Why It Matters
When you look at the 5-year price appreciation number on a large metro's data page — Dallas, Miami, Chicago, New York — odds are that number comes from FMHPI, not FHFA HPI. Here's why: FHFA splits 13 large metros into Metropolitan Divisions (NYC becomes 5 separate markets, LA becomes 2), so FHFA has no parent-MSA series for those metros. Freddie Mac publishes a parent-MSA series for every metro without splitting. That's the whole reason FMHPI exists as a common reference for large-metro appreciation. For investors, FMHPI answers "what did prices do in this metro over the last 5 years?" with monthly granularity, free data, and documented methodology. The catch: FMHPI only sees mortgages Freddie Mac insured or purchased — roughly half the conventional market. The other half is Fannie Mae (FHFA HPI uses both Fannie + Freddie combined) and private-label. Jumbo loans, cash sales, and non-conventional financing are missing. That doesn't make FMHPI wrong; it makes it a measure of the conventional conforming slice of the market.
At a Glance
- What it is: Monthly repeat-sales house price index covering the Freddie Mac mortgage portfolio.
- Who publishes it: Freddie Mac, a Government-Sponsored Enterprise under FHFA conservatorship.
- Grain: ZIP (for ZIPs with enough repeat-sales data), metro (parent-MSA level for all 387 metros), state, and national.
- Publication: Monthly series with ~6-week publication lag. March 2026 data typically publishes mid-May 2026.
- Coverage difference from FHFA HPI: FMHPI uses only Freddie-insured mortgages; FHFA HPI uses Fannie + Freddie combined. For 13 large MSAs, FMHPI has parent-metro data where FHFA does not.
- Historical series: Starts January 1975 at national grain. Metro coverage expands over time as Freddie's portfolio grew.
How It Works
Repeat-sales methodology, explained once. FMHPI uses the same underlying technique as FHFA HPI and Case-Shiller: identify homes that sold at least twice (or were refinanced with a new appraisal at least twice) and compute the price change between transactions, controlling for that same property's characteristics. By comparing the same property across time, repeat-sales controls for quality differences — you're not comparing a 2,000 sq ft home to a 3,500 sq ft home, you're comparing one specific 2,400 sq ft home to itself over 5 years. Weighted across thousands of properties in each geography, the result is a monthly index showing how prices moved for the typical home in that market. Freddie publishes both a raw and seasonally adjusted series; for investor analysis, use the seasonally adjusted series unless you specifically need month-to-month seasonal patterns.
Why FMHPI exists alongside FHFA HPI. Both indexes are GSE-sourced, both use repeat sales, both cover overlapping portfolios. The difference: FHFA combines Fannie + Freddie into one index; Freddie publishes its own. For most metros, FHFA HPI and FMHPI track each other closely — FHFA is the broader measure. But for the 13 MSAs where FHFA splits the metro into Metropolitan Divisions — NYC, LA, Chicago, Dallas, Washington DC, Philadelphia, Miami, Atlanta, Boston, San Francisco, Detroit, Seattle, Tampa — FHFA publishes only the MD-level series, not a parent-MSA series. If you want "the 5-year price change for the Dallas-Fort Worth MSA," FHFA makes you sum 2 MDs, weighted by population, with assumptions. Freddie publishes a parent Dallas-Fort Worth MSA series directly. That's why REI Prime uses FMHPI for those 13 metros — one clean number per parent metro.
What FMHPI covers and what it misses. Coverage is limited to properties with at least one Freddie Mac-insured mortgage on record during the observation window. What that includes: ~60-70% of the U.S. conventional conforming mortgage market — owner-occupants buying single-family homes at or below the conforming loan limit. What it misses: (1) jumbo loans above the conforming limit (which is higher in expensive markets like San Francisco or New York), (2) all-cash purchases — no mortgage, no data point, (3) non-conventional financing (FHA, VA, USDA, portfolio lenders), (4) investor-owned properties that never refinanced, (5) Fannie Mae-insured properties (these go into FHFA HPI but not FMHPI). The shorthand: FMHPI is the Freddie conforming slice. In metros where jumbo and cash activity are large (coastal California, Manhattan, Miami ultra-luxury), FMHPI undercounts the high end.
How FMHPI appears in investor analysis. Three common uses. First, 5-year appreciation tracking: pull the FMHPI series for your target metro, compute total appreciation over 5 years, compare to peer metros. Second, YoY trend monitoring: is monthly FMHPI still growing, or has it turned negative? Third, methodology cross-check: if Zillow ZHVI says one thing and FMHPI says another, dig in. The two measure different slices (FMHPI = Freddie conforming, ZHVI = all home values via algorithmic AVM). Disagreement usually reveals something about the market mix — for example, FMHPI flat while ZHVI up means the high-end was the story. FMHPI distributes through Freddie's research portal and through FRED as series like `FRBKC...HPI`. Free, monthly, documented.
Real-World Example
Miguel Herrera compares 5-year price appreciation across three metros using FMHPI.
Miguel is considering where to deploy capital next. He has three metros on his short list: Dallas-Fort Worth TX, Atlanta GA, and Cleveland OH. He pulls 5-year FMHPI for each from Freddie's research portal.
- Dallas-Fort Worth-Arlington MSA (FMHPI, 5-year through March 2026): +51.2% total appreciation
- Atlanta-Sandy Springs-Alpharetta MSA (FMHPI, 5-year): +64.8% total appreciation
- Cleveland-Elyria MSA (FMHPI, 5-year): +38.1% total appreciation
On pure price appreciation, Atlanta leads. But Miguel knows FMHPI is the conforming slice of the conventional market — it might be overstating or understating the full picture depending on what the mix looks like in each metro.
He cross-references with FHFA HPI. For Atlanta and Cleveland (where FHFA has parent-MSA series), FHFA shows +63.4% and +37.9% — within 1-2 points of FMHPI. That confirms the FMHPI read is representative for those metros. For Dallas (where FHFA only has the two Metropolitan Divisions — Dallas-Plano-Irving and Fort Worth-Arlington-Grapevine), he averages FHFA weighted by population and gets +50.8%. FMHPI's +51.2% matches. Clean read.
He cross-references with Zillow ZHVI for the high-end check. Atlanta ZHVI is +68.2% over the same window — 4 points higher than FMHPI. That tells Miguel the Atlanta jumbo and higher-end segment ran hotter than the conforming slice. For an investor focused on sub-$500K workforce housing, FMHPI is the more relevant signal. For someone chasing luxury, ZHVI tells the better story.
His decision: Atlanta wins on appreciation momentum, but Cleveland's lower FMHPI signals cheaper entry with room for catch-up. Dallas sits in the middle with the cleanest FMHPI/FHFA consistency. He adds Cleveland to his acquisition pipeline and Dallas to the watch list.
Pros & Cons
- Free, monthly, documented — no paywall, no API key, direct download from Freddie
- Parent-MSA coverage for all 387 metros — no MD-splitting problem like FHFA HPI
- Repeat-sales methodology controls for property-level quality differences
- Historical depth to January 1975 at national grain
- Distributed through FRED alongside other federal and GSE series, making integration easy
- Cross-validates FHFA HPI — divergence is a diagnostic signal, not a problem
- Limited to Freddie-insured conforming mortgages — misses ~40% of the conventional market (Fannie side) and all cash/jumbo/non-conventional activity
- ZIP-level coverage is spotty for low-volume ZIPs — small markets have noisy or missing series
- ~6-week publication lag — slower than Zillow ZHVI (2-week lag) for real-time monitoring
- Methodology revisions periodically reshape the historical series — the Freddie CHPI (Choice HPI) release in 2021 updated earlier data
- Not universally cited — some lenders and analysts still default to FHFA HPI or Case-Shiller, creating version confusion
Watch Out
- Coverage gap at the high end: FMHPI doesn't see jumbo loans. In metros with heavy high-end activity (San Francisco, Manhattan, Miami Beach), the FMHPI number is the conforming-segment number, not the market-wide story. Pair with ZHVI or Case-Shiller for the high-end read.
- Methodology changes move history: Freddie has revised methodology at least twice (most recently the CHPI transition in 2021). Historical series you pulled in 2020 may read differently today. Use the most recent vintage.
- ZIP coverage is noisy below ~100 repeat sales/year: If you're using FMHPI at ZIP level, check the repeat-sales count. Below ~100/year, the series has wide confidence intervals and can swing 5-10% from noise alone.
- Watch for Fannie-heavy markets: Some metros — particularly military towns, rural conforming markets — lean more toward Fannie than Freddie. FMHPI undercounts them. FHFA HPI (Fannie + Freddie combined) gives the broader read.
- Not the same as Case-Shiller: FMHPI and S&P Case-Shiller are both repeat-sales but cover different portfolios and use different weighting schemes. For the same metro, FMHPI and Case-Shiller can diverge 2-5 points over long windows. Cite one or the other, never mix.
Ask an Investor
The Takeaway
FMHPI is the cleanest source for parent-MSA price appreciation across large U.S. metros — the series investors cite when FHFA HPI can't because FHFA splits the metro into Metropolitan Divisions. For the 13 affected MSAs (NYC, LA, Chicago, Dallas, DC, Philly, Miami, Atlanta, Boston, SF, Detroit, Seattle, Tampa), FMHPI is the canonical choice. For all other metros, FHFA HPI is the broader measure (Fannie + Freddie), but FMHPI still works as a secondary cross-check. Pull FMHPI from Freddie's research portal or FRED, compare 5-year and 1-year changes to peer metros, and cross-validate with FHFA HPI or ZHVI when the number looks surprising. Free, monthly, 6-week lag. The underlying index belongs to Freddie Mac, the GSE under FHFA conservatorship.