Price-to-Rent vs 1991–2000: +37.8% (Jul 2026)

Is the housing market healthy?

Housing tab score

4/10Alert

worst block 4.00

Housing is on alert: buying with a mortgage costs 33% more relative to hourly pay than in 1991–2000 (30-year mortgage at 7.35%); also, homes cost 38% more relative to rents than in 1991–2000 and existing-home sales per person are 29% below 2017–2019. Holding up: recorded delinquency is low (0.99% of mortgage balances are 90+ days delinquent).

The 5 blocks

AccessWhat do buying and renting cost relative to hourly pay?4.0 ActivityAre homes being built and sold at a normal pace?5.5 SupplyIs there too much or too little housing?7.3 PricesAre prices outpacing inflation or drifting away from rents?6.5 CreditAre mortgages being paid and is debt contained?10.0

What would turn it red

Delinquency > 5% of balances (today 0.99%): the score could not go above 2.
Debt > 90% of income (today 59.6%): it could not go above 4.
Buying, relative to pay, more than 40% above 1991–2000 (today 33%): access would drop to 3.

The score is the number of bars lit: 10 is healthiest and 1 is worst. The spark marks where it is today.

Price-to-Rent vs 1991–2000

0

%

-0.2

%

-1.5

%

-0.4

%

Last

37.8 %

37.8 %

All-time: -3.8 – 47.5 % · since Jan 91 · now at 81% of that range

Price-to-Rent vs 1991–2000 stands at 37.8% (Jul 1, 2026), unchanged from 37.8% on Jun 1, 2026. Since Jul 1, 2025 it has moved -0.4 pp.

The FHFA home price index divided by CPI rent, versus its 1991–2000 average. Far above means a stretched valuation (2006: +38%, 2022: +47%). Calculated by us.

Loading chart…

Y axis zoomed to the selected range, so week-to-week moves are visible. It does not start at zero.

Source: FHFA and BLS via FRED — own calculation (HPIPONM226S / CUSR0000SEHA, vs. 1991–2000)

Housing Indicators

Sort:

Indicator	Last	Previous	Change	Highest	Lowest	Range %	Gauge	Date
Home Cost vs Hourly Pay idx▲ 2024	133.47	123.64	+9.83idx · 1W	143.17	58.43	89%	—	Oct 26
Housing Starts K	1,221	1,288	-67-5.2% · 1M	2,494	478	37%	—	Sep 26
Existing Home Sales K	3,990	3,930	+60+1.5% · 1M	4,270	3,930	18%	—	Sep 26
Months' Supply of New Homes months	8.5	9	-0.5months · 1M	12.2	3.3	58%	—	Aug 26
Homeowner Vacancy Rate %	1.2	1.1	+0.1% · 3M	2.9	0.7	23%	—	Apr 26
Rental Vacancy Rate %	7.3	7.3	0% · 3M	11.1	5	38%	—	Apr 26
Real Home Prices YoY (FHFA/CPI) %	-0.7	-1.1	+0.4% · 1M	12.9	-13.9	49%	—	Jul 26
Price-to-Rent vs 1991–2000 %	37.8	37.8	0% · 1M	47.5	-3.8	81%	—	Jul 26
Mortgage Delinquency 90+ Days (NY Fed) %	0.99	1.09	-0.1% · 3M	8.89	0.37	7%	—	Apr 26
Mortgage Debt to Disposable Income %	59.6	59.6	0% · 3M	101.5	16.2	51%	—	Apr 26

The 10 indicators

Access: What do buying and renting cost relative to hourly pay?

average 4.0 · Alert

Cost of buying and renting

Home Cost vs Hourly Pay

4/10Alert

133.47week of Oct 2026 22 · rent 102.77 · 30-year mortgage at 7.35% (Oct 2026 average)

Buying with a mortgage costs 33% more relative to hourly pay than in 1991–2000; renting, 3% more.

See scale and last 10 years

The monthly cost of a 30-year mortgage on the FHFA price index, and CPI rent, relative to average hourly pay. 100 is the 1991–2000 average; 133 means 33% more demanding, not a payment of 33% of income. It is not the payment of a specific family: it does not measure unemployment, hours or other income.

Measures: Cost relative to hourly pay (1991–2000 = 100).

Oct 2016last 10 yearsOct 2026

Score	Buying with a mortgage	Renting
10	100.00 or less	100.00 or less
9	100.01 to 105.00	100.01 to 103.00
8	105.01 to 110.00	103.01 to 106.00
7	110.01 to 115.00	106.01 to 109.00
6	115.01 to 120.00	109.01 to 112.00
5	120.01 to 130.00	112.01 to 115.00
4	130.01 to 140.00	115.01 to 120.00
3	140.01 to 150.00	120.01 to 125.00
2	150.01 to 165.00	125.01 to 135.00
1	more than 165.00	more than 135.00

Buying with a mortgage: 133.47 → 4 · Renting: 102.77 → 9. The worse one rules.

FHFA (HPIPONM226S), Freddie Mac (MORTGAGE30US) and BLS (CUSR0000SEHA, AHETPI); own calculation · FRED

Activity: Are homes being built and sold at a normal pace?

average 5.5 · Caution

Housing starts

Housing Starts

7/10Caution▼ below normal

−19.8%Sep 2026 · 1,221K a year · −4.1% over a year

Builders are starting 19.8% fewer homes per person than normal (4.7 per 1,000) and 4.1% fewer than a year ago.

See scale and last 10 years

Homes whose construction begins, 3-month average. Normal is 4.7 per 1,000 people a year (the 1991–2019 median). Building far above that is bad too (2005).

Measures: Starts per 1,000 people vs. normal and vs. a year ago (%).

Oct 2016last 10 yearsSep 2026

Score	Per person vs. normal · Above	Per person vs. normal · Below	Vs. a year ago
10	0.0 to 10.0	−5.0 to −0.1	−3.0 or more
9	10.1 to 20.0	−10.0 to −5.1	−6.0 to −3.1
8	20.1 to 25.0	−15.0 to −10.1	−10.0 to −6.1
7	25.1 to 30.0	−20.0 to −15.1	−15.0 to −10.1
6	30.1 to 35.0	−25.0 to −20.1	−20.0 to −15.1
5	35.1 to 40.0	−30.0 to −25.1	−25.0 to −20.1
4	40.1 to 50.0	−40.0 to −30.1	−30.0 to −25.1
3	50.1 to 60.0	−50.0 to −40.1	−35.0 to −30.1
2	60.1 to 75.0	−60.0 to −50.1	−40.0 to −35.1
1	more than 75.0	less than −60.0	less than −40.0

Per person vs. normal: −19.8 → 7 · Vs. a year ago: −4.1 → 9. The worse one rules.

Census (HOUST) and BEA (POPTHM) · FRED

Existing-home sales

Existing Home Sales

4/10Alert

−29.2%Sep 2026 · 3.99M a year (NAR)

Existing-home sales per person are 29.2% below 2017–2019.

See scale and last 10 years

Existing homes sold (NAR), per person, vs. 2017–2019. FRED only publishes 13 months: before that, the history uses home-purchase mortgages from the NMDB (FHFA), which leave out cash purchases.

Measures: Sales per person vs. 2017–2019 (%).

Oct 2016last 10 yearsSep 2026

Score	Existing-home sales (NAR)	Purchase mortgages (NMDB)
10	−5.0 or more	−5.0 or more
9	−10.0 to −5.1	−10.0 to −5.1
8	−14.0 to −10.1	−14.0 to −10.1
7	−18.0 to −14.1	−18.0 to −14.1
6	−22.0 to −18.1	−22.0 to −18.1
5	−26.0 to −22.1	−26.0 to −22.1
4	−30.0 to −26.1	−30.0 to −26.1
3	−35.0 to −30.1	−35.0 to −30.1
2	−40.0 to −35.1	−40.0 to −35.1
1	less than −40.0	less than −40.0

Existing-home sales (NAR): −29.2 → 4 · Purchase mortgages (NMDB): no data → —. The worse one rules.

NAR via FRED (EXHOSLUSM495S) and FHFA (NMDB, financed purchases) · FRED

Supply: Is there too much or too little housing?

average 7.3 · Caution

New-home months of supply

Months' Supply of New Homes

4/10Alert

8.5 monthsAug 2026

There are 8.5 months of new homes for sale: too many.

See scale and last 10 years

How many months it would take to sell the new homes for sale at the current pace. Above 7–8 there are too many; it reached 12.2 in 2009.

Measures: Months of supply.

Sep 2016last 10 yearsAug 2026

Score	Months of supply
10	5.0 or less
9	5.1 to 5.5
8	5.6 to 6.0
7	6.1 to 6.5
6	6.6 to 7.0
5	7.1 to 7.5
4	7.6 to 8.5
3	8.6 to 9.5
2	9.6 to 11.0
1	more than 11.0

Census (MSACSR) · FRED

Homeowner vacancy

Homeowner Vacancy Rate

8/10Healthy▼ below normal

1.2%Q2 2026

1.2% of owner homes are vacant for sale (normal is 1.6%).

See scale and last 10 years

Normal is 1.6%. Above that there are too many homes (2008: 2.9%); far below, too few. 2020 Q2–Q3 do not count.

Measures: % of owner homes vacant for sale.

Jul 2016last 10 yearsApr 2026

Score	Above	Below
10	1.6 to 1.8	1.4 to 1.5
9	1.9	1.3
8	2.0	1.1 to 1.2
7	2.1	1.0
6	2.2	0.9
5	2.3 to 2.4	0.8
4	2.5 to 2.6	0.7
3	2.7	0.6
2	2.8	0.5
1	more than 2.8	less than 0.5

Census (RHVRUSQ156N) · FRED

Rental vacancy

Rental Vacancy Rate

10/10Healthy

7.3%Q2 2026

7.3% of rentals are vacant (normal is 7.5%).

See scale and last 10 years

Normal is 7.5%. 2020 Q2–Q3 do not count.

Measures: % of rentals vacant.

Jul 2016last 10 yearsApr 2026

Score	Above	Below
10	7.5 to 8.2	7.0 to 7.4
9	8.3 to 8.7	6.7 to 6.9
8	8.8 to 9.2	6.4 to 6.6
7	9.3 to 9.7	6.1 to 6.3
6	9.8 to 10.0	5.8 to 6.0
5	10.1 to 10.2	5.5 to 5.7
4	10.3 to 10.5	5.2 to 5.4
3	10.6 to 10.7	4.9 to 5.1
2	10.8 to 11.0	4.5 to 4.8
1	more than 11.0	less than 4.5

Census (RRVRUSQ156N) · FRED

Prices: Are prices outpacing inflation or drifting away from rents?

average 6.5 · Caution

Real home prices

Real Home Prices YoY (FHFA/CPI)

10/10Healthy

−0.7%Jul 2026

Home prices are down 0.7% in a year after inflation.

See scale and last 10 years

Healthy is a home rising roughly in line with other prices. If a month's CPI is missing, that month shows the nominal change, marked as such.

Measures: FHFA price vs. a year ago, after CPI (%).

Aug 2016last 10 yearsJul 2026

Score	After inflation · Above	After inflation · Below	Nominal (CPI missing) · Above	Nominal (CPI missing) · Below
10	0.0 to 3.0	−2.0 to −0.1	4.0 to 6.0	2.0 to 3.9
9	3.1 to 5.0	−4.0 to −2.1	6.1 to 7.0	0.5 to 1.9
8	5.1 to 7.0	−6.0 to −4.1	7.1 to 8.0	−1.0 to 0.4
7	7.1 to 9.0	−8.0 to −6.1	8.1 to 9.0	−2.0 to −1.1
6	9.1 to 11.0	−10.0 to −8.1	9.1 to 10.0	−3.0 to −2.1
5	11.1 to 13.0	−12.0 to −10.1	10.1 to 11.0	−4.0 to −3.1
4	13.1 to 16.0	−15.0 to −12.1	11.1 to 12.0	−5.0 to −4.1
3	16.1 to 20.0	−18.0 to −15.1	12.1 to 14.0	−6.5 to −5.1
2	20.1 to 25.0	−22.0 to −18.1	14.1 to 16.0	−8.0 to −6.6
1	more than 25.0	less than −22.0	more than 16.0	less than −8.0

After inflation: −0.7 → 10 · Nominal (CPI missing): no data → —. The worse one rules.

FHFA (HPIPONM226S) and BLS (CPIAUCSL); own calculation · FRED

Price-to-rent

Price-to-Rent vs 1991–2000

3/10Alert

+37.8%Jul 2026

Homes cost 38% more relative to rents than in 1991–2000.

See scale and last 10 years

The FHFA index divided by CPI rent. Far above the 1990s means a stretched valuation (2006: +38%, 2022: +47%).

Measures: Price-to-rent vs. 1991–2000 (%).

Aug 2016last 10 yearsJul 2026

Score	Price-to-rent vs. 1991–2000 (%)
10	5.0 or less
9	5.1 to 10.0
8	10.1 to 15.0
7	15.1 to 20.0
6	20.1 to 25.0
5	25.1 to 30.0
4	30.1 to 35.0
3	35.1 to 40.0
2	40.1 to 45.0
1	more than 45.0

FHFA (HPIPONM226S) and BLS (CUSR0000SEHA); own calculation · FRED

Credit: Are mortgages being paid and is debt contained?

average 10.0 · Healthy

Mortgage delinquency

Mortgage Delinquency 90+ Days (NY Fed)

10/10Healthy

0.99%Q2 2026

0.99% of mortgage balances are 90+ days delinquent.

See scale and last 10 years

New York Fed (Consumer Credit Panel / Equifax). It reached 8.9% in 2010. Before 1999, banks' 30-day delinquency (Fed), with its own cuts.

Measures: % of mortgage balances 90+ days delinquent.

Jul 2016last 10 yearsApr 2026

Score	New York Fed (90+ days)	Banks, before 1999 (30+ days)
10	1.00 or less	2.00 or less
9	1.01 to 1.25	2.01 to 2.50
8	1.26 to 1.50	2.51 to 3.00
7	1.51 to 2.00	3.01 to 3.50
6	2.01 to 2.50	3.51 to 4.00
5	2.51 to 3.00	4.01 to 5.00
4	3.01 to 4.00	5.01 to 6.50
3	4.01 to 5.00	6.51 to 8.00
2	5.01 to 7.00	8.01 to 10.00
1	more than 7.00	more than 10.00

New York Fed (90+ days): 0.99 → 10 · Banks, before 1999 (30+ days): no data → —. The worse one rules.

New York Fed Consumer Credit Panel / Equifax (own calculations); before 1999, Fed (DRSFRMACBS) · FRED

Mortgage debt to income

Mortgage Debt to Disposable Income

10/10Healthy

59.6%Q2 2026

Household mortgage debt equals 59.6% of a year of disposable income (61% in the 1990s).

See scale and last 10 years

All household mortgage debt relative to a year of disposable income. In the 1990s, 61%; in 2007, 101%.

Measures: Mortgage debt / disposable income (4-quarter average, %).

Jul 2016last 10 yearsApr 2026

Score	Mortgage debt / disposable income (4-quarter average, %)
10	62.0 or less
9	62.1 to 65.0
8	65.1 to 68.0
7	68.1 to 71.0
6	71.1 to 75.0
5	75.1 to 80.0
4	80.1 to 85.0
3	85.1 to 90.0
2	90.1 to 95.0
1	more than 95.0

Fed (HHMSDODNS) and BEA (DPI); own calculation · FRED

If this gauge had existed since 1991

Each month's score, with today's rules and cuts. Sales start in 1998 (with the NMDB) and New York Fed delinquency in 1999: before that, the score uses the indicators available, at least half. Vacancies do not count in 2020 Q2–Q3. The history ends in the last month with all ten indicators; today's score combines the latest data for each. Gray bands are recessions.

A2005 · mortgage debt at 91% of income and homes 34% pricier relative to rents
B2009 · starts fall 49% in a year and delinquency reaches 5.6%
C2010 · 8.9% of mortgage balances 90 days unpaid
D2021 · homes 42% pricier relative to rents than in the 1990s
E2023 · buying with a mortgage costs 41% more relative to pay than in the 1990s
FToday · score 4: housing is on alert

How it is calculated

The rules

Tables of 9 cuts. Two-sided ones also score lower when they fall short. With two readings, the worse one counts.
Each block is the average of its indicators and the score is that of the worst block. Access, supply and prices alone do not go below 3: only activity or credit lead to 1–2.
Debt > 90% of income → at most 4. Delinquency > 5% → at most 2; > 7% → 1.
Today's score needs all ten up to date; the history, half.

Deliberate choices

Access is measured against hourly pay, not as a family's payment: 133 means 33% more demanding than in 1991–2000, not 33% of income, and it does not measure unemployment, hours or other income.
Sales are NAR's per person; FRED only publishes 13 months, and before that the history uses NMDB home-purchase mortgages (FHFA), which leave out cash purchases.
Delinquency is the New York Fed's (90+ days); before 1999, banks' 30-day rate, with its own cuts, which does not trigger the alarm rule.
Vacancies in 2020 Q2–Q3 do not count: the survey lost responses during COVID.
Debt is measured against the 4-quarter average of disposable income, so a one-off government payment does not move it.
It describes the housing market from the latest data. It does not forecast prices and is not a buy or sell signal.

Sources: FHFA (prices and NMDB, CC0; This product uses FHFA data but is neither endorsed nor certified by FHFA), Freddie Mac (mortgage rate), BLS (wages, rent and CPI), Census (construction, supply and vacancies), BEA (population and income), NAR (existing-home sales), Fed (mortgage debt and bank delinquency) and New York Fed Consumer Credit Panel / Equifax (delinquency; own calculations), via FRED and official sources. Recessions: NBER. Not investment advice.