Over the past few months, I have developed a particular distaste for the analyst community over their clustering of opinions around the critical issue of measurement of Dubai property prices. This is an issue that deserves critical oversight, and it is imperative that there be a proper framing of the root causes at play here.
House price indices serve as the most important market monitoring tool for assessing the health and efficiency of real estate markets. These indices are used for the valuation of portfolios and loans, as well as assessing the affordability and investment viability of markets. This way, they serve as a critical variable that is widely reported.
In Dubai, as well as in most markets, the most popular method for index construction is the repeat sales technique. The basic notion is to calculate the average capital appreciation by selecting individual apartments and/or houses that sell more than once and computing the price change between transactions.
However, there are two critical problems with this approach: 1) it ignores the effect of any renovation that might have taken place at the properties and 2) it ignores the payment terms that different homes have in place from different developers. As private sector developers, and even individual sellers, resort to offering multi-year payment plans, the time series data of the same houses are starting to show increasingly erratic movements, as the cost of such payment plans are embedded in the price.
Given the increasing usage of such payment plan options, index providers who hitherto ignored such offerings in their calculations, will be unable to do so going forward, implying that there will be a jump in prices recorded. However, in actuality, the real issue at hand has been the sample size of the data, the methodology that has been put into place, and the systemic bias that has resulted due to this asymmetry of information flows.
A US Federal Reserve study conducted in March of 2017 found that the effect of the above two variables was greater than 15 per cent in primary housing markets such as San Francisco, Los Angeles, Boston, Chicago and New York. Similar literature has been generated for other housing markets such as Frankfurt, Munich, Stuttgart, Paris, Milan and Sicily.
Despite the abundance of such studies that essentially replicate these findings, it remains a source of astonishment that index providers have not adjusted their methodology to incorporate for greater statistical rigour. The error is often compounded in time series data when gaps between transactions are filled in through listings, rather than primary market offerings.
And where the listings are included, there is no factoring in for the payment plans. None of the sample sizes included units that were offered via payment plans, despite evidence that nearly 50 per cent of homes sold in 2018 had such offerings. The end result is a considerable distortion of index performance, and it is this, more than any other measure, that has been at the heart of the fundamental misunderstanding of what is transpiring on the ground within Dubai realty.
The more data intensive method is through hedonic regression analysis, where both differential quality composition in heterogenous housing units (allowing for home renovation) as well as differential payment plans are incorporated ... and are explicitly controlled for. This is the case for the more analytically rigorous Case Shiller index in the US.
Using this methodology for the previous five years, we find a variance in price performance indices to be approximately 15 per cent. This is not only statistically significant, it is also relatively free of systemic bias. This has a considerable impact on mark-to-market loan to value ratios, decision making at the granular market level, where the bulk of the purchases is being conducted at the primary, as opposed to the secondary market level.
And it has the perverse impact of understating the house price appreciation in non-trivial ways. These statistical methodologies have been studied at length and have been adopted in developed markets.
Statistical rigour involves having a soupcon of the nuance, in order to glimpse at the underlying granular nature of truth. It is about time that the same criteria be adopted in Dubai realty as well.
Sameer Lakhani is Managing Director at Global Capital Partners.