Costing the earth

What on earth is going on with the land values on housebuilders’ balance sheets? Over the past couple of months there seems to be more ins and outs than an East End hokey cokey. The impression one gets is that firms themselves cannot make up their minds on the values to be applied, so what hope is there for investors (or competitors) to make rational judgments? Let’s try and make sense of it all.

First, why do housebuilders make land writedowns? Well, theoretically the value of land is correlated to the price of the units built on it. If the average plot cost (APC) is equivalent to say 25% of the average selling price (ASP) then there is a multiplier effect on residual land value from changes in ASPs. In simple terms, the land multiplier is about three or four times the percentage rate of change in ASP. If selling prices rise then so does the value of land, but this is not recorded on the balance sheet. Rises in land values are reflected in the income statement through increased profit, and affect the balance sheet only by way of retained profits.

Of course ASPs have fallen in the past 18 months, so the value of plots in a landbank is worth less. Given that housebuilders all suggest a peak-to-trough decline in prices of more than 20%, land value depreciation may be about 60-80%.

Every major housebuilder except Berkeley has written-down its landbank over the past 12 months. (The rationale behind Berkeley’s decision is that its high density development means that APC forms a low percentage of ASP.) The top six quoted housebuilders have written-off £3.2bn so far, with Taylor Woodrow accounting for £1.35bn of that.

Net realisable value (NRV) is selling price minus selling cost. Its ratio compared with original land value varies between 14% at Bovis to 47.5% at Redrow. More interestingly, after NRV provision, average plot costs are between £28,000 at Redrow and £44,000 at Barratt, and yet the ASP of the two is only £26,400 apart and closing fast in Redrow’s favour. No prizes for guessing which you’d expect to make more profit per unit – but does it really mean Redrow is a better housebuilder than Barratt? Or simply that Barratt needs to increase its NRV provisions?

The impression one gets is that firms themselves cannot make up their minds on the value of land, so what hope is there for investors?

In fact, it’s even more complicated than that. Base land values aside, there is no uniformity on what NRV provision is required. The standard assumption of provisioning “to a zero margin” is interpreted as zero margin before or after (a) interest costs, (b) central overheads, (c) site overheads and (d) selling and marketing expenses. Thus company A might chose to establish a zero margin before interest costs and overheads whereas company B might adopt the principle of zero margin after central and site overheads and sales and marketing, which could create an earnings before income and tax (EBIT) difference of 10%. Once again selective analysis can be dangerous.

To bring this whole debate to the present, and further confuse matters, let’s consider the case of Persimmon. In its results for the six months to June it actually released £28m of the £624m NRV provision it made last year! Subsequently, both Taylor Wimpey and Redrow significantly increased their land provisions over the same period, a period when selling prices stabilised and may have improved. Moreover, Persimmon’s £28m net release masks much larger gross movements; it actually released £149m or 23% of its original writedown, but at the same time made an additional £121m in provisions, equivalent to 18.5% of land value. Persimmon cited price stability versus its NRV assumption (that is: 27% off selling prices), lower build costs, mix changes and site replanning as reasons for the changes. We must presume the first two factors to be positive and the final point to be a major negative. Persuasive arguments just muddy the waters even more and scream out for some sort of industry conformity on the issues of land evaluation and provisioning.

So what can we conclude from all this? I think the following:

• Persimmon’s release makes a mockery of rivals’ aggressive revisiting of provisioning in June. It may also make it harder for, say, Barratt to make further writedowns to equally conservative levels
• Provision releases will make genuine profits even more opaque and future results more volatile, so firms will be harder to compare
• Given the current industry predilection for replanning, should we worry about hidden devaluation. Under PPG3, replanning was formerly presumed to “add value” to the landbank. Not now, it appears
• It raises serious questions about the industry’s cyclical profits recovery, in my view. If companies are going to writeback NRV provisions as prices stabilise rather than crystallise those “gains” through completions in the profit and loss accounts, it will only serve to constrain margins and earning.

However, maybe it’s wrong to be concerned about land value provisioning, since the one thing all housebuilders seem to be agreed on is that it’s right to be back in the land market. That probably answers the crucial question about where land values are headed … but it also poses lots of new ones.