins and operating margins maxed out at around 26.7% and 21.7%, respectively.

When I first analyzed NVR, the challenge I faced was determining where these margins would be in 5 years. Would it head back to that 2022 level like a strong magnet, stabilize around current levels, or crater further?

To answer this, I had to look at the root causes of what has happened since 2022. The first issue is that both lot and land costs have risen sharply. Next, since demand has softened, NVR has had to dangle incentives to buyers to get them to buy NVR’s homes. Then you have mortgage rates rising to levels not seen since 2009, further suppressing demand.

In my view, mortgage rates are probably the biggest deciding factor in where NVR's margins go in the future. If you think we are getting back to those COVID rates of 2-3%, then perhaps the company will likely revisit operating margins in the 20%+ range. Alas, I don’t foresee us getting there anytime soon, given where inflation is right now. Rates are high by historical standards, but they also seem to be stabilizing here. So I’m most comfortable assuming that NVR’s margins will stay within its current range, plus or minus 1-2 percentage points.

The good thing about NVR is that they are still turning a profit even in a pretty mediocre environment of elevated rates and uninspiring demand. Even when looking at competitors, it’s clear that the industry is experiencing margin compression now. Until rates decrease, demand changes, and inflation subsides, I assume margins across the industry will remain low.

I just spent some time discussing interest rates and how they are arguably the biggest factor for NVR’s currently depressed margins. But I want to discuss the risks that I see for NVR, too.

Regarding mortgage rates, if they remain elevated in the 7%-8% range, I still don’t see them posing an existential risk to NVR. They are still turning a profit, and the business model will continue humming along, but if rates continue to climb, the risk is that NVR’s declining top line and margins will create a double-whammy for profits.

The other risk has to do with their LPAs. On a quarter-over-quarter basis, it hasn’t looked that bad with contract land deposit impairments of $8.9m vs $8.1m in the prior year’s quarter. But for FY 2025 vs FY 2024, the numbers are eyebrow-raising. They had $75.8m in impairments for FY 2025, vs only $7.1m in 2024.

It’s great to see the recent quarterly number come down a little bit after the elevated annual number, but it’s still much higher than any number NVR has shown since the GFC.

It would be great if NVR could fully insulate itself from cyclicality, but that’s simply not possible. Shawn made the excellent point, though, that the best time to buy cyclicals is counterintuitive. It’s actually when the multiple is highest, because earnings are most depressed and due for a cyclical turn.

Since NVR focuses on building lots in very specific states and communities (16 states and 432 communities), there is a risk that a specific community might fail to meet NVR’s standards. Maybe a community sees an exodus of its population, or NVR is wrong about the demographics and buying power of potential customers. If they get this wrong, they will be stuck with LPA’s that are unattractive to develop. This is the situation I think we are now witnessing, which is why the impairments have risen.

The final risk I want to mention is that much of NVR’s value creation has come from EPS growth driven by buybacks. Since NVR is generating lower profits, its balance sheet and cash balance haven’t been replenished at the same rate as they were post-COVID. If they are unable to continue buying back as many shares, combined with the headwinds I discussed above, EPS will almost certainly decelerate substantially.

NVR is an interesting business. It has a good business model, which has created outstanding shareholder value for decades. But when looking at NVR, we have to be very careful not to extrapolate historical numbers into the future. There have been times in NVR’s history that EPS growth would make nearly any business on the planet jealous. Below, I have NVR’s EPS growth from the bottom of the GFC to the peak of COVID. And the numbers are spectacular, with a 27% CAGR over 14 years.

But it’s not as simple as just assuming they continue growing like this. The recent past paints a much different story. Since the height of COVID, EPS has a CAGR of -5.7%.

So the real question is, where is the middle ground? No one knows where rates will go, but I’m perfectly fine assuming they stay similar to where they are today and perhaps get a little better over the medium term. This would enable the revenue growth rate to return to pre-COVID levels.

As a result of that assumption, my base case supposes a 5-year revenue CAGR of about 6%. At the same time, I assume we get some demand back. This should allow their average selling prices to increase moderately, while keeping margins up. With that in mind, I assume profit margins will remain in the 12% range, a little lower than they are now, just to stay conservative.

There are a few other areas to consider when evaluating a business like NVR. The first is debt. NVR has historically carried debt just like any other homebuilder, albeit at a much lower ratio to its profits. For this reason, I still think NVR will carry debt in the future, but will also have a lot of cash left over, meaning cash will continue to be a larger number than debt. This will result in a negative net debt number, which affects their enterprise value.

Next, we have to factor in that NVR will continue to meaningfully reduce its share count. In the base case, I assume shares continue to come down to about 2.4 million. Lastly, we assign a multiple of 12x EV/Earnings, and we get a return of about 8%. However, I’m not super comfortable with this number as the business is in a highly cyclical industry, so I’m adding a 25% margin of safety. This brings the price to $6,483, offering only a 2% annual expected return.

This return is well below our typical target of about 12%. To get that kind of return, shares would have to drop about 40%. And even in that case, I’m still not sure I like the business model or that I have enough understanding of how the home-building cycle works to want to put this in the portfolio.

The other problem that I’ve had to deal with in the past is holding businesses through down cycles. For investors who truly understand cycles, this can be a massive wealth generator. But for me, it just doesn’t suit my investing style and preferences. I like businesses where I can generally assume the top and bottom line will go up annually, with limited cyclical impacts.

So we will be taking a pass.

For more, you can listen to our podcast on the company here.

More updates on our Intrinsic Value Portfolio below 👇

“There is magic in buybacks. And the magic really starts happening when you start taking out more than 80% of shares. So you buy back half the shares, you get a 2X if there's no change in anything else. You buy back two-thirds, you get a 3X, buy back 80%, you get a 5x, 90% is 10x.”

— Mohnish Pabrai

📺 WATCH: Joel Greenblatt’s Columbia Lecture on the NVR Case Study

🎧 LISTEN: Mohnish Pabrai’s Mental Models for Exceptional Capital Allocation

📖 READ: Doing Nothing by Petty Cash about the difficulties of inactivity

You can also read our archive of past Intrinsic Value breakdowns, in case you’ve missed any, here — we’ve covered companies ranging from Alphabet to FICO, Transdigm, Lululemon, PayPal, DoorDash, Crocs, LVMH, Uber, and more!

Enjoy reading this newsletter? Forward it to a friend.

Was this newsletter forwarded to you? Sign up here.

Join the waitlist for our Intrinsic Value Community of investors

Shawn & Daniel use Fiscal.ai for every company they research — use their referral link to get started with a 15% discount!

Use the promo code STOCKS15 at checkout for 15% off our popular course “How To Get Started With Stocks.”

Follow us on Twitter.

Read our full archive of Intrinsic Value Breakdowns here

Keep an eye on your inbox for our newsletters on Sundays. If you have any feedback for us, simply respond to this email or message [email protected].