Gera Local Price Momentum (GLPM) — Methodology

Data source

Step-by-step formula

1. 1

   Download the Price Paid Data CSV files

   Fetch pp-2026.csv and pp-2025.csv from https://www.gov.uk/government/statistical-data-sets/price-paid-data-downloads. No API key required. Files are plain CSV with no header row. Column 4 (index 3) is the full postcode (e.g. "SW3 4JU"); column 2 (index 1) is the price in GBP; column 16 (index 15) is the record_status ("A" = added, "C" = changed, "D" = deleted).

2. 2

   Filter valid records

   Exclude rows where record_status = "D" (deleted). Parse the postcode field to extract the outcode (first group before the space). Discard rows with an invalid postcode or price ≤ 0.

3. 3

   Group by outcode × period

   Create two maps — current (pp-2026.csv) and prior (pp-2025.csv) — keyed by outcode, values = array of sale prices.

4. 4

   Apply N≥10 threshold

   For any outcode with fewer than 10 transactions in the current map, skip it entirely — the published dataset omits "insufficient data" districts.

5. 5

   Compute medianPrice

   Sort the current-period prices array ascending. median = prices[⌊n/2⌋] if n is odd; (prices[n/2−1] + prices[n/2]) / 2 if n is even. Round to the nearest integer GBP.

6. 6

   Compute pctChange12m

   If the prior-period array for the same outcode has ≥10 entries, compute its median identically. pctChange12m = (currentMedian − priorMedian) / priorMedian × 100, rounded to 1 d.p. If prior period has <10 entries, pctChange12m = null.

7. 7

   Record txCount

   txCount = length of the current-period prices array (before any deduplication — one transaction = one row in the PPD CSV).

8. 8

   Sort and publish top 600

   Sort all qualifying outcodes by txCount descending. Publish the top 600 as the GLPM dataset. This ensures coverage prioritises the most liquid markets.

Pseudocode

# Input: pp-2026.csv (current), pp-2025.csv (prior)
# Output: GLPM_DISTRICTS array

for each csv_file in [pp-2026.csv, pp-2025.csv]:
  for each row in csv_file:
    if record_status == "D": skip
    outcode = postcode.split(" ")[0].toUpperCase()
    price = parseInt(fields[1])
    if price <= 0 or outcode is empty: skip
    period_map[period][outcode].append(price)

for each outcode in current_map:
  if len(current_map[outcode]) < 10: continue  # insufficient data

  medianPrice = median(current_map[outcode])
  txCount     = len(current_map[outcode])

  if outcode in prior_map and len(prior_map[outcode]) >= 10:
    priorMedian  = median(prior_map[outcode])
    pctChange12m = round((medianPrice - priorMedian) / priorMedian * 100, 1)
  else:
    pctChange12m = null

  results.append({ outcode, slug, medianPrice, pctChange12m, txCount })

results.sort_by(txCount, descending=True)
GLPM_DISTRICTS = results[:600]

Limitations & caveats

• Lag: Land Registry typically publishes transactions 2–3 months after completion. The January–April 2026 file may not yet include all April 2026 completions.
• Mix effect: A median over all property types (detached, semi, terraced, flat) is influenced by the mix of properties sold, not just price-level changes. A shift toward more expensive types can inflate the median even if like-for-like prices are flat.
• 12-month comparison window: GLPM compares 2026 data (Jan–Apr) vs full-year 2025 data. The windows are not identical in length — interpret pctChange12m as a directional indicator, not a like-for-like repeated-sale index.
• Not professional valuation advice: GLPM is a statistical summary of past transactions. It is not a property valuation, mortgage advice, or investment recommendation.