According to Savills, 2016 and the World Bank, the global real estate market is worth $217tr, 75% of which is residential property. Annual real estate trading has averaged $683bn annually since 2007 and reached $900bn in 2015. This represents turnover of around 0.3-4% of the capital stock. Inventory turnover in the S&P 500 company averages around 15 times, a multiple compared to global real estate of approaching 5,000. The global real estate market is huge – making up more than half the value of all mainstream assets in the world. But it is horribly illiquid.
Ignoring taxes on property transactions, round trip costs (legal, valuation and structural due diligence, contracts, investment advisory fees and brokerage fees) are likely to produce an average round trip trading cost of around 3-6% of the price of the real estate; and this estimate excludes abortive costs on transactions that do not complete. Annual fees of 5% on $1tr of property value would provide income for built environment consultants, real estate professionals, lawyers and accountants of around $50bn a year. Shaving 10% off that through greater efficiencies would release potential revenues of $5bn.
Clearly, the focus of endeavour will be on improving the efficiency of the transaction process, and the relative scale will encourage serious investment primarily in the residential market; commercial market applications are likely to follow behind.
In the UK, estate agents have traditionally operated on a ‘sole agency’ basis, and are reliant on instructions to sell properties on behalf of the owner for a success-only fee of around 1-3% of the purchase price. In the US, sole agency is less common than the pooling of instructions across groups of brokers through what is known as a multiple listing service. The listing data stored in a multiple listing service’s database is the proprietary information of the broker who has obtained a listing agreement with a property’s seller. Broker fees can as a result of this sharing model be as much as 6%, although the broker also facilitates exchange of contracts and legal professionals are less involved, saving fees elsewhere.
During an interview in his office in San Francisco, Jeremy Sicklick of HouseCanary told me that it takes on average over 100 days to sell or buy a house in the US, and a combined total of 11% of the price of the house is spent by both sides in fees and taxes. The HouseCanary mission and opportunity is to take a slice of that 11% while reducing the total burden of cost.
In most developed markets, where debt is used in the majority of house purchases, the bank or lending party commissions a valuation by a qualified professional. This inevitably takes time – form filling by the buyer, processing of the application by the lender, commissioning of the valuation, setting up the inspection, preparing, writing and returning the valuation and processing the information received – which can eat into a large proportion of the 100 days.
Uncertainty over the value of the property can also delay the initial sale process, risking gazumping and a long drawn out negotiation. The HouseCanary proprietors believe that they can develop intelligent AI algorithms which can be accurate for the vast majority of US homes to within a 2% error range. If this thesis were to be accepted by market participants and lenders, perhaps half of the 100-day lag can be taken out of the process.