Single-family rentals (SFRs) are undoubtedly the new darling of residential investment. While build-to-rent strategies are taking off in both the UK and the US, soaring construction costs are driving investor interest towards existing stock.
Institutional investors such as pension funds and insurers ultimately want scale. This is hard to achieve in SFR using traditional property industry methods for acquisitions, due to the granular and disparate nature of the assets.
In the States, where SFR first emerged as an asset class, investors were able to quickly achieve scale by snapping up thousands of foreclosed homes in the wake of the subprime mortgage crisis.
In Europe, where SFR is only beginning to take off, the starting landscape is very different. Far from causing a housing market crash, Covid-19 saw house prices boom globally, with homeowners or prospective buyers benefiting from policies such as mortgage holidays and tax breaks. While there are signs of housing markets cooling and warnings of a potential recession, few are predicting a 2008-style collapse. In fact, it is office values that many investors are worrying about.
Overcoming the obstacles that come with acquiring granular, hard-to-access assets is why investment managers wanting to grow their exposure to Europe’s SFR sector – worth some €50trn – are turning to proptech firms like ourselves or Bricklane, which recently formed a joint venture with Moorfield.
IMMO’s platform rests on three pillars: workflow automation, data gathering and data analysis.
Data informs our approach at each stage of the investment process, from capital allocation through to underwriting, leasing and management.
Yet for all the excited talk of the power of ‘big data’, what matters most to us is the quality – not the quantity – of the data we gather.
Our AI-powered Automated Valuation Model, which sits at the heart of our platform, draws primarily from three data sources.
First is, as you would expect, real estate datasets looking at historic rent levels and house prices.
Second is macroeconomic indicators at a district level (Eurostat NUTS-3) like GDP per capita, demographics and transport connectivity. While these are also backward looking, they help us determine what areas to invest in, and predict where will best hold value in a recession.
Third is live listings for properties available on the market, as there is typically a one- to two-month lag between current prices and what the agents or public bodies such as the UK Land Registry and HMRC have recorded.
Ensuring all this data is clean – meaning there are no mistakes or duplications – is crucial, especially if you are focused on mis-priced or value-add opportunities like us. You can have all the data points in the world, but if they are inaccurate – even by just a fraction – you can wreck your whole model. Many in real estate are already suspicious of machine learning systems like ours, which are seen as something of a ‘black box’ of information, and prefer regression analysis instead.
In Europe, the primary challenge is a lack of standardisation in how data is collected and reported, with the quality and availability of data varying from geography to geography. There will always be human error, but bad practices and overly strict data privacy laws in some countries inhibit market transparency.
Data cleanliness is typically less of an issue for macroeconomic indicators, which are usually collected by government statisticians. You also have a pan-European source in the form of Eurostat. However, for property market data, what Europe really needs is its own CoStar-equivalent, like the US has.
These barriers have not prevented us from amassing an initial £66m portfolio in the UK and Germany. Yet if the industry is serious about unlocking the full potential of Europe’s SFR sector, then it needs to act to ensure we have access to not only more data, but cleaner data too. As they say, bigger is not always better.
Souad Chatty is head of market intelligence platform at IMMO