The Competition and Markets Authority have just released the responses to their second round of consultation on the operation of the housebuilding market. The CMA had identified two issues which they considered might merit a market investigation reference – landbanking and problems with the adoption of public roads and other amenities by local authorities. There were 20 responses including most of the listed housebuilding firms (Barratt, Bellway, Crest Nicholson, Persimmon, Redrow, Taylor Wimpey and Vistry) as well as the Home Builders Federation.
Sticking to landbanking, the response of the housebuilders was fairly predictable. It’s pretty well summed up in Barratt’s response.
The land banks of large housebuilders are not acquired or used for anticompetitive purposes. They are a necessary part of a large housebuilder’s business model and, as the CMA recognises in its Update Report, they are driven in part by the nature of the planning regime.
As ever, in most of the responses of the housebuilders, the planning system gets it in the neck, the knee, the eye, the back etc.
I had rather thought that the planning system was being used as a scapegoat by the housebuilders but then I remembered the reasonable housebuilders that I’ve met who can get genuinely emotional (mainly anger and frustration) about the uncertainties, vagaries, costs and delays that characterise the planning system. I don’t think that it is performative.
A planning consultant that I know well constantly reinforces these exasperations. I have to remind myself that we didn’t design the planning system to be this way deliberately. Rather that it is more like a bureaucratic version of a Heath Robinson contraption formed from lots of legacy policies with botched quick fixes attempting to reconcile lots of conflicting social and political forces.
It’s important not to be too parochial. These conflicting social and political forces and frustrating land use regulatory systems (perhaps simplistically but increasingly NIMBY v YIMBY/YIYBY) are happening in plenty of other cities globally.
The frustrating thing about the housebuilder submissions is that they are largely evidence free. Whilst they repeat the mantra that there are sound operational reasons for housebuilders to secure access to land for future development (Taylor Wimpey), they don’t actually say how much land they need to hold for operational reasons.
Swimming against the tide a bit compared to other academic commenters, I’m almost embarrassed to write that the more I’ve looked at the topic, the more sympathy I have for the large listed housebuilders as far as landbanking accusations are concerned.
There’s a number of reasons for this that I’ll try to set out briefly.
Firstly, land-banking ties up a lot of capital and creates risks to the security of this capital for housebuilders.
Secondly, there is a risk of confusing effect and cause here. If housebuilders buy land, it becomes part of their inventory and it could perform well as an asset and it could prevent competition from other developers in the local market. It could do all three. Is it fair to presume that they’re mainly hoarding?
Thirdly, a proportion of their short-term land bank plots with planning consent can’t be built out because they are on sites that need detailed planning consent and/or to satisfy other planning requirements and/or require on and offsite site works before works can commence. The housebuilders could have said what the proportion was in their responses.
Fourthly, a proportion of landbanked plots will be on sites that are being built out. In simple terms, if you’ve just completed 200 units on a site in a single year that has permission for 1,000 units and expect to keep building at this rate, then you have a land bank of four years on that site. As the average site becomes larger, it’s likely that this will be more of an issue. Again, the housebuilders could say how much of their landbank consists of this category of plots in their consultation responses.
Fifthly, in the US, the large housebuilders have shifted to a ‘land-light’ strategy and have been able to control rather than have to buy plots with a significant lag before construction. The interesting thing has been that, when trying to minimise their exposure to investment in land, the US housebuilders still control around five plots for every one they complete – and still tend to own one or two as well. I should find out why this is. However, the key point is that, even when US residential developers are able to structure their land acquisition to enable them to buy plots a short time before construction, they still seem to need to control quite a lot of plots presumably for operational reasons.
Sixthly, it’s not just the US where the housebuilders have significant landbanks. The listed housebuilders in Australia and Ireland have larger landbanks in terms of years of inventory.
Finally, most of the volume housebuilders strategic land is controlled through options and these sites haven’t got planning consent. Since the housebuilders typically haven’t paid for it yet, they won’t want the land value to go up (although they’ll probably be happy because the causes of growth in land values will usually be good for their business). They typically want to get a consent on optioned sites. Otherwise they’ll have wasted a lot of time and money. The sooner that they can get consent, buy the land and build out the scheme, the sooner they can recover their investment in land promotion and acquisition.
One more…there must be some operational issues as well. Getting the right site in the right place at the right time is hard. In terms of providing continuity of employment and workflow to contractors, there are likely to be challenges in ensuring a suitable site is ready for the workforce to move onto once a scheme is built out. It can take a long time and there’s lots of uncertainties in finding suitable sites, winning the bidding process and executing the transaction. Alternatively, if prices and demand are falling, then they’re stuck with a site that may no longer be viable.
And another one…with the exception of Berkeley Homes, the housebuilders’ short-term land banks are all fairly similar in terms of years’ supply. They all tend to hover between four and six years’ worth of plots. When the completions go down, their landbanks go down.
Too much sympathy and understanding? I could go on about the fact that land values haven’t grown very much in the last decade, that the housebuilders don’t promote most of the land through the planning system, that SME housebuilders don’t really compete with the large listed firms…
I was surprised that they didn’t make a stronger, more evidence-based case in their submissions to the CMA. It’s not just, mainly or even the planners’ fault that they need five or six plots in their landbanks for every one they complete? But that’s the main message from their submissions that seem to have been co-ordinated by Slaughter and May incidentally.