Agents have always claimed to add value that exceeds their fee. A new study by Collateral Analytics (CA) provides the most powerful evidence to date that this is true.
The CA study, released August 1st, shows that homes sold with the help of real estate agents sell for 5.02% to 7.26% more than homes sold directly by owners without agent assistance.
This means that DIY sellers have the worst of both worlds. They do all the work of marketing, showing, negotiating and closing on their homes and put less money in their pockets than their counterparts who hire an agent. They save an agent fee of 5% to 6% but get 5% to 7% less for their house, making them worse off in some cases than if they hired an agent to handle the transaction.
Furthermore, FSBOs may be losing more from their bottom line because they end up paying the fee of the agent representing the buyer.
More on that issue of the buyer’s agent fee in a moment, but let’s first look at why the CA research has more credibility than prior studies:
- Collateral Analytics is not a real estate brokerage firm or real estate industry advocacy group. CA creates “automated valuation models” (AVMs) that predict the value of real estate based on public information. The Zillow Zestimate is the most well-known AVM.
- The CA study included a massive sample of homes. It included 1.35 million residential properties sold in 770 different zip codes nationwide from January 1, 2016 to June 30, 2017. This sample represents approximately one in every six homes sold during that 18-month time frame. Prior research on this topic was limited to single metropolitan areas or involved very small sample sizes.
- CA used three different methodologies instead of just one and their methodologies reduced or eliminated problems with “selection bias” that have plagued prior efforts to investigate this matter.
One of the methodologies is particularly intriguing. CA looked at the sale price of FSBO and agent assisted sales compared to CA’s estimate of value for each home. If 123 Main Street was a FSBO, they checked its actual sale price against CA’s predicted value from its automated valuation model. Likewise, if 789 Center Drive was an agent assisted sale, its actual selling price was compared to the AVM’s predicted value.
The findings showed that homes sold by agents tended to bring prices above predicted value, while FSBO properties typically sold at a discount to predicted value. The differential was computed for all properties in each zip code and the average differential in all 770 zip codes was 5.626% in favor of the agent assisted owners. Owners using an agent got, on average, a 5.626% higher sales price than sellers going it alone.
The two additional methodologies used in the CA study indicated higher sale prices for agent assisted transactions of 5.02% and 7.26%.
The sellers using an agent got 5.02% to 7.26% more for their property, which more than offset the 5% to 6% fee paid to the agent.
Let’s go back to that issue of the buyer’s agent fee. How often do FSBO sellers end up paying it on behalf of the buyer? The CA study had no way to evaluate this variable, but data from other sources allows us to make a reasonable guess about the impact of this factor.
We know that 92% of resale buyers use a real estate agent to help with their purchase. As a practical matter, this makes it likely that many or most FSBO owners end up paying the fee due to the buyer’s agent. Many FSBO sellers proactively offer to do so to entice agents to show their properties.
Bottom-line, FSBOs get 5% to 7% less for their home on average and may cough up another 3% of their sales price to pay the buyer’s agent. They can easily be at a 10% deficit to their agent assisted brethren.
The sellers using an agent pay 5% to 6% in agent fees and end up being better off because (1) They get a 5% to 7% higher sales price that more than covers their agent fee and (2) They avoid a separate payment of 3.0% to the buyer’s agent.
Agents are a good value, bringing higher net proceeds to their seller clients than sellers that decide to go it alone.