My Notes on the Real Estate Industry

A Quick Preface

I’ve recently heard several intriguing pitches from startups in the real estate sector. The way these brilliant entrepreneurs explain the future of real estate makes me think that this archaic industry is evolving. As a result, I have spent the last two weekends trying to “get up to speed” on the industry and key trends. The way I typically do this is read papers, dissect raw data sets, and speak to industry experts. I usually keep track of my notes by turning them into whitepaper-ish industry overviews.

For the real estate market, I’ve broken my notes into two sections. The first section focuses on the TAM and market growth. The second focuses on key trends in the market.

I’ve decided to publish these notes in case they are helpful for others. As a caution, this is pure research, without a thesis, and pretty raw. If anyone has other perspectives or findings, I’d welcome the chance to pick your brain over coffee (

Real Estate Market Overview

Market Size: The BEA calculates real estate and construction to be a $2.8 trillion market, or 16% of total GDP. This includes new construction, rent (gross and owner imputed rent), and services. While this initially seems like a huge number, it is dwarfed by the realization that it excludes resale property.

To get a more realistic view of market size, you need to aggregate: (1) real estate GDP, (2) plus existing home sales, (3) plus existing commercial sales, (4) plus new mortgage originations. If you aggregate all of this[1], you get to approximately a ~$5T market — almost 2x what GDP captures! Even if you further reduce this to just transaction fees, real estate is a GIGANTIC market (assumptions range from 3–10%).

Breaking down the market further, you’ll see it is comprised of ~40% commercial real estate and ~60% residential real estate. Within each of these areas are various sub-sectors, with unique drivers and trends. I decided to dive into the residential sector because of a hypothesis that the market is highly fragmented and under-penetrated by technology (there are opportunities in commercial as well).

Market Growth: The second requirement for a strong market is growth. There are several ways to look at growth in real estate, and rather than pick the “right” metric, I triangulate around a few data points.

Per Ibis, real estate sales and brokerage has grown at ~5–6% a year. However, what does this mean? Is pricing or volume driving this? Will this sustain going forward?

The first, and maybe most obvious data point is price. Residential real estate prices have grown by ~6% a year since 2012, which is well above historical rates of 1–2%[2]. This blended rate doesn’t show the whole picture as price varies drastically by city[3]. Real estate is broken down in primary and secondary markets. The primary markets are Boston, Chicago, Los Angeles, New York, San Francisco and Washington, D.C. Secondary include, but are not limited to, Miami, Denver, Dallas, Atlanta, Salt Lake City, Memphis, etc. If you look at the below chart, you can see that both are experiencing price increases, however there is a gap between primary vs. secondary cities that started around 2007 (think: flight to quality during the recession). Net / net…price will likely continue to increase, but potentially at lower rates and will vary by city.

Source: Here.

A second measure could be the number houses being built, or housing starts. The growth here looks a bit sluggish compared to pricing. The chart below suggests that new houses are starting to recover, but are still low for our growing population (the yellow line). However, it is important to remember that new home construction is a smaller market than resale. In order to get a full picture of how the market is trending, we should look at existing home sales.

Source: Here.

Looking at existing residential home sales we see an interesting trend. Existing home sales seem to be shaky, but still growing when you look at long-dated trends. Looking at the period from 1980–2005, we see to be at a similar slope, which gives me hope that we’re starting to see a return to more normal growth in existing home sales, despite being well below 2006 peaks.

Source: Here.

Conclusion: So my initial question of whether real estate is a large and growing market seemed to pass. Definitely a large market. Growth seems to be good from a price perspective, and slowly recovering from a volume growth perspective. Going forward, I think momentum will slow as we trend from “hyper recovery” mode to “sustaining” mode. However, the market still seems interesting.

Real Estate Trends

Trend #1: Resurgence of Rental

Today homeownership rates are close to a 50-year low; a major reversal of the historical trend of increasing home ownership. This is a big change. From 2000 to 2005, homeownership rates increased from 64% to 69%. This trend reversed during the recession and has continued downward to 63. This means ~8M people switched from homeowners to renters.

So, what happened? The answer is not straight forward.

Source: Here.

First, nearly ~7M Americans lost their homes during the recession, which combined with high unemployment, caused many to turn to the rental market. In addition, credit tightened making it difficult to secure a mortgage. The housing market collapse was a very deep, but perhaps temporary shift (meaning there is some level of expected recovery).

The second macro factor is more permanent. The under 40 population is delaying homeownership until later in life. The chart below shows that the greatest increase in the number of rentals from 2004–2013 came from the 25–45 age groups.

Source: Here.

So, why is this population choosing to rent? It is a confluence of factors that relate to demographics, debt levels, underemployment, limited credit history, and low inventory of starter homes. From the chart below we can decipher at least one clear trend: mortgage debt has been flat while student debt has been increasing.

Source: Here.

More people graduating college with large debt burdens, and don’t want to take on a mortgage. This, combined with lower housing starts and stricter lending requirements, means that homeownership is being put off until later in life. Additionally, demographic changes such as the average age people get married and have kids, is happening later in life. The result of this is an inflection point and a permanent change in the real estate market: renting is becoming the new-normal for folks under 45.

Trend #2: New Financing Structures

The 30-year fixed rate on mortgages is at a 50-year low. In the past two months, we finally saw this rate increase from ~3.5% up to ~4%, however this is still dramatically below historical rates.

Source: Here.

Here is the weird thing: despite these low rates, mortgage originations have trended downwards. Whenever I see a long-dated chart like these, and see that we are at an all-time low, it makes me think through “what innovations will come, that address this market change?”. Traditional mortgage products aren’t work for everyone. The only population that has increased originations with these “cheaper” mortgage rates are the super-prime. What products will work for those of us with a FICO score below 760?

Source: Here.

Several startups have come up with creative financing solutions.

- Online mortgage lenders are creating marketplace that allows individuals to secure mortgages from non-traditional lenders.

- Some originators are creating shared mortgages. This allows you to take on a fractional share of a mortgage so you don’t have the debt burden of a full mortgage.

- Some startups are providing a platform where you can sell off part of the equity in your residential property. Proceeds can either be used to shore-up liquidity or pay down part/all of a mortgage.

It is important to note that >95% of all mortgages are written or guaranteed by the US government. Therefore, while there are several new financing structures being created by startups, this is an overhanging factor in real estate: the government.

It feels like we are at an inflection point in the mortgage industry where the current product isn’t working. I liken it to the following: if you are lowering the price of your product, and it still isn’t selling, maybe you should change your product. While this is an oversimplification, it gets my point across. Our current mortgage structure is archaic and not working for young folks. Creative solutions that are enabled by technology are the future.

Trend #3: The Application of Data

While hard to quantify, there has been a transition within the real estate market from being a “relationship” business to a data-focused business. On the residential side, Redfin, Trulia, Zillow, Homeaway, and all provide a plethora of information on apartment and home sales. In my mind, the first wave of real estate investing was largely focused on bringing this fragmented and offline knowledge base into repositories that could be accessed by the masses. This, in combination with county level data publicly available, gives many startups a good entry point into using these large datasets to predict trends in housing. Redfin, a DFJ portfolio company, allows the public to look at neighborhood level data and understand how pricing, inventory, and sales are trending. This level of transparency within the industry is unprecedented.

With these large data sets, I am positive innovation will come. From what I’ve seen, deep-learning can have huge impacts when applied to specific verticals with large data repositories. I recently spoke to a friend who was a data scientist with me at Square and now works at Opendoor. He commented to me how much more robust the models are in real estate, rather than lending, because of the swaths of large public databases. In an industry where access to information is key, the ability to analyze large data sets through deep learning techniques is critical.

So are we at an inflection point in the use of data in deep learning for real estate? Deep learning is definitively at an inflection point. And real estate data is finally available in mass scale. Therefore, I can only conclude that the combination of the two will come in short order.

Bringing real estate data online was version 1 of real estate startups. Using this data to intelligently drive new markets and products will be version 2.

Trend #4: Deregulation?

This isn’t a trend per se, but wanted to flag that changes could come from deregulation of the banking sector. The Trump administration has suggested that there will be an easing of bank regulations and oversight. In fact, just this past Friday, Trump signed an executive order to scale back Dodd Frank. This could be a major change, as it would impact mortgage origination and potentially reverse the trend of credit tightening.


What does this all mean? In my view, the housing market went through a really, really, tough time during the recession. As the market recovered, it emerged as a changed entity. How people buy houses, when people buy houses, and what they purchase, is not the same as it used to be. And rightfully so. This huge market is recovering at a time when new enabling technologies will allow it to operate differently than in the past.

While I can’t say for sure how real estate will look in 10 years from now, I can confidently say that it will be unlike the last 50 years. This old industry is getting a makeover.


[1] Existing residential 2016 home sales $1.3T (5.45M x $232,200), plus existing commercial real estate sales of $567B, plus new mortgage originations of ~$400B.



Like what you read? Give Adena Hefets a round of applause.

From a quick cheer to a standing ovation, clap to show how much you enjoyed this story.