Following listener pushback after Episode 22's debunking of the "18-year property cycle" theory, Jeremy revisits the topic in more depth, this time testing the more specific "18.6-year" version of the claim against city-level Australian data, as well as the original US and UK historical data the theory is meant to be based on.
Recapping the Theory
Jeremy briefly recaps the claim: that property markets follow a consistent, repeating 18.6-year cycle between major booms. Looking again at Australia's national growth history from 1981 to 2024, he shows that however the "boom" threshold is defined, the resulting gaps between growth peaks are inconsistent and nowhere near 18.6 years, whether using a 25% or 10% per annum growth threshold.
Addressing the Feedback
Jeremy notes that after publishing Episode 22, several YouTube commenters pushed back, asserting the cycle was real without providing supporting data, though one suggested testing Sydney and Melbourne specifically rather than Australia as a whole, which Jeremy takes up directly in this episode. He also raises a further question: where does the specific figure of 18.6 years actually originate? He understands the underlying claim traces back to historical US real estate data spanning over a century, though he hadn't sourced the original reference material directly.
Testing Sydney
Looking at Sydney's growth history over the past 44 years, Jeremy identifies several growth peaks, but notes the gaps between them are inconsistent and far from 18.6 years regardless of threshold used. Using a 20% per annum boom threshold, four peaks emerge with an average gap under 10 years; raising the threshold closer to 30% leaves three peaks with an average gap of 16 years, still off the mark, and with individual gaps ranging as widely as 3 to 8 years apart.
Testing Melbourne
Using a 25% per annum threshold, only two Melbourne booms qualify, both more than 30 years ago, meaning no further boom has occurred under that definition in three decades. Lowering the threshold to 20% produces four peaks spanning 32 years, an average gap of 8 years, again with no real consistency between individual gaps.
Testing Brisbane
Brisbane shows four clear peaks over 44 years, averaging a 13-year gap between them, again nowhere near 18.6 years.
Testing the Original US Data
Seeking to test the claim at its original source, Jeremy sourced historical US property price data from the Federal Reserve Bank of St. Louis, covering just over 60 years (1963 to July 2024). He notes that if an 18.6-year cycle held true, this period should show approximately three to four clean cycles, not more. Depending on how peaks are defined, the data shows around five to six peaks, with an average gap of roughly 11 years, again inconsistent between individual gaps, meaning even the US data the theory is supposedly built on doesn't actually support an 18.6-year pattern over this more recent 60-year window.
Testing UK Data
Turning to UK Land Registry data spanning roughly 50 years (from the late 1960s to around 2019), Jeremy identifies four clear peaks across three gaps spanning about 30 years, averaging just 10 years between peaks, again far from 18.6, and again inconsistent.
Jeremy's Conclusion
Having now tested Australia nationally, Sydney, Melbourne, and Brisbane individually, and the original US and UK datasets the theory is meant to be grounded in, Jeremy finds no support for an 18.6-year cycle in any of them. He notes that even if a different, more accurate average cycle length could be derived from this data (for example, closer to 10 years), the inconsistency between individual gaps would still make it useless as a forecasting tool, since there's no reliable way to know when the "next" cycle will actually begin. His closing message to commenters: if you believe the theory holds, submit data and analysis directly (via email), rather than opinion alone, and he'd be happy to reconsider based on genuine evidence.
An Unexpected Final Twist
To close the episode on a lighter note, Jeremy shares that he asked ChatGPT whether the "18.6-year cycle" is real. The response confidently confirmed the existence of an 18.6-year cycle, but referring to the lunar nodal cycle, a genuine astronomical phenomenon describing a gradual shift in the orientation of the moon's orbit, entirely unrelated to property markets or economics. Jeremy jokes that this may plausibly be where the original property cycle theory's specific figure was borrowed from, drawing a wry parallel between a real astronomical cycle and an economic theory that doesn't hold up under scrutiny.
Closing Thoughts
Jeremy and Damien close by reiterating that they welcome genuine data-backed challenges to their analysis, and would be glad to feature a listener's research on the podcast directly if submitted, but remain unconvinced by opinion alone. They encourage listeners to like, comment, subscribe, and share the episode with anyone who still believes in the 18.6-year cycle.