Size has its limits

The current round of consolidation in housebuilding brings the prospect of huge companies operating on a scale never before seen in this country. For some time, it’s had me thinking about just how big a company can grow in a decentralised industry without encountering diseconomies of scale.

Currently we have two giants – Persimmon and Barratt – that have proven track records in managing growth, and which clearly have the capability of building more than 20,000 homes a year. A merged Wimpey and Taylor Woodrow should be able to do the same. In theory, of course, there’s no reason why a British housebuilder couldn’t produce 30,000 or more homes a year. While it would hold a dominant position in the industry, it would still represent less than 20% of the UK market, so there should be no competition issues. And it has been done elsewhere: in the US, there are three companies that have built more than 30,000 units a year. There, however, the market is much bigger, the land and planning difficulties smaller, and the building techniques different.

In Britain, these waters are as yet uncharted. No company here has yet built more than 16,500 homes a year. Currently, typical production per site in Britain is about 35 unit completions a year. So 30,000 a year would need 860 outlets across the country. Under the existing model, volume housebuilders deliver their product through decentralised local subsidiaries. Even if every subsidiary managed 600 units a year – possible but most unlikely, in my experience – this leviathan would still need a whopping 50 divisions run by 50 different teams – potentially, a structural and management nightmare.

For the future, the biggest constraint on production for the biggest companies is not going to be the supply of land or financial firepower or even planning, but the ability of management to handle the sheer scale of the business while keeping profitability and quality intact.

So where do we go? Do we change the model and increase each subsidiary’s optimum production target to, say, 1,000 units a year, thus trying to maintain national coverage with only 30 subsidiaries? It’s one route, although attempts to do this have not been successful so far. I suspect we might also see a real trend towards “horizontal expansion” – movement into related sectors requiring less decentralisation, such as commercial, leisure, and retail and property management. I can also see some forsaking the pursuit of growing volumes and profits to concentrate instead on cashflow, rather like Berkeley, where the concept seems to be paying off for both shareholders and management. What I don’t yet foresee is a 30,000-units-a-year British housebuilder.