Middlemen: Matchmakers or Sources of Strategic Information?

Probing that question using a database of for-sale-by-owner home listings

What does a middleman do in today’s economy? Traditionally seen bringing together buyers and sellers, pure go-betweens — think executive recruiters, brokers of everything from logistics services to food ingredients, and residential real estate agents — have in recent decades fashioned themselves as strategic advisors. They offer a menu of data and analysis on the markets in which they operate, hoping to elevate their roles beyond mere finders.

And yet their compensation is nearly always tied to the traditional middleman role — finding a company’s new CFO, booking truck or rail capacity for a shipper, brokering the delivery of fruits and nuts, and bringing a buyer to the seller of a home.

Brigham Young’s Darren Aiello and Taylor Nadauld and UCLA Anderson’s Mark Garmaise examine the competing views of the middleman, or intermediary, employing an unusual and little-studied data set: for-sale-by-owner and for-rent-by-owner real estate listings. These are sellers who set out to avoid middlemen — and their fees. How some of them then come to employ agents — and the resulting impact on the local housing market — affords a different view onto the much-studied role of intermediaries.

Don’t Take FSBO for an Answer

The research begins with a company that is indisputably an information purveyor, REDX, which, along with some competitors, does the messy work of scraping the internet for FSBO and FRBO listings, cleans them up and presents them in a useful and uniform fashion. For paid access to these listings, REDX’s customers are real estate agents, always hungry for leads for new listings and for additional properties they might show a buyer they represent. Real estate agents tend to view a FSBO listing as merely a seller who hasn’t yet retained an agent.

Aiello, Nadauld and Garmaise analyzed more than 385,000 FSBO listings on REDX from January 2016 through July 2018, with a particular focus on how the ranking of listings on the website impacted the likelihood of an agent converting a FSBO seller into an agent-represented seller, and the speed and price of transactions that resulted.

The focus on website listing order follows the realities of online search. Businesses have armies trying to game the search engine algorithms to get their links to land on the first page of search results. (Or they pay Google to plaster an advertising link on the initial search page.)

REDX updates listings multiple times a day and bases the order of listings on when the data was collected. That the listing order isn’t dependent on characteristics of the home, or the owner or pay-for-play, gave the researchers a clean insight into the pure power of placement.

They sorted the listings into quartiles based on where they landed in the display. A move from the third to the first quartile ranking increased the probability of the owner later hiring an agent by 3%. That is, prominent display seems to have increased agent focus on which properties to go after.

And this doesn’t seem to be a function of the owners just tiring of going it alone. While the researchers found a strong positive correlation between listings displayed high on REDX’s site and a REDX-affiliated agent snagging the listing, there was a negative correlation when looking at whether the client signed with an agent who didn’t use REDX.

Once the listing hired an agent — the median number of days it took for owners to relent and hire an agent was 77 — the researchers found that the home was more likely to sell quickly. They also found that a higher quartile rank translated into an average sale price that was 0.65% higher. (Note: Given that their data “do not contain precise measures of agent fees,” they are unable to address whether the benefit of hiring the agent is equal to the cost.)

Real estate agents typically collect about 5% of the sale price, split between buyer’s and seller’s agent.

Many real estate agents specialize by location, over time winning referrals from neighbors and building a hyperlocal reputation, offering potential clients local market studies and abundant anecdotes about recent transaction; this resembles the information agent role of a middleman.

The REDX data doesn’t fit into that model. The researchers found that REDX listings resulted in agents signing clients farther out from their core marketing area. This geographic reach was of particular interest to the researchers.

The information theory of intermediary value would suggest that a home sold through an agent would effectively disseminate data to nearby homeowners that could spur them to also sell or buy near that recent transaction. Geez, what a bargain, a buyer might think. Golly, they got $1 million for a shack like ours, a seller could conclude.

But incorporating listing and sales data from Zillow, the researchers found just the opposite: When an agent converts a FSBO listing, it reduced, rather than increased, nearby listings.

This suggests middleman as finder, not information expert. In an email exchange, Garmaise summed up: “Yes, we find that real estate agents mainly match existing potential buyers with sellers rather than reducing information frictions and bringing new buyers into the market.”