In addition to being a mainstay of conversation in rather dull dinner parties the topic of property valuation is central to many covered bond jurisdictions. As I have commented before there are at least eight ways to value a house in different covered bond laws in Europe but by far the most prevalent are market value and mortgage lending value.
Which should any jurisdiction use? It should be obvious but just in case: mortgage lending value is valid if home ownership levels are low, market value if they are high. If renting a property is the norm, is not correlated to private wealth and if each community and type of property has a clearly observable market value, then it does make sense to talk about a theoretical floor to property values. All of these apply in, for example Germany, few in, for example, the UK (as an aside, I have never met a British person who even knows the floor area of their own property, a fact that continues to mystify German friends).
On one level, it doesnt really matter. This is the value which when multiplied by 80% (or whatever LTV cap applies locally) gives the value which can be used for the purposes of calculating the statutory over-collateralisation. This in almost all cases is a useless number as the over-collateralisation demanded by the rating agencies and/or the regulator is invariably far higher.
Rating agency models adjust their required over-collateralisation to take into account the valuation method and usually know the actual gap between the two valuation methods, so the actual driver of collateral levels doesnt really care what the basis is.
What concerns me more is the regulator. There seems to be a growing trend towards the gratuitous use of mortgage lending value because it is a) lower, therefore more conservative. Regulators like conservative. b) less volatile, therefore less likely to generate sudden needs for more collateral in a stress scenario and c) the way it is done in Germany.
In some cases the regulator even uses the mortgage lending value for the sake of covered bond calculations and the market value for the sake of bank capital needs. At best this looks indecisive, at worst it is a significant extra cost for the banks who need to value each property twice under different assumptions.
How much do they differ? By more during a price expansion (because rent prices are very slow to react to market conditions), less during a downturn. Also, presumably be less in a market with an efficient rental market that regular reprices (a student or old people dominated community for example). Im not aware of it ever going negative but Ive heard 11% suggested as the long term average difference.
But the usefulness of mortgage lending value stands and falls on one thing; whether market values are mean reverting? Over time do they tend to move back toward the mortgage lending value? If not, theoretical property values seem to be difficult to defend.
Event, Articles and Videos that might interest you
No related events !