Executive Summary
Critical internal dissent from a self-declared cycle fan-boy. Mat Smith has built his investment career on the 18-year cycle since 2003, but the UK data since mid-2022 forces him to publicly question the theory. Shows 5% falls 2022–24 during what the model says should be the most explosive growth phase — larger than any fall 1954–1990. Historical review of UK data 1845–1990 finds almost no evidence of the cycle outside the 1990–2022 window.
Key Claims (CRITICAL for Contradictions Log)
- “Since the summer of 2022 we have instead seen falls in the 5% region. To put that in context, that is a bigger fall than seen at any time between 1954 and 1990. It is bigger fall than the mid-cycle recessions in 2000 and 2019… All at a time when the model would predict massive end-cycle gains in the 10–20% per year region. This is a huge problem for the model.” — confidence: high
- Schroders data 1845–1920: UK house price to income fell 0.4%/yr while incomes rose 1.11%/yr — no 18-year cycle visible across this 75-year stretch.
- Nationwide data 1954–1990: not a single quarter posted a fall in UK nominal property values. 36 years of continuous growth = model requires 2 crashes, neither occurred.
- Inflation-adjusted analysis 1968–present: cycle “starts to behave” only from ~1990. Peaks/troughs: 1973/1975, 1989/1991, 2003/2009, 2023–2025? (unconfirmed).
- Land values: recent development-land price falls in London and wider UK since Q4 2021. Per Harrison, “movement in land prices precedes any large movement in house prices” → suggests cycle end already reached.
Notable Quotes
“For the 36 year period from 1954 to 1990 the UK saw only growth in prices, even through the turbulent 1970s. Not one quarter posted a fall in property values. So 36 years and just going up to the right without a mid-cycle pause or crash.”
“With our recent experience of the cost of living crisis, perhaps it is high inflation that is throwing the cycle off course?”
“It could mean we have had the growth part of the cycle already and we are in the recession phase. Or we have moved into the recovery phase. Or this is just a mid-cycle dip. Or not.”
Critical Takeaway
The 18-year cycle is reliably predictive only in the 1990–2022 UK window. Pre-1990 UK data does not show the pattern. Mat Smith models out possibilities:
- COVID disruption threw cycle off course
- Cycle actually ended early (2023–24 = recession not mid-cycle)
- Nominal vs real mattered — using real inflation-adjusted prices makes the pattern reappear 1968–present
Concepts Referenced
- 18.6-Year Real Estate Cycle — severely tested
- Fred Harrison — author being critiqued
- Land-Credit Feedback Loop — land price moves first (per Harrison)
- Contradictions Log feed
Cross-References
- Direct contradiction to foldvary-riding-upswing-2015 2026 prediction
- Supports Foldvary’s own caveat in georgist-journal-18yr-2012: “just a statistical average”
- Cross-link: harrison-moneyweek-2026-interview-2022 (2026 peak forecast under challenge)
[Source: raw/mat-smith-substack-2024.md, fetched 2026-05-03]