Why It Matters
YoY is the workhorse change measurement in real estate analysis. "Home prices are up 4.2% YoY." "Unemployment is down 0.3pp YoY." "Rents grew 3.6% YoY." The power of YoY: because you're comparing the same month (or quarter) in both years, seasonality cancels out automatically — no statistical adjustment required. That makes YoY valid on every series, whether seasonally adjusted or not. It's the default comparison for metro-level LAUS unemployment (NSA), ZORI rent indexes, FHFA HPI levels, and any series where you want to strip the calendar without relying on a statistical model. The tradeoff: YoY has a 12-month horizon, so it lags real-time market shifts. Pair YoY (context) with MoM or QoQ on SA data (recent direction) for complete analysis.
At a Glance
- What it is: Change from the same month/quarter/week one year ago.
- Formula: `YoY % = (Current period − Period 12 months ago) / Period 12 months ago × 100`
- Why it strips seasonality: Same-month comparison naturally cancels out the seasonal pattern in both values.
- When to use it: Always valid — on SA or NSA data. The default for metro-level unemployment, rent indexes, home prices.
- Tradeoff: 12-month lag. A sharp market shift in the last 3 months is only partially reflected in YoY.
- Pair it with: MoM for SA data, QoQ for quarterly data, TTM (trailing 12 months) for rolling annual totals.
How It Works
The math cancels seasonality. When you compare March 2026 to March 2025, both values sit in the same season (spring). The seasonal component of each value is (roughly) identical, so when you subtract, the seasonal effect cancels. What remains is the trend change plus some noise. Unlike seasonal adjustment, which uses a statistical model, YoY does seasonality-stripping through direct comparison. That's why YoY is valid on NSA data — you don't need the model.
Why metro unemployment uses YoY by default. BLS LAUS publishes metro unemployment NSA — no seasonal adjustment at the metro grain. If you compare Austin metro November to Austin metro December, the seasonal pattern dominates (December always has more layoffs). The only way to read an underlying trend from NSA metro data is YoY: Austin December 2025 vs Austin December 2024. Same season, same geography, year apart. Whatever difference you see is real — not calendar.
When YoY shines and when it doesn't. YoY shines when: (1) the series is NSA and MoM is calendar-dominated, (2) you want annual context ("how is the market vs one year ago?"), (3) seasonality is strong and the statistical adjustment is unreliable. YoY lags when: (1) a major market shift happens in the last 3-6 months (YoY still reflects mostly the prior year), (2) base effects distort the signal (comparing a post-pandemic surge to the pandemic trough makes YoY look huge even if the recent trend is flat), (3) monthly data has plenty of observations and SA is available.
The base effect trap. A common YoY distortion: the prior-year base. Q2 2022 home sales were depressed by spiking mortgage rates. Q2 2023 home sales weren't much better, but comparing to Q2 2022 (the low base) made YoY look positive. The underlying market was flat; YoY was artificially improved by the weak comparison period. When using YoY, always check the base year's context. If the base is anomalous (pandemic, recession, supply shock), YoY readings overstate or understate the real trend.
Real-World Example
Carmen Vargas uses YoY for a metro underwriting decision.
Carmen is deciding between acquiring in Columbus OH and Indianapolis IN. Both metros have similar price points and rental demographics. She pulls three YoY readings for each metro from federal sources:
- Columbus OH, YoY through March 2026:
- LAUS unemployment: 3.4% this year vs 3.6% prior year → -0.2pp YoY (labor market tighter)
- ZORI rent: $1,420 vs $1,374 → +3.4% YoY (rent growing)
- FHFA HPI: 412.8 vs 395.2 → +4.5% YoY (prices appreciating)
- Indianapolis IN, YoY through March 2026:
- LAUS unemployment: 3.2% vs 3.1% → +0.1pp YoY (labor softening slightly)
- ZORI rent: $1,295 vs $1,268 → +2.1% YoY (rent growing slowly)
- FHFA HPI: 389.4 vs 378.1 → +3.0% YoY (prices appreciating less)
All six readings are YoY, all strip seasonality automatically, all valid comparisons.
The story: Columbus is outperforming Indianapolis on all three metrics. Tighter labor market (better YoY improvement), faster rent growth (3.4% vs 2.1%), stronger price appreciation (4.5% vs 3.0%). Each YoY reading answers a real question without calendar noise.
She does NOT rely solely on YoY. She cross-checks with MoM on the SA FRED mortgage rate series, checks QoQ on FHFA HPI for the most recent inflection, and reads the SAAR pace from the latest NAR existing home sales release. YoY gives context; the shorter horizons catch recent direction. Both matter.
Her decision: Columbus. But if YoY had been her only tool, she would have missed that the last 2 quarters of QoQ showed Indianapolis accelerating faster than Columbus. YoY is necessary but not sufficient.
Pros & Cons
- Strips seasonality automatically — valid on both SA and NSA data
- Standard across virtually every economic and real estate release
- Easy to compute and easy to communicate
- Provides annual context missing from short-horizon comparisons
- Doesn't require a statistical model
- 12-month lag — misses or underweights recent shifts
- Base-year effects can distort readings — comparing to an anomalous prior year (pandemic, recession) produces misleading YoY
- Doesn't answer "what's happening right now?" — use MoM or QoQ for that
- Single-point comparison — doesn't capture the path between the two periods
- Lagged series (ACS, IRS) publish YoY by default but with multi-year publication lag
Watch Out
- Base effects: YoY comparing 2023 to 2022 overstates improvement because 2022 was mortgage-rate-shock-depressed. Always check the base year for anomalies.
- "Headline YoY" hides pace shifts: A YoY of +4% could be a steady linear climb, OR could be a +8% first half followed by a 0% second half that averages to 4%. Check the monthly/quarterly path within the YoY window.
- Not a forecast: YoY is backward-looking. A 4% YoY today doesn't predict 4% next year. Use it for assessment, not projection.
- Apples-to-apples issue: YoY works only when the geography, definition, or methodology hasn't changed. BLS methodology revisions, Census geography reclassifications, or publisher data changes can break the YoY comparison.
- Revisions can move the base: Prior-year values get revised periodically. The YoY you calculated last month might read differently today if the base-year value changed.
Ask an Investor
The Takeaway
YoY is the universal standard for real estate change comparison — same-month-to-same-month, seasonality strips out automatically, valid on every series. Use it for metro LAUS unemployment (NSA), rent indexes, FHFA HPI level comparisons, and any monthly or quarterly series where you want the annualized context. Pair with MoM on SA data for recent direction, and with QoQ on quarterly data. Don't rely on YoY alone — it lags by 12 months and base effects can distort single readings. Check the path within the window. Data sources: FRED, BLS, Census, NAR — every one publishes YoY alongside the raw series.