Existing vs. New Home Sales Indicators

Two satisfying-looking data releases hit each month—existing home sales from the National Association of Realtors (NAR) and new home sales from the Census Bureau—and most investors treat them as interchangeable "housing data." They are not. Existing home sales capture 85–90% of transaction volume but reflect decisions already made (closings of contracts signed weeks earlier). New home sales capture only 10–15% of volume but reveal where builders are betting real capital on future demand. Reading them together (and knowing which one to trust when they diverge) gives you a genuine edge in interpreting the housing cycle, consumer spending trends, and even GDP trajectory.

TL;DR: Existing home sales measure the bulk of housing activity but lag real-time demand; new home sales are volatile and heavily revised but forward-looking. Use pending home sales to bridge the gap, and never overreact to a single month's new-home number.

What Each Indicator Actually Measures (And Why the Difference Matters)

Existing home sales count completed transactions of previously owned homes. The NAR surveys Multiple Listing Services and local boards, then annualizes the figure. When you see a headline like "existing home sales fell to 3.96 million," that means the monthly pace, multiplied by twelve, implies roughly 3.96 million closings per year.

Key metrics in the release include total sales volume (annualized), median and average sale prices, inventory levels (expressed as months of supply), and days on market. These numbers describe the largest segment of housing (around 5 million transactions in a normal year), so they dominate headlines.

New home sales count contracts signed (not closings) for newly constructed homes. The Census Bureau collects this data from a sample of builders and permit offices. Because the count is based on contract signing—often before a shovel hits dirt—new home sales are inherently more forward-looking than existing sales.

The point is: Existing home sales tell you what already happened. New home sales tell you what builders expect to happen. When the two indicators diverge, the divergence itself is the signal.

Why the Data Sources Differ

The NAR is an industry trade group with access to MLS data (a comprehensive record of listed and sold homes). The Census Bureau relies on sampling—a smaller dataset subject to larger statistical noise. This explains why new home sales carry confidence intervals of roughly ±10% on any given month, while existing home sales are more stable. Understanding the source helps you calibrate how much weight to put on each number.

How to Read the Numbers in Practice

Existing Home Sales: Volume, Inventory, and Mix

A typical existing home sales release gives you three things worth tracking:

1. Sales volume vs. prior year. Month-over-month changes are noisy (seasonal patterns are strong—spring selling season consistently outperforms winter). Year-over-year comparisons strip out seasonality and reveal the real trend.

2. Months of supply. This is the most actionable number in the release. The calculation is straightforward:

The calculation: Months of Supply = Active Listings ÷ Monthly Sales Rate

A range of 4–6 months is considered balanced. Below 4 months signals a seller's market (prices likely rising). Above 6 months signals a buyer's market (prices likely softening or declining).

3. Median price (with a caveat). NAR reports a simple median—the middle sale price. This is not a home price index. If expensive homes sell disproportionately in a given month, the median rises even if no individual home appreciated. For true price trends, cross-reference with Case-Shiller (repeat-sales methodology, 2-month lag) or the FHFA House Price Index (repeat-sales on conforming loans, misses jumbo and cash transactions).

Why this matters: Investors who treat NAR median price as a value indicator get misled by mix shift every single month. A rising median during falling volume can mean the market is weakening (only affluent buyers are transacting), not strengthening.

New Home Sales: Builder Confidence and Incentive Signals

New home sales releases are smaller in magnitude but richer in forward-looking content:

• Contract volume tells you whether builders are successfully selling into current demand.
• Median price for new homes can reveal builder incentives. When builders offer mortgage rate buydowns, upgrades, or closing-cost assistance, the effective price drops even if the sticker price holds steady. A falling new-home median price often means builders are cutting deals to move inventory—a leading signal of softening demand.
• Months of supply for new homes. The healthy range is 5–7 months. Above 7 months indicates builder stress (too much unsold inventory). Below 5 suggests builders can't keep up with demand and may raise prices.

The point is: New home sales data embeds builder behavior—and builders are professional forecasters of local housing demand. When builders pull back (rising inventory, deeper incentives), pay attention even if existing sales look stable.

Worked Example: Reading an Actual Release

Scenario: You're reviewing the October 2024 housing data to assess whether the housing market is stabilizing.

Existing home sales (NAR release):

• Sales volume: 3.96 million annualized
• Median price: $407,200
• Active listings: 1.37 million
• Monthly sales rate: approximately 330,000

Your calculation: Months of supply = 1,370,000 ÷ 330,000 = 4.2 months

Interpretation: Sales volume is historically weak (pre-pandemic normal was closer to 5.3–5.5 million), but inventory at 4.2 months is still below the 4–6 month balanced range. Prices remain supported because supply hasn't normalized, even though fewer people are buying.

New home sales (Census Bureau release):

• Sales volume: 679,000 annualized
• Median price: $437,300
• Months of supply: 7.7 months

Interpretation: New home inventory at 7.7 months is above the stress threshold of 7 months. Builders are sitting on unsold homes. The elevated median price suggests incentives haven't fully compressed sticker prices yet, but the supply overhang signals that more aggressive discounting (or construction pullbacks) are likely coming.

The combined read: Existing home sales are constrained by low inventory and the lock-in effect (more on this below). New home sales are showing early signs of builder stress. The divergence tells you that housing demand isn't collapsing—it's being redirected. Buyers who can't find existing inventory are turning to builders, but builders have now overbuilt relative to that redirected demand.

Revision reality check: That 679,000 new home sales figure will be revised. Historical revision patterns show first estimates routinely shift by ±10%, and final revisions can exceed ±15% from the initial print. A "679,000" reading could easily become 610,000 or 750,000 after revisions. Never make a portfolio decision on a single month's new home sales number. Use a three-month moving average instead.

The Lock-In Effect (Why Existing Sales Collapsed After 2022)

Existing home sales dropped from 6.1 million annualized in early 2021 to roughly 3.8–4.0 million by 2024. The primary driver wasn't weak demand—it was the lock-in effect.

The mechanism: An estimated 60% of outstanding mortgages carried rates below 4% as of 2024. Homeowners with a 3% mortgage face a brutal calculation: selling means surrendering that rate and financing the next home at 7%+. On a $400,000 mortgage, that's roughly $900 per month in additional interest cost—just for moving. The rational response is to stay put.

The result: Existing inventory dried up, transaction volume cratered, but prices stayed elevated (because the few homes that did list sold quickly to pent-up buyers). New home sales gained market share because builders could offer mortgage rate buydowns (effectively subsidizing the buyer's rate to 5.5–6%) that individual sellers couldn't match.

Why this matters: If you're using existing home sales as a demand proxy, the lock-in effect makes the signal unreliable. Low existing sales volume in 2023–2024 didn't mean nobody wanted to buy—it meant nobody wanted to sell. New home sales became the cleaner demand signal during this period precisely because they weren't distorted by the lock-in dynamic.

Pending Home Sales: Bridging the Gap

The NAR also publishes pending home sales—contracts signed but not yet closed. Since closings take 30–60 days, pending sales are a genuine leading indicator for existing home sales.

How to use pending data:

• Rising pending sales signal that existing home closings will increase 1–2 months later.
• Falling pending sales signal closings will decline 1–2 months later.

The caveat: Approximately 15–20% of contracts fall through before closing (financing denials, inspection failures, buyer remorse). This "contract fallout rate" means pending sales consistently overstate eventual closings. Don't simply take the pending number and project it forward—discount it by the historical fallout rate for a more accurate forecast.

The point is: Pending home sales give you advance notice on existing sales direction, but the magnitude requires adjustment for fallout. Track the trend (rising or falling), not the exact level.

Price Data: Which Source to Trust

The practical rule: Use NAR median price for a quick read on what's transacting. Use Case-Shiller or FHFA for actual price trend analysis. If NAR median price and Case-Shiller are moving in opposite directions, trust Case-Shiller—it's methodologically sound, even if slower.

Common Pitfalls (And How to Avoid Them)

Comparing existing and new home sales levels directly. They use different collection methods, different reporting lags, and different revision schedules. Compare each series to its own history, not to each other.

Treating median price as a home value index. A month where luxury homes dominate closings pushes the median up without any individual home appreciating. This is mix-shift bias, and it fools investors every month. (If you catch yourself saying "home prices rose 5% this month" based on NAR data alone, you've fallen for it.)

Ignoring regional variation. National aggregates mask enormous differences. Florida and Texas inventory has normalized (or overshot) while Northeast and West Coast markets remain tight. A national "4.2 months of supply" number can hide a 2-month market in Boston and a 9-month market in Austin.

Overreacting to new home sales volatility. A single month's new home sales print has a confidence interval wide enough to reverse the entire narrative. Three-month moving averages are the minimum smoothing you should apply before drawing conclusions.

Missing seasonal context. Spring selling season (March–June) consistently accounts for a disproportionate share of annual transactions. A "surge" in April existing home sales is often just seasonality, not a trend change. Always compare year-over-year.

Housing Sales and the Broader Economy

Strong housing markets support GDP and consumer spending through three channels:

1. Wealth effect. Rising home values make owners feel richer, boosting discretionary spending. (This effect is well-documented but asymmetric—declines in home values reduce spending more than equivalent increases boost it.)

2. Home equity extraction. Homeowners tap HELOCs or cash-out refinances to fund consumption—renovations, education, even daily spending. When rates are high and equity growth slows, this channel shuts down.

3. Transaction-linked spending. Every home sale triggers purchases of furniture, appliances, moving services, and renovations. At 5+ million existing home sales per year, this spending is material. At 3.8 million, it's noticeably weaker.

The point is: Housing data isn't just about housing. Existing and new home sales are proxies for a meaningful slice of consumer spending and residential investment (a direct GDP component). Weak housing sales today signal softer consumer spending 2–3 quarters out.

For deeper context on how residential construction feeds into the macro picture, the relationship between Housing Starts, Permits, and Builder Confidence provides the supply-side complement to the demand signals discussed here. And for tracking business investment alongside housing, Durable Goods Orders and Capex Signals covers the corporate spending channel.

Checklist for Interpreting Housing Sales Data

Essential (do this every release)

• Check existing home sales volume year-over-year (not just month-over-month)
• Calculate months of supply from active listings and sales rate
• Note whether NAR median price is moving with or against Case-Shiller
• Check pending home sales for the forward signal on closings

For deeper analysis

• Compare new home sales to its own three-month moving average (ignore single-month prints)
• Check new home months of supply—above 7 months signals builder stress
• Look for builder incentive signals in new home median price trends
• Track the ratio of new-to-existing sales for lock-in effect severity

Revision awareness

• Flag any new home sales figure as preliminary (expect ±10–15% revisions)
• Revisit prior month's number when the revision drops—did the story change?
• Use three-month averages for trend calls, never single-month readings

Next step: Track the new-to-existing home sales ratio over six months. When this ratio rises, existing inventory is constrained (lock-in effect is dominant). When it falls, existing inventory is normalizing and builders lose their competitive advantage. This single ratio tells you more about housing market dynamics than either headline number alone.