Property trends on the money for AFL success

The article describes how Knight Frank researcher Richard Jenkins applied commercial property forecasting models to AFL results and found statistical correlations. He analysed a decade of data and compared major market indicators (like vacancy rates and industrial production) with the performance of most of the league’s 18 teams. The article emphasises these are correlations from the model, not proven causes.

The model was created by Knight Frank researcher Richard Jenkins. He used the complex mathematical models commercial agencies use to forecast property trends, analysing about a decade of data with indicators such as net absorption, vacancy rates, stock levels, industrial production and unemployment (and consumer confidence in some cases).

According to the article, the model’s inaugural run last year correctly predicted that the Sydney Swans would win the premiership and it correctly picked the ranking of four other teams. Seven clubs were out of place on the ladder by just one spot, showing a surprisingly good fit for a novel application of property models.

The model’s 2013 ladder predictions, as listed in the article, are: 1 Sydney, 2 West Coast, 3 Hawthorn, 4 Richmond, 5 Geelong, 6 Collingwood, 7 Fremantle, 8 Essendon, 9 St Kilda, 10 Adelaide, 11 Carlton, 12 Western Bulldogs, 13 North Melbourne, 14 Brisbane, 15 Port Adelaide, 16 Gold Coast, 17 Melbourne, 18 GWS.

The article highlights a notable correlation between Richmond and consumer confidence — Richard Jenkins says consumer confidence is actually a good barometer of where Richmond will finish. It also mentions an unclear connection between St Kilda’s performance and the St Kilda Road office market, but says the reason for that link remains anyone’s guess.

The article suggests conditional links: if office tenants prove reluctant to take up space in Sydney’s CBD in 2013, the Sydney Swans would be well positioned to win again. For West Coast, the model says the club’s place in the grand final could be assured if strong demand continues to push down Perth’s office vacancy rate. In short, changing office vacancy dynamics in those cities were used as part of the model’s input and interpretation.

The article notes that no punters have asked to see the model’s algorithms, and it does not indicate that the algorithm or full methodology is publicly available. The piece presents the work as an interesting application of commercial property models rather than a published, open tool.

For everyday investors the key takeaway is that commercial property indicators (vacancy rates, net absorption, stock levels, industrial production, unemployment and consumer confidence) can show surprising correlations with broader social outcomes — in this case, AFL performance. However, the article implies caution: correlations don’t prove causation, some links are unexplained, and this is presented more as an intriguing experiment than investment advice.