Guide

Pending home sales explained

Harbor Realty’s regional economist used to staff agent headcount off existing home sales closings — a number that reflects offers written six to ten weeks earlier. When mortgage rates jumped in 2022, closings kept falling for months after buyer traffic had already bottomed. Branch managers over-hired into a shrinking pipeline and burned six figures in desk fees before the next downsizing round. The fix was to anchor forecasts on pending home sales — NAR’s count of resale contracts signed but not yet closed — and rebuild the three-month agent-capacity model around the Pending Home Sales Index (PHSI). Pipeline error fell from 18% to 6% on a rolling quarter basis, and recruiting spend tracked actual signings instead of lagging closings.

Pending home sales measure the moment a buyer and seller agree on price and sign a purchase contract on a previously owned home. That event precedes the closing counted in existing home sales by roughly one to three months (longer when underwriting is slow or contingencies drag). The PHSI compresses signings into an index rebased to 2001 = 100, seasonally adjusted, with regional splits. For macro readers, pending sales are the earliest NAR-published signal of resale demand turning; for operators, they are the pipeline between mortgage affordability shocks and commission revenue. This guide covers PHSI methodology, the signings-to-closings lag, regional indexes, cross-checks against MBA purchase applications, a Harbor Realty refactor, a technique decision table, pitfalls, and a production checklist.

What pending home sales measure

NAR’s Pending Home Sales report tracks contracts signed on existing (previously owned) homes reported through local Realtor associations and MLS feeds. A pending sale is not a closing: the deal can still fail inspection, appraisal, financing, title, or buyer cold feet. But among NAR’s monthly releases, signings sit closest to the buyer’s decision to transact at today’s rate and price.

Signings vs closings vs new-home contracts

Pending (PHSI) — contract signed on a resale; leads closings by 30–90+ days.
Existing home sales — transaction closed; coincident-to-lagging demand read.
New home sales (Census) — contracts on new construction; separate supply chain from resale pipeline.

Confusing pending signings with closings double-counts the same deal at two stages or mis-times rate shocks. A March signing often closes in May; March closings reflect January signings when rates may have been different.

PHSI index math and seasonal adjustment

The headline Pending Home Sales Index is not a raw contract count. NAR indexes contract activity so long-run comparisons are possible:

Base year — 2001 average monthly signings = 100 on the index.
Seasonal adjustment — spring buying season is smoothed so January and June are comparable.
Month-over-month and year-over-year — markets focus on m/m % and y/y % changes in the index, not the level alone.

A PHSI print of 75 means seasonally adjusted signings are 25% below the 2001 average pace — not that 75 homes sold. When the index rises from 72 to 74 (+2.8% m/m), that is the actionable momentum signal; the absolute level matters mainly for historical context.

Regional pending indexes

NAR publishes PHSI for Northeast, Midwest, South, and West. Regional splits expose migration and affordability divergence: the South can show firm signings while the West contracts under high prices and property taxes. Harbor Realty weights its Southeast offices using the South index plus state MLS overlays rather than the national headline alone.

Lead-lag relationship with existing home sales

Pending sales are a leading indicator for existing home sales closings at cyclical turning points. Typical lag:

30–45 days — cash or lightly contingent deals in liquid markets.
45–75 days — financed purchases with standard inspection and appraisal.
90+ days — complex contingencies, short sales, or slow underwriting.

Practical rule: two consecutive months of PHSI gains while closings are still falling often foreshadow stabilizing existing sales within one quarter — if mortgage rates do not re-spike and cancellation rates stay normal. The lead fails when contracts collapse at the financing stage (rate lock expirations, appraisal gaps) or when inventory is so thin that signed contracts cannot find closable matches.

Cancellation and fall-through rates

Not every pending sale closes. NAR and industry surveys estimate fall-through rates roughly in the high single digits to low teens in normal conditions, rising when rates jump mid-pipeline. A PHSI bounce with rising cancellations is weaker than the index alone suggests — pair PHSI with MBA purchase application trends and lender pull-through commentary.

Cross-checks: MBA apps, rates, and builder sentiment

Pending sales sit in the middle of the housing-frequency stack:

Indicator	Frequency	What it captures	Lead/lag vs closings
MBA purchase applications	Weekly	Loan application volume (intent)	Earliest; noisy
Pending Home Sales (PHSI)	Monthly	Signed resale contracts	Leading (1–3 months)
Existing home sales	Monthly	Closed resales	Coincident/lagging
NAHB HMI	Monthly	Builder sentiment	Leads new construction

Weekly MBA purchase apps can inflect before PHSI but include refi noise in headline indices — use the purchase-only subindex. Builder sentiment from the NAHB Housing Market Index leads new supply, not resale signings directly, but tight new inventory can push buyers toward existing homes and lift PHSI with a lag.

Release calendar and market read-through

NAR typically releases Pending Home Sales on the last business day of the month (or the first day of the next month) at 10:00 a.m. ET for the prior month’s signings — confirm on the economic calendar. Because PHSI leads existing home sales, it often prints before that month’s closings report, giving traders an early read on resale volume.

Who reacts and how

Mortgage originators and banks — purchase pipeline staffing; PHSI directly feeds funded-loan forecasts.
Real estate services and brokerages — commission revenue expectations; more sensitive than homebuilder equities.
Home improvement retailers — turnover drives remodeling; signings lead big-ticket projects by a quarter.
Treasuries and equities — secondary reaction unless PHSI surprises sharply; housing is one input to growth narratives.

PHSI rarely moves Fed policy alone, but persistent weakness alongside soft retail sales and rising unemployment feeds recession pricing in bonds.

Worked example: Harbor Realty pipeline refactor

Harbor Realty replaced its closings-based capacity model with a PHSI-driven pipeline after the 2022–2023 forecast miss:

Estimate regional conversion — historical ratio of South PHSI m/m change to Southeast office signings (Harbor internal MLS) over 36 months; coefficient ~0.82 with six-week lag.
Apply cancellation haircut — subtract 9% fall-through in normal rate environments; 12% when 30-year fixed rises more than 25 bp in the signing month.
Map to closings revenue — multiply net pending by average commission and expected close month using 55-day median lag.
Set agent desk caps — if projected net signings fall more than 5% q/q, freeze new agent onboarding; if PHSI rises two months in a row, reopen recruiting at 80% of prior pace until closings confirm.
Weekly MBA cross-check — if purchase apps diverge more than 8% from PHSI-implied trend, widen forecast confidence bands and delay hiring commits.

Example log line: “PHSI 74.2 (+1.6% m/m), South 78.1 (+2.1%). MBA purchase apps +3% w/w. Net Harbor signings forecast +4% next 60 days. Hold recruiting; closings should stabilize near 4.1M SAAR in Q3 if rates hold 6.5%.”

Technique decision table

Question	Use pending home sales (PHSI)	Use instead
Will resale closings turn in the next 1–2 months?	Yes — primary leading signal	MBA apps for higher frequency
What was last month’s closed volume?	No — use existing home sales SAAR	Existing home sales
Are builders adding supply?	No — resale signings ≠ construction	Housing starts, permits
Where are home prices heading?	Partial — volume lead, not price level	Case-Shiller, median price in existing sales
New-construction demand?	No	New home sales
Shelter CPI trajectory?	Indirect — turnover affects comps with long lags	Shelter CPI guide, rents

Common pitfalls

Treating PHSI level as a unit count — it is an index rebased to 2001, not thousands of homes.
Ignoring regional splits — national PHSI can mask a booming South and contracting West.
Assuming every pending sale closes — cancellation rates rise with rate volatility and appraisal gaps.
Comparing PHSI to new home sales — different universes (resale contracts vs new construction).
Single-month overreaction — weather and holiday distortions hit signings; use three-month trends.
Missing the rate window — signings reflect rates at contract; closings reflect earlier offers.
Skipping MBA purchase apps — weekly data catches inflections between monthly PHSI prints.
Equating PHSI to inventory — thin listings cap signings even when buyer intent is strong.

Production checklist

Log PHSI national m/m and y/y plus all four regional indexes each release.
Track median lag from internal signings to closings for your market (30–90 days).
Estimate fall-through rate by rate-volatility regime; haircut pending forecasts accordingly.
Cross-check weekly MBA purchase application index (purchase-only subindex).
Compare PHSI trend to prior two months of existing home sales closings for confirmation.
Monitor 30-year fixed rate at contract date, not closing date, for affordability reads.
Pair with months’ supply from existing sales for pricing-power context.
Document forecast error when PHSI and closings diverge — refine lag and cancellation assumptions.
Flag inventory-constrained metros where PHSI undercounts latent demand.
Review NAHB HMI only for new-supply implications, not as a PHSI substitute.

Key takeaways

Pending home sales count resale contracts signed, not closed — they lead existing home sales by roughly one to three months at turning points.
PHSI is an index (2001 = 100), seasonally adjusted, with regional Northeast/Midwest/South/West splits.
Cancellation rates and inventory constraints can break the lead-lag signal; cross-check MBA purchase apps and fall-through data.
Harbor Realty cut pipeline forecast error from 18% to 6% by staffing off PHSI plus regional conversion instead of lagging closings.
Use closings for coincident volume, pending for pipeline direction, starts and permits for future supply, Case-Shiller for prices.