I hear that the RICS are going to produce a new version of their Valuation Information Paper on the valuation of development land. Like some other valuation methods in the UK, development appraisal techniques seems to have got stuck in some type of time warp. All development appraisal models simplify but some simplify a lot more than others. All development appraisal models are flawed but some are much more flawed than others.
The simple residual approach involves making up some stuff that doesn’t actually reflect what happens in development. Most development appraisal methods are based on the calculation of a residual land value estimated as the difference between development revenues and development costs including a return for the developer. As the name suggests, the simple residual approach makes some simplistic assumptions.
- It assumes that all development costs are paid in advance in each period.
- Development costs (excluding land) are spread equally over the development period.
- All revenues are received at the end of the development period.
- The costs and revenues don’t change over the development period.
- All development costs can be borrowed.
- Return is in the form of a cash lump sum and developers want the same cash sum whether it is in one or five years.
This can be made more realistic.
In the academic context, in the 1970s (I think that was Peter Byrne, Stuart Morley and Andrew Baum who started to apply DCF techniques to development appraisal), it was realised that the simple residual method was far too simplistic and did not reflect the reality of development. However, the advantage of the simple model was its tractability and the fact that it avoided lots of calculations. Crude simplifications and unrealistic assumptions may have had some justification in a world before computers or even calculators but not anymore. It’s hard to think of any reason for it to be used these days. Mea culpa, since we’re worried that it might still be used in practice, we still teach it to our students so that they won’t face the potential embarrassment of not knowing about it. Perhaps we should bite the bullet here.
I think that the problem with what Stuart and Andrew did is that they retained some of the original simplifying assumptions from the simple residual model when they didn’t really need to. Applying DCF was a major improvement – the timing of costs and revenues could be treated much more realistically and accurately. However, other simplifying assumptions persisted in textbooks and software etc and in practice. In DCF models, it is still generally assumed that all development costs can be borrowed. Often costs and revenue inflation is ignored (implicitly zero change is forecasted) and return is treated as a simple margin on revenue or costs. It’s notable that, when proper real estate economists stumble upon development appraisal, they tend to scratch their heads in puzzlement and wonder what’s going on. Their underlying question is “Why does real estate development appraisal ignore long established and basic project appraisal theory? What’s so special about a real estate development project?” If they asked me, I’d have no answer for them.
Mainstream project appraisal theory does not require any oversimplification or assumptions that are unreal (apart from payments being paid in neat sequences at the beginning/end of months/quarters/years). It tells us to estimate the expected net cash flow from a project and discount it at an appropriate target rate of return i.e. predict how much will be received from the project and when, predict how much will be paid out from the project and when and then discount the resultant net cash flow at a target rate of return. This requires some view on how costs and revenues will change. Most project appraisals have to deal with change in revenues and costs and it’s standard in the appraisal of the Investment Values of leased real estate assets .
Financing the project is a separate issue. It’s worth remembering many non-development projects need loans which are not included in the unlevered project appraisal and that real estate investment appraisals to calculate Investment Value don’t include finance costs – albeit gearing is very common. The unrealistic assumption that all costs are borrowed is not made.
Return is not factored in as a crude mark-up but rather as a risk-adjusted required rate of return (albeit this is a very difficult number to estimate rigorously).
Some developers actually use standard project appraisal methods as standard but many don’t and you certainly don’t see it in viability guidance for planning. Does these deviations from reality in development appraisal matter for their reliability? I suspect that the answer is “not much”. The uncertainty caused by a poor model is likely to be minor compared to the uncertainty caused by uncertain appraisal inputs (costs, revenues etc.). This probably explains the persistence of weak models in practice.
Will things change? I doubt it. Generations of professionals have been taught the simplified, unrealistic stuff. It’s embedded deeply in textbooks, development appraisal software and, er, guidance notes. As I said, I still teach it myself because students will be expected to know it by their employers. There’s not really a good reason to move towards a robust model since there doesn’t seem to be much clear-cut competitive advantage in using one. However, I don’t envy whoever is drafting the new VIP. I suspect that whoever it is may be forced to regurgitate stuff that they think is guff. Or worse, they may not realise that it’s guff.