Land is the fundamental ingredient in the construction of new homes. Many of the issues limiting the rate of new home building can be traced back to the pricing and availability of land for residential development.

The availability of development land is constrained by the planning system but also by other factors. Land is generally an appreciating asset and many landowners’ price expectations will be firmly set. Even if planning regulations were eased further, there are limited incentives for landowners to sell at a faster rate than they are currently as that might compromise the price they achieve. That is particularly the case where the land is already generating an income through other uses.

Development land is typically valued using a residual approach with reference to comparable transactions. A developer assesses what new build house price is achievable in that location with reference to prices and sales rates in the second hand market and on nearby comparable new build sites. At a very basic level (assuming no affordable housing, S106 or CIL), multiply that new build house price by the number of homes to be built on the land and you arrive at the gross development value (GDV) of the site. The underlying value of the land is then the GDV less the cost of development and less an appropriate allowance for profit as the formula opposite shows.

When in competition with other developers and assuming discipline on appropriate profit margins, the winning bidder will typically be the one that pushes for a combination of the highest new build price, the highest density (subject to planning) and the lowest build cost (unless one offsets the other). With land typically bought up front, this approach sets the target new build house price in stone. Developers will then only build and sell homes at the rate dictated by market demand for this target new build price. It is through this mechanism that the 10 to 1 ratio between overall housing market transactions and private housing starts is thought to arise.

We can demonstrate the link between house prices and land values using a simple model. It uses Nationwide new build house prices as a proxy for GDV and the only input. It is based on an old land buyer’s rule of thumb: land is 1/3rd of GDV. Therefore the modelled land value is 1/3rd of the house price. The volatility in land prices is accounted for by assuming that the remaining 2/3rds (effectively costs and profit) are sticky and do not fall. Therefore land values absorb the full impact of any falls in house price and the full benefit from any rises while house prices are below their previous peak. The model output (red line) does a reasonable job of tracing actual land values up until two years ago.

Other Factors

At first sight, the model’s output suggests that there is substantial room for uplift in land values. However, the model is oversimplified by intention and the reality is more complex and varies by local market. Calculating actual land values has to account for a mix of property use classes; a mix of tenures; site remediation; infrastructure and CIL; affordable housing contributions; and other costs. It will also include forecasts for rates of sale, house prices and construction costs. These will then be balanced against the developer’s required rates of return. This all adds a degree of uncertainty to the calculation.

Many developers and housebuilders will remain cautious despite the current strength of the new build sales market. Housebuilders’ profit margins may be returning to pre-recession levels but there are other potential economic, political and market risks that warrant some caution. Build cost inflation and labour availability have been top of the constraints list recently and explain some of the underperformance. The need to absorb CIL and the lower value of affordable housing will continue to contribute to the gap between the model and actual land values in some markets. With large scale Government support of the housebuilding sector through the likes of Help-to-Buy, there will be political uncertainty created by the election cycle and a lack of long-term cross-party strategic planning for the provision of new homes.

Our house view is more balanced and suggests that the full uplift in land values will only be realised in markets with high housing demand and constrained land availability. There are currently a relatively high number of consents coming through the planning system compared to new housing starts although they are unevenly distributed across the country. Where starts exceed planning consents, this could lead to increased demand for land and an associated rise in land values. Thanks to the mechanisms detailed in this note, higher land values could then limit sales rates in markets unable to absorb higher new build prices. This approach means that developers and housebuilders will continue to build homes at a rate dictated by how fast they can sell them based on the price they had to pay for the land. The need for more land to be released is apparent.

Planning issues will always be a valid go-to complaint for developers but recent evidence suggests that the availability (and cost) of materials and labour has been a bigger constraint over the last couple of years.

Survey responses from the HBF survey suggest these issues are declining in importance but remain high relative to recent years. The uncertainty around future build cost inflation will inevitably limit land prices to some degree.

Prior to 2001, indicators for build costs suggest that they generally matched the trend in house prices.

That relationship broke down between 2001 and 2007 when house price inflation substantially outpaced build costs.

The relationship has resumed post recession and the recent house price rises have been matched by increases in build costs.

The performance of local land values since the market peak in 2007 broadly reflects the underlying performance of the local housing market.

There are exceptions that highlight local approaches to development and land release. Cambridge has seen similar or higher house price growth than other markets yet land values are lower relative to peak. This reflects the local approach where more land has been made available, including some from the Greenbelt, and there is clarity on the level of affordable housing required.

Although not directly relevant to the above analysis, the chart opposite shows the trend in land values between 1892 and 1969.

The substantial premium for commercial land across the period and the divergence in value between residential and agricultural land from the 1930s are both interesting features and reflect the importance of location in land values.

Care should be taken in reading too much into any single year given the data collection periods.

Taken from
Housing Market Notes:

• Approaching peak loneliness?
• Size matters
• Housing delivery
• Pension reform and the housing market
• Housing market activity
• Rental Britains
• First time buyer affordability
• Local house price indices
• Measuring housing need
• Magnetic London