The relativity challenges

For many years, relativity had been assessed with reference to scarce market evidence from unenfranchiseable lease sales and the graphs of relativity produced by various surveyors’ firms. However, the evidence behind these curves varies in quality and in many cases has been lost, preventing them being tested by parties to current disputes over the appropriate premium to be paid.

In June 2014, a challenge to this approach was heard in the Upper Tribunal, advocating the use of an hedonic regression analysis of transactions from 1987-1991, before the 1993 Act which extended enfranchisement rights to all houses and brought in the right for leaseholders of flats to extend their leases or buy the freehold of the building.

This challenge failed, largely because it was felt that analysis of an historic market was not relevant to assessing relativity in the current market. Relativity is effectively a function of discount rates. The lower the discount rate, the lower relativity becomes. It is clear that we should expect relativity to have changed over the last 20 years, with current relativity being substantially lower than it was in the 1990s.

A further challenge was mounted in January 2016, again promoting use of the hedonic regression analysis of pre-1993 Act transactions. This time the decision was clearer, comprehensively rejecting the analysis of pre-1993 Act transactions and finding it probable that “economic and market changes since 1987-1991 have indeed lowered relativities.”

The Tribunal considered detailed evidence on many of the graphs of relativity produced by surveyors’ firms, but gained little or no assistance from most in arriving at its decision. One exception was the Gerald Eve graph, which was found to have become “the industry standard”, but evidence was heard that suggested it “may indeed now overstate relativities for leases without rights under the 1993 Act” and it was regarded as secondary to market evidence.

Modernising relativity

Evidence was also heard regarding the new Savills 2015 enfranchiseable graph. Although it was found that this graph is “subject to some possible technical criticisms”, the Tribunal commended it as “a significant improvement on the Savills 2002 enfranchiseable graph, being based upon recent market transactions which have been objectively analysed”. The analysis presented here is unchanged from that provided in evidence to the Tribunal.

Using data from Lonres on flat sales in prime central London, we carried out an hedonic regression analysis of the current market using data from January 2010 to June 2015. Hedonic regression is an approach widely used in housing market analysis, including as the basis for the Nationwide and Halifax house price indices.

The approach relies on the assumption that the total value of a property is made up of the cumulative effect of its characteristics, for example number of bedrooms, size, parking, garden, location, etc., and allocates a proportion of value to each of these. For leasehold property, the length of the unexpired term is clearly a key characteristic in determining value. Our model produces a range of coefficients to describe the contribution to value made by each characteristic. These all make intuitive sense, building our confidence in the model.

Our most robust model analyses 5,904 sales of leasehold flats and provides the point estimates of relativity for each lease length shown in Graph 1. The size of the point indicates the number of data points at each lease length. We have then fitted a curve to the point estimates of relativity. The equation for this curve is shown on the graph.

The analysis provides a starting point for the assessment of relativity; it gives an estimate of relativity for a lease with Act rights in the current market. A deduction from this level would need to be made in a valuation of a lease on the hypothetical assumption that the lease being valued does not have the rights conferred by statute, i.e. to extend the lease or join in a collective purchase of the freehold.

^{Source: Savills Research using Lonres (data from January 2010 to June 2015)}

Discount

The effect of the statutory assumption is to reduce the rights a leaseholder possesses in holding the subject lease compared to leases in the market. The current market curve we have derived therefore marks the maximum relativity for leasehold reform valuations, with adjustment required to take account of the statutory assumption, reducing relativity below this market level.

The evidence available to assess the discount required from the real world curve shown in Graph 1 is scarce. We have compiled a range of reference points for the discount required from the current market relativity to calculate relativity under the statutory assumption (Graph 2).

^{Source: Savills Research}

As best we can, we have fitted a curve through these reference points from which the discount from enfranchiseable relativity can be estimated for every lease length. The resulting unenfranchiseable curve is shown alongside the enfranchiseable curve in Graph 3.

^{Source: Savills Research using Lonres data}

SAVILLS RELATIVITY GRAPHS

Click table to enlarge

^{Source: Savills Research using Lonres data}