Data Sources &
Methodology
数据来源与
算法说明
Every number on this site comes from an Irish government agency or official industry body. Every calculation formula is documented here. If you find an error, please tell us. 本站所有数据均来自爱尔兰政府机构或官方行业组织。所有计算公式均在此页面完整记录。如发现错误，欢迎联系我们。
We show our work 我们公开所有工作过程
Too many property tools hide their data sources and formulas behind marketing language. We think investors deserve better. 太多房产工具将数据来源和计算公式隐藏在营销话术背后。我们认为投资者值得更好的对待。
Government sources only 仅使用政府数据
All 13 datasets come from Irish government agencies (CSO, Revenue, SEAI, RTB), statutory bodies (PSRA, BPFI) or the ECB. No paid vendors. No proprietary data feeds. Every source is freely accessible online. 全部13个数据集来自爱尔兰政府机构（CSO、Revenue、SEAI、RTB）、法定机构（PSRA、BPFI）或欧央行。不使用付费数据商，不依赖专有数据。每个来源均可在线免费访问。
Every formula is public 每个公式均公开
Gross yield, net yield, crime rate, LPT estimate — every metric is calculated with a documented formula. You can reproduce any number on this site using the source data and the formulas below. 毛收益率、净收益率、犯罪指数、地方物业税估算——每项指标均有记录在案的公式。您可以使用原始数据和以下公式重现本站任何数字。
Limitations disclosed upfront 局限性提前披露
BER ratings use software assumptions, not measured energy. Daft rents are asking prices. Crime data is by Garda division. We name every known limitation next to every data source — not buried in small print. BER评级使用软件假设而非实测能耗；Daft租金为挂牌价；犯罪数据按警区而非县统计。我们在每个数据来源旁边明确列出所有已知局限——而非藏在小字免责声明中。
Data Sources 数据来源
Organised by category. Click any source link to access the original dataset directly. 按类别整理。点击任意来源链接可直接访问原始数据集。
A — Property Market Data A — 房产市场数据
B — CSO Government Statistics B — CSO政府统计数据
C — Economic & Financial Data C — 经济与金融数据
D — Tax & Regulation D — 税务与法规
Methodology 算法说明
Click each section to expand the full formula and explanation. Every metric on the site is derived from one of these calculations. 点击每个板块可展开完整公式和说明。本站每项指标均来自以下计算之一。
The Property Price Register records sale prices as declared for stamp duty purposes. For new residential builds, this price is VAT-exclusive — the developer pays VAT (at 13.5%) separately. Secondhand properties are VAT-exempt and used as-is. To make prices comparable, we apply a 13.5% uplift to all new builds.房产价格登记册按印花税申报目的记录成交价。对于新建住宅，该价格不含增值税——开发商单独缴纳13.5%增值税。二手房产免征增值税，直接使用。为使价格具有可比性，我们对所有新建住宅应用13.5%上调。
adjusted_price = ppr_price × 1.135
# Secondhand (VAT-exempt): no adjustment
adjusted_price = ppr_price
VAT rate reference: Reduced VAT rate of 13.5% applies to residential construction in Ireland (Revenue.ie).增值税参考：爱尔兰住宅建设适用13.5%的降低增值税税率（Revenue.ie）。
The raw PPR table (property_sales) contains ~775,000 rows and is never modified. All analysis queries via a database view clean_residential_sales that applies five filters in sequence:原始PPR表（property_sales）包含约77.5万行，永不修改。所有分析通过数据库视图clean_residential_sales查询，该视图按序应用五层过滤：
-
Price floor — remove transactions below €50,000价格下限——去掉低于€50,000的交易
Family transfers where a parent sells to a child for €1 or €100 do not reflect market prices. The €50,000 threshold captures these nominal transfers while preserving genuine low-value market transactions. 父母以€1或€100将房产转给子女的家庭内部转让不代表市场价格。€50,000的阈值捕获这些名义转让，同时保留真实的低价市场交易。 -
Price ceiling — remove transactions above €3,000,000价格上限——去掉高于€3,000,000的交易
245 transactions totalling €3.06 billion. These are typically institutional bulk purchases of entire apartment blocks or large development schemes. Retaining them would significantly distort county medians, especially in Dublin. 245笔交易，总金额达€30.6亿。这类交易通常是机构投资者批量购买整栋公寓楼或大型开发项目。保留会严重拉高郡级中位数，尤其在都柏林。 -
PSRA not-full-market flag — remove known non-market transactionsPSRA非市场价标记——去掉已知非市场价交易
The PPR source data includes anot_full_marketfield where PSRA flags transactions they know were not at market value (e.g. receivership sales, social housing transfers). This filter removed 33,301 records. PPR原始数据包含not_full_market字段，PSRA会标记已知非市场价交易（如破产清盘销售、社会住房转让）。此过滤器去掉了33,301条记录。 -
VAT adjustment — normalise new build prices增值税调整——标准化新建住宅价格
New builds (wheredescription LIKE 'New%'andvat_exclusive = true) are recorded at VAT-exclusive prices in the PPR. We multiply by 1.135 to produceadjusted_price. All analysis usesadjusted_price, not the raw price. Secondhand properties are VAT-exempt and used as-is. 新建住宅（description LIKE 'New%'且vat_exclusive = true的记录）在PPR中以不含增值税的价格登记。我们乘以1.135生成adjusted_price字段。所有分析使用adjusted_price而非原始价格。二手房产免征增值税，直接使用。 -
Deduplication — remove repeated entries for the same address and date去重——去掉同一地址同一日期的重复记录
UsingDISTINCT ON (address, date_clean), retaining the lower-priced entry. 870 duplicate rows removed: 496 were exact price repeats (data entry duplication), 299 had slightly different prices (possible correction filings or multi-unit properties). 使用DISTINCT ON (address, date_clean)，保留价格较低的记录。共去除870条重复行：496组价格完全相同（数据录入重复），299组价格略有不同（可能是修正记录或多单元物业）。
CREATE VIEW clean_residential_sales AS
SELECT DISTINCT ON (address, date_clean)
*,
CASE WHEN description LIKE 'New%' AND vat_exclusive
THEN price * 1.135
ELSE price END AS adjusted_price
FROM property_sales
WHERE price >= 50000
AND price <= 3000000
AND not_full_market IS NOT TRUE
ORDER BY address, date_clean, price ASC;
property_sales: ~775,000 rows (never modified). After all 5 filters: 699,653 transactions used for analysis. The raw data is always available for independent verification.
原始表property_sales：约77.5万行（永不修改）。经全部5层过滤后：699,653笔交易用于分析。原始数据随时可供独立验证。
We use the statistical median (50th percentile) rather than the arithmetic mean. The median is more robust to outliers — a handful of super-prime transactions can inflate the mean significantly in low-volume markets.我们使用统计中位数（第50百分位数）而非算术平均数。中位数对异常值更稳健——少数超豪华交易可能在低交易量市场中大幅抬高均值。
median_price = prices[len(prices) ÷ 2]
Dublin postcodes: Minimum 10 transactions required. Counties: Minimum 20 transactions required for a county-year combination to be included.都柏林邮区：至少需要10笔交易。郡：郡-年组合至少需要20笔交易才被纳入。
Dublin transactions in the PPR can be assigned to 22 postcodes (D1–D24, excluding D19, D21, D23, with D6W as a special case). We use a two-method approach to maximise coverage:PPR中的都柏林交易可分配到22个邮区（D1–D24，不含D19、D21、D23，D6W为特殊情况）。我们使用双重方法最大化覆盖率：
routing_key = eircode[:3].upper()
if routing_key == "D6W": return "Dublin 6W"
if routing_key[0] == "D" and routing_key[1:].isdigit():
if int(routing_key[1:]) in VALID_POSTCODES: return "Dublin " + n
# Method 2: Address regex fallback (pre-2021 and missing Eircodes)
match = re.search(r"Dublin\s+(\d{1,2})(W?)", address, re.IGNORECASE)
if match: return "Dublin " + match.group(1) + match.group(2)
VALID_POSTCODES = {1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,20,22,24}
Two variants of gross yield are calculated, using different price and rent sources:使用不同的价格和租金来源计算两种毛收益率变体：
gross_yield_daft = (daft_monthly_rent_2bed × 12) / daft_asking_price × 100
# Variant B: PPR yield (shown as secondary)
gross_yield_ppr = (rtb_monthly_rent × 12) / ppr_median_price × 100
Daft yield uses Daft.ie Q4 2024–Q2 2025 2-bedroom asking rent and Daft asking price. PPR yield uses RTB/ESRI registered rent and PPR 2025 VAT-adjusted median. Both are gross figures — net yield will be materially lower after fees, vacancy and tax.Daft收益率使用Daft.ie 2024年Q4–2025年Q2两居室挂牌租金和Daft要价。PPR收益率使用RTB/ESRI登记租金和PPR 2025年增值税调整后中位价。两者均为毛收益——扣除费用、空置率和税费后净收益将明显偏低。
The Net Return Calculator applies all landlord costs and Irish income tax to arrive at true net annual income. Steps:净回报计算器对所有房东成本和爱尔兰所得税应用完整计算，得出真实净年收入。步骤：
Step 2 vacancy_loss = gross_rent × vacancy_rate
Step 3 effective_rent = gross_rent × (1 − vacancy_rate)
Step 4 mgmt_fee = effective_rent × mgmt_rate
Step 5 maintenance = property_price × maint_rate
Step 6 insurance = f(price) # see table below
Step 7 taxable_income = effective_rent − mgmt_fee − maintenance − insurance
Step 8 income_tax = taxable_income × tax_rate
# Non-resident: 20% | Standard: 20% | Higher rate: 40%
Step 9 lpt = f(property_price) # Revenue 2022 bands
Step 10 net_income = taxable_income − income_tax − lpt
Step 11 net_yield = net_income / property_price × 100%
Insurance estimate (Step 6):保险费估算（第6步）：
| Property Price房产价格 | Annual Insurance年保险费 |
|---|---|
| < €300,000 | €600 |
| €300,000 – €600,000 | €800 |
| > €600,000 | €1,200 |
LPT bands (Revenue 2022 revaluation, annual charge):LPT分档（Revenue 2022年重估，年度收费）：
| Property Value Band房产估值分档 | Annual LPT年度LPT |
|---|---|
| ≤ €200,000 | €90 |
| €200,001 – €262,500 | €225 |
| €262,501 – €350,000 | €315 |
| €350,001 – €437,500 | €405 |
| €437,501 – €525,000 | €495 |
| €525,001 – €612,500 | €585 |
| €612,501 – €700,000 | €675 |
| €700,001 – €787,500 | €765 |
| €787,501 – €875,000 | €855 |
| €875,001 – €962,500 | €945 |
| €962,501 – €1,050,000 | €1,035 |
| > €1,050,000 | Self-assessed (~€1,350+ est.)自行评估（~€1,350+估算） |
CSO CJQ06 reports crime by 28 Garda divisions. 6 divisions cover multiple counties and must be split. Dublin is the sum of 6 Dublin Metropolitan Region (DMR) divisions.CSO CJQ06按28个警区报告犯罪数据。其中6个警区覆盖多个郡，需要拆分。都柏林为6个都柏林大都市区（DMR）警区之和。
crimes_county_A = crimes_division × pop_A / (pop_A + pop_B)
crimes_county_B = crimes_division × pop_B / (pop_A + pop_B)
# Dublin = sum of 6 DMR divisions
crimes_Dublin = Σ(DMR_Northern, DMR_East, DMR_N.Central, DMR_S.Central, DMR_Southern, DMR_Western)
# Crime rate per 1,000 population
crime_rate = avg_annual_crimes / pop_2022 × 1000
Multi-county divisions and population split sources:跨县警区及人口拆分来源：
| Garda Division警区 | Counties覆盖郡 | Split method拆分方法 |
|---|---|---|
| Cavan/Monaghan | Cavan, Monaghan | Pop proportion人口比例 |
| Sligo/Leitrim | Sligo, Leitrim | Pop proportion人口比例 |
| Roscommon/Longford | Roscommon, Longford | Pop proportion人口比例 |
| Laois/Offaly | Laois, Offaly | Pop proportion人口比例 |
| Kilkenny/Carlow | Kilkenny, Carlow | Pop proportion人口比例 |
CSO NDQ07 reports completions for 140 Eircode Output Areas (identified by 3-character routing key prefix, e.g. "D01", "T12"). We map each to a county using the Eircode routing key county table, then sum.CSO NDQ07按140个Eircode输出区域（以3字符路由键前缀标识，如"D01"、"T12"）报告竣工量。我们使用Eircode路由键郡对照表将每个区域映射到郡，然后求和。
D* → Dublin C* → Cork G* → Galway L* → Limerick
W* → Waterford K* → Kildare E* → Clare T* → Tipperary
# (full 26-county mapping applied)
completions_county = Σ(completions_oa for all OA in county)
Annual figures are derived by summing quarterly completions. The 3-year average (2022–2024) is shown as the headline supply figure.年度数据通过汇总季度竣工量得出。2022–2024年3年均值作为供给主要指标展示。
The ECB changes its main refinancing rate at scheduled Governing Council meetings (typically 8 times per year). Between meetings, the rate is unchanged. We construct a monthly series by forward-filling each rate until the next change.欧央行在定期管理委员会会议（通常每年8次）上变更主要再融资利率。两次会议之间利率不变。我们通过向前填充每次利率直到下次变动来构建月度序列。
for each month in [2022-01 … 2026-01]:
rate[month] = most recent ECB rate change on or before month
The Irish mortgage rate (Central Bank B.3.1) is already a monthly average and used directly. The chart shows both series from January 2022 to January 2026 (49 data points).爱尔兰按揭利率（央行B.3.1）已是月度均值，直接使用。图表展示2022年1月至2026年1月（49个数据点）的两条序列。
Update Schedule 数据更新计划
This is a static site. Data is updated manually when source agencies publish new releases. Current data as of March 2026. 本站为静态网站，数据在来源机构发布新版本时手动更新。当前数据截至2026年3月。
| Dataset数据集 | Current Coverage当前覆盖 | Source Frequency来源更新频率 | Next Expected Update下次预计更新 |
|---|---|---|---|
| PPR | 2010 – Feb 2026 | Monthly | Q2 20262026年Q2 |
| BER | 2009 – Q4 2025 | Quarterly | Q2 20262026年Q2 |
| RTB/ESRI | 2007 – Q3 2025 | Quarterly | Q4 2025 release (due Apr 2026)Q4 2025数据（预计2026年4月） |
| Daft.ie | Q4 2024 – Q2 2025 | Quarterly | Q3 2025 report (due May 2026)Q3 2025报告（预计2026年5月） |
| CSO Census (Pop + Vacancy) | 2022 | 5-yearly | Census 20272027年人口普查 |
| CSO NDQ07 (Completions) | 2022 – 2024 | Quarterly | Q1 2025 data (due May 2026)2025年Q1数据（预计2026年5月） |
| CSO BHA14 (Planning) | 2022 – 2024 | Annual | 2025 annual (due mid-2026)2025年年度（预计2026年中） |
| CSO CJQ06 (Crime) | 2022 – 2024 | Annual | 2025 annual (due mid-2026)2025年年度（预计2026年中） |
| IDA FDI | 2025 Annual Report | Annual | IDA 2025 Annual Report (due early 2026)IDA 2025年报（预计2026年初） |
| BPFI Mortgage | 2010 – Q4 2025 | Quarterly | Q1 2026 (due Apr 2026)2026年Q1（预计2026年4月） |
| ECB + Central Bank Rates | Jan 2022 – Jan 2026 | Monthly | Ongoing — updated with each ECB meeting持续更新——每次欧央行会议后更新 |
Found an error? Have a question? 发现错误？有疑问？
Data errors, methodology questions, and suggestions are all welcome. We read every message. 数据错误、方法论问题和建议均欢迎反馈。我们阅读每一条留言。
imfan.yang@gmail.comBuilt in Dublin by Fan Yang — a data exercise, not a licensed financial advisory service. 由Fan Yang在都柏林建立——数据分析项目，非持牌金融咨询服务。