For years now, the theory has been mooted that more energy efficient and better insulated buildings should attract a corresponding premium when the time came to take it to market.
After all, with low costs of capital (interest rates) low returns on investment, government subsidies and the rising cost of energy itself, if you can produce your own, or drastically reduce the amount you use, then over a long enough time frame you should be ‘in the money’ right? But does that theory reflect the reality of the market? More importantly, right now, are the latest improvements in energy efficiency being priced into residential properties?
I’ll admit here at the start that I have a bias in this argument, because I want to find this evidence, and encourage people to invest more into renewable energy and efficiency. Where to begin? I chose the past 12 months of transactions in South West London, an area that covers everything from the mews of Knightsbridge to the terraced houses of Merton Park.
However, the first chart was a real kick in the teeth. The data below shows the average £/sqft value of properties based on their energy rating according to Energy Performance Certificate.
Surprisingly, ‘A’ Rated properties were less than half as valuable as those that were less efficient. They comprised only a small section of the sample properties (5 ‘A’ rated properties out of 8.5k total) but there was no clear relationship between energy efficiency and pricing for any of the other energy ratings, with all the others hovering at around £900/sqft.
I then considered the different property types in the sample, the variance between big old houses, the small modern flats, perhaps size was affecting our results? Sure enough, more inefficient properties also tended to be larger – as we can see below:
So bigger houses are less efficient, and in aggregate there is no preference for energy efficiency, so is there a preference for larger properties? The general assumption is that smaller properties can be sold for higher values, as customers are effectively paying for the number of doors and are less conscious of the actual space itself. The blue line chart below shows the average £/sqft value for each size of properties in South West London.
Interestingly, it shows that whilst there is a reduction in £/sqft values between 500-700sqft – which aligns with our theory from the development world – properties larger than 1,000sqft actually showed increasing £/sqft values. This demonstrates that given the choice, those that can afford to do so are willing to pay a higher premium to get that all important extra space.
For example, a 700sqft property might typically be worth £850/sqft whereas a property twice its size, could be worth 20% more at £1,030/sqft. For a detailed look at the exact sample size, it is shown in orange for each property size with the count on the right hand axis.
So, we have seen that more energy efficient properties are smaller, and people actually pay less for small to mid-sized properties on average and pay more for larger ones, but in aggregate there is no difference between a B and G rated property because of size difference. Which must mean that similarly sized properties, with different energy ratings should have different £/sqft values.
The problem though, is then finding them.
The heatmap below shows the frequency of properties with different energy efficiencies in differ-ent postcode districts of South West London. I excluded A and G rated properties as the sample size was too small, and selected SW11, SW6, SW18, SW15, SW16, SW19 and SW17 as they had the largest and most consistent sample sizes available.
The heatmaps below show the £/sqft values for properties based on their size and energy rating. Unfortunately this didn’t show what I was looking for either, and continued to suggest that smaller and less efficient properties attracted £sqft premiums. The irony being that according to the data, a small inefficient property would sell at a higher £sqft value than anything else!
One other version of this dataset was attempted, using only data from newbuild properties. Surely a customer, if given the choice between a new property with an energy rating of B or F, they would be willing to pay more if it had a higher rating and therefore their bills were lower?
Unfortunately not it seems. No clear relationship could be shown, and whilst some areas like SW11 and SW16 showed evidence of the expected buying behaviour, with lower £sqft values for lower energy efficiency, other areas such as SW18 and SW6 showed the exact opposite, with higher values for less efficient properties.
This makes it highly likely that some currently non-numeric element is at play in the data. Perhaps the less energy efficient properties have higher ceilings, larger windows, a more desirable design or are in more specifically enviable locations.
These items can’t be picked up by simple analysis as we’ve seen, so if want to save the planet one house at a time, we will have to revisit this later on once our data science team have time to examine it more closely!
© Treex 2020