What happened next? Homebuilder stocks went on a tear. From last July through its peak this May, the S&P homebuilders index rose by 51%, more than doubling the S&P 500’s return. PulteGroup shares, as one example, jumped to $24 from $9. It seemed everyone on Wall Street had piled into the trade.

The case for homebuilders last year boiled down to two key points. First, the supply of new homes was spectacularly thin. A historically low number of new homes existed on the market, so it was clear that builders would have to build more to meet demand, even though demand for new homes was less than half of normal. Second, even though homebuilder shares had already risen, the stocks still traded at huge discounts to what they would be worth in more normal times when demand for new homes returned. Those discounts gave investors a healthy “margin of safety,” which meant that even if you timed the stocks wrongly, you wouldn’t lose boatloads of money because the stocks were cheap to begin with.

MORE: El-Erian: Fed’s taper should start small

Foreclosures and rising mortgage rates and all the other variables in the housing market are important, but their effect on homebuilder stocks is far smaller than these two points. If you have a tight market for new homes and homebuilders stocks’ are cheap, the stocks scream buy.

So after a big run-up, how do they look today?

Here’s where things get interesting. Back in May, investors awakened to the fact that mortgage rates could rise considerably after the Fed talked about ending quantitative easing. Investors acted as if demand for new homes would plummet. Homebuilders’ stocks declined throughout the summer, falling nearly 30% from their peak in mid-May through early September. The good news is that the decline has given investors a possible window. If homebuilders are 30% cheaper than they were a few months ago, and their future success is still probable, then the stocks should be even better buys today.

So have conditions changed negatively for homebuilders? The answer is no. The tight inventories of a year ago remain virtually unchanged, says Mike Castleman, Jr., a senior vice president at Metrostudy, a research firm that surveys nearly 70% of new home markets across the U.S. One telling statistic that Castleman follows is the number of finished vacant inventory, which means new homes that are completed but not yet occupied. At the end of the second quarter of 2012, the number of finished vacants was 29,428. That remained virtually unchanged at the end of the second quarter of 2013, at 29,283.

Across all of Metrostudy’s markets, there’s a 1.9-month supply of finished homes, which is below the market equilibrium of 2 months to 2.5 months of supply. Castleman ticks through markets facing shortages: the Twin Cities have a one-month supply of new homes; Suburban Maryland, 1.1 months; San Diego, 1.6 months; Salt Lake City, 1.1 months; Northern California, 1.3 months.

“That’s in the face of higher prices, deals that have been pulled back, and mortgage rates that have gone up,” says Castleman. If the total number of finished homes remains at these depressed levels, homebuilders will be able to raise prices even more, a phase of the housing recovery we’re currently experiencing. “Right now, it’s a very builder-friendly market,” he says.

Back to homebuilder shares. If the story is still positive for homebuilders, how much of a window did this summer’s selloff create? Any bad news of rising mortgages rates is probably priced into shares, says Ivy Zelman, CEO of Zelman & Associates. Zelman was one of the only Wall Street analysts to correctly call the housing downturn in 2006 and she remained mostly bearish on the market until early 2012. In other words, she’s worth listening to.

MORE: Don’t bet on a European recovery yet

Zelman follows more than a dozen homebuilders and says those stocks trade at seven times her 2015 earnings estimates, making homebuilders very cheap on a normalized basis. Despite recently reducing her fair value estimates because homebuilders have risen so quickly, Zelman recommends buying 14 builders, which on average enjoy a 50% upside over the next 15 months. Her recommendations with the highest potential gains include MDC Holdings (MDC), PulteGroup (PHM), and William Lyon Homes.

Zelman writes in her latest report that previous housing recoveries have experienced similar pauses as the one experienced this summer. “For example,” she writes, “in the middle of the early 1990s recovery, homebuilding stocks underperformed the broader market by more than 20% over a six-month period in 1992, erasing roughly half of the group’s outperformance in the year prior. However, beginning in the fall of 1992, housing stocks resumed their outperformance, ultimately climbing an additional 100% through the end of 1993.”

Zelman expects housing stocks to continue to outperform the market over the coming years as demand for new homes returns to normal levels.

Don’t be so surprised this time when homebuilders continue their ascent.