Big data could lead to automation in real estate investment

Speakers at PERE’s annual Asia Summit in Hong Kong still believe gut instinct cannot generally be automated when it comes to underwriting deals. But the market’s biggest risk-return strategy may be the exception to the rule.

Once again, the discussion around the opportunities and challenges facing the private real estate sector is centering on changing occupier habits in the face of technology-led advances.

A US-China trade war was sporadically identified as a real threat to business at the PERE Asia Summit, but it didn’t displace technology as the primary disruptor for the 500-plus delegates at Hong Kong’s Grand Hyatt hotel this week.

Key to the debate was the role ‘big data’ has to play for investment managers as they underwrite their outlays. And the overriding consensus was that as long as humans require real estate, humans will be the ultimate judges when it comes to working out what to pay for it. “I just can’t see robots making investments,” said keynote speaker Goodwin Gaw.

While the expanding reams of data at their fingertips were playing an increasingly important role in optimizing how they buy real estate, panellists refused to accept investment processes could become automated.

Other reasons for that were offered too. Sung Lee, an executive vice-president at Altus, a software solutions provider, pointed out that while data collection among investment managers is improving, data are still being collected from disparate sources, which makes aggregation and standardization a challenge, and in turn makes it a less effective underwriting tool. Capitaland’s Ervin Lim, meanwhile, argued that data still require human interpretation in order to make decisions. His firm is one of the more bullish prop-tech investors, running a dedicated venture capital platform.

But not everyone was so certain that data would not lead to automated real estate investment. While Starwood Capital Asia’s managing director Kevin Colket labelled it “past-looking”, “unable to predict black swans” and “a tool that will never replace people’s judgment and desire to get together”, he also conceded that, in the future, core real estate might be one area where machine-learned pricing could happen and transactions lose the human touch.

Intentionally or not, then, he effectively pointed to the lion’s share of real estate investment as potentially becoming subject to technological takeover. Indeed, the $92 billion of value-add and opportunity fundraising recorded last year by PERE research is miniscule compared with the almost uncountable equity amassed for property’s lowest risk-return strategy. Of the approximately $650 billion of investment in 2018 predicted by broker JLL, it is a safe bet to say the vast majority will go to low-risk, low-return strategies. Core real estate provides a suitable fixed-income alternative able to meet most institutional liabilities, and therefore all real estate strategies are oriented in either its creation or retention.

We’re talking about a hell of lot of bricks and mortar, therefore, with the potential to change hands without much human involvement. Of course, that still leaves the likes of our speakers from Gaw and Starwood Capital – the runners of higher risk and return money – with jobs to do as they continue to rely on their intestinal instincts to make property core. “You won’t find tech replacing a gut thesis,” Gaw insisted.

Still, what was heard on PERE’s biggest stage this week has its editorial team thinking of a future where big data has core real estate selling like exchange traded funds – containers for much of the equities markets’ capital – even if property’s hedge fund equivalents continue on in their quest to find the asset class’s alpha.