Let’s talk about housing prices

In my last post, I talked a little about the Grizzly’s experience in the Bay Area housing market. For the two of us, it was an amazing opportunity – we made about $500k from an initial $50k down in just 4 years, not bad returns if you can get them. But I want to take a step back and talk about some more boring details of housing market valuation and why we ultimately thought it was best to take our winnings off the table.

Measuring housing prices

First, we need to lay down some basics about how folks who know about these things talk about housing prices and affordability. Basically, when is the market hot when is it not. Housing prices are complicated. Compared to housing, stock and bond markets are miracles of transparency. In the stock market, you have assets that are very well-defined and homogenous. One share of Apple stock is exactly like any other share of Apple stock. Know the interest rate and maturity date on a US treasury bond and you have a pretty complete picture of what your returns will look like. On the stock market, prices are well known, vast quantities of shares change hands daily at published prices. If you want to know what someone will pay for your Apple stock – just look it up and you can get that price almost instantly. And if you want to get a general gauge of valuations in the stock market there are some well-known indexes (e.g. SP500, Russell 3000) that can be compared against well-known fundamentals (earnings, revenue, etc.). It is by no means simple but at least the metrics are commonly available.

Not so for housing. Housing markets are relatively opaque things. Every house or apartment is a little bit different. Transactions on any particular house are rare. And while there are some indexes measuring affordability and overall price levels these have nowhere near the amount of history we have for similar trackers in the stock market. However, let’s take a look at what is available and see what sort of picture we can develop for housing market trends. I’ll focus on the Bay Area because 1) it’s crazy! 2) it’s where we just made our own decision to exit.

Data!

Let’s get into some cold hard data on the current state of the SF housing market. We’ll look at two simple price indexes that we can use to track the overall evolution of housing prices over time. Then we’ll get into a couple additional data sets that add color to the picture.

Case-Shiller

The Case-Shiller Home Price Index is perhaps the most widely used measure of home prices in the US. Two pretty cool econ nerds, Karl Case and Robert Shiller, came up with it back in the late 80’s (Shiller is also the author of the book Irrational Exuberance, a great read on the causes and impacts of speculative bubbles). Surprisingly, prior to their work, there was no reliable tracking of housing prices in the US over time. Their index is a relatively complicated measure that tracks same home sales – i.e. sales included in the index are when the same home is sold at least twice across multiple years. The index also only uses single-family homes, so multi-units, condo’s and apartments are not included. The reason for this added layer of complication is to remove any changes in the total market caused by relative changes in housing quality – i.e. changes in types of size of houses on the market. Below is the Case-Shiller Index for the entire US from 1890 until today. That giant spike right after 2000 is the real-estate bubble. The new spike forming to the right is where we are right now.

Case-Shiller Housing Index 1890-2017. http://www.econ.yale.edu/~shiller/data.htm

The most significant finding to come out of their work was that across time there has been almost zero real price appreciation in the housing market (after accounting for inflation) over time. Essentially, national real-estate prices keep up with inflation and building costs, but that’s about it. If you’re expecting significantly higher returns from price appreciation alone on real-estate you should probably check your assumptions.

The Case-Shiller Index isolated for San Francisco looks basically the same. Spike in the 2000’s followed by yet another spike today.

Case-Shiller Housing Index for San Francisco MSA 1987-2017

All Transactions Index

The Case Shiller Index is a great measure of national housing trends, but due to its method of construction, it misses some information for some local markets. Specifically, the reliance on same house sales and lack of data for multi-family units can make it somewhat less reliable when you have an area, such as San Francisco, that has seen a relatively large amount of new unit construction and specifically new unit construction in condos and apartments. Just take a look at the SF skyline and count the number of new luxury condo/apartment developments and you’ll get a good indication of what I mean. As a result, I’m partial to looking at another index compiled by the Federal Housing Administration. This index is based on all transactions in a given area. It won’t have the same house to house accuracy of Case-Shiller but will pick up movements not captured by Shiller in rapidly expanding housing markets. The overall story is the same, but the all transactions index captures a much larger upswing in the last few years.

All Transactions Housing Index San Francisco-Redwood City- South SF. https://fred.stlouisfed.org/series/ATNHPIUS41884Q

Indexes vs. Income in San Francisco

Just looking at these graphs alone has one big gap. These indexes are not adjusted for inflation and (more importantly within a dynamic city like San Francisco) they are not adjusted for changes in income. San Francisco and the surrounding area has seen a dramatic run-up in per-capita income over the last few decades. The first argument that one would make rebutting any notion of overpriced housing would be to point out these increases. Housing is more expensive, but people can still afford it! Unfortunately, that’s not really the case. The two graphs below show the same data adjusted for per-capita income growth in the Bay Area. With this adjustment, we are now taking into account both inflation and real income growth.

Case-Shiller Index/Per-capita Income

All Transactions Index/per-capita income

While the Case-Shiller index vs. income has not reached the dizzying heights of the previous decade it has still risen to levels equivalent to those that occurred after the 1980’s boom that preceding the SF real-estate crash of the early 90s. The all-transactions index paints a much bleaker picture – prices across all transactions are nearing those seen at the top of the previous cycle, which saw a drop of almost 50%.

Affordability Index

The last dataset I want to examine is the California Association of Realtors Affordability Index. This index attempts to take EVERYTHING into account and estimate the % of families living in any given that are able to afford the median-priced home in that area. Everything includes property taxes, interest rates, down-payments, and insurance. It is intended to be a reliable indicator of how under or overpriced an area has become. The graph below shows the index for San Francisco.

CAR Affordability Index for San Francisco. http://www.car.org/marketdata/data/haitraditional/

The data shows that at current levels only about 13% of families can currently afford the median home in San Francisco – not an inherently stable place to be. The index hasn’t reached these levels since the bubble of the 00’s. This dynamic is even less stable if you believe there may be an upward skew to current San Francisco incomes due to an expanding bubble in the tech market (but that is another post and another story!).

What really caused the last bubble and crash

In any discussion of housing prices right now there will be inevitable comparisons to the bubble and subsequent crash in the last decade. When examining those comparisons we should make sure we have a full understanding of the actual underlying causes of the last crisis. History does not repeat, but it does often rhyme and it seems to be rhyming quite a bit these days.

A group of intrepid researchers led by this nice lady has been publishing a series of very insightful papers examining a new narrative of the credit and housing crisis of the last decade. The conventional wisdom is that a bunch of unscrupulous lenders funneled risky mortgages to unqualified buyers and repackaged them into a bunch of risky securities. Pieces of this are still true, but the element of this story they are quietly debunking is who exactly the borrowers were who drove both the expansion in borrowing and the subsequent expansion in defaults.

The thrust of their research is that the ones sucking up all this cheap credit were not a bunch of destitute subprime borrowers. Rather they were actually well-qualified real estate investors. Both the expansion of credit and the subsequent spike in defaults can almost entirely be explained by the behavior of this group. I.e. the subprime crisis wasn’t caused by a bunch of strippers flipping condo’s a la the Big Short. It was caused by a bunch of middle, upper-middle class, and upper-class real-estate speculators flipping condos and housing – boring folks like you and me. Data can be found in the paper here.

The graphs above show something pretty simple. The first set shows that the growth in borrowing was concentrated in those with 2 or more properties (investors) and in the highest quartiles of credit scores (lowest supposed risk). The second set of graphs shows the rise in both delinquencies and foreclosures attributable to those investors. They find that essentially all of the increase in foreclosures and delinquency during the crisis can be explained by these two factors – increase in borrowing by investors, and increased delinquency from investors.

There was little to no actual subprime story. The authors go on to talk about where the original subprime stories came from – essentially it was a time of the measurement error. The subprime narrative got started based on a couple studies that came out right at the time of the recession. Those studies looked at the credit scores of folks at the beginning of the bubble and used those scores to classify folks by risk. The problem with that methodology is simple, looking at a credit score from 1999 misses the fact that credit scores usually go up as people age. The subprime borrows originally identified were actually a bunch of 30-year-olds starting their real-estate investment careers.

The mortgage crisis was unfortunately caused by a bunch of folks that look like you, your rich neighbor, and the Grizzlies. A bunch of ‘investors’ looking for quick profits and then abandoning houses when those profits didn’t materialize.

Is it happening again?

So if we’re worried about real-estate prices we might start to look for evidence that real-estate ‘investors’ are starting run amok again. Unfortunately, the evidence is worryingly widespread. House flipping appears to be back with a fury – below is the simple Google trend analysis for ‘flipping houses’ and ‘flip house’. We’re now comfortably back to the levels seen last decade.

There’s a new generation of wonderful programs once again extolling the virtues of this particular form of ‘investment’ such as the aptly named ‘Flip or Flop‘, ‘Texas Flip and Move‘, ‘Masters of Flip‘, and more.

Additionally, the number of startups focusing on real-estate investing is exploding. I am now bombarded almost hourly by Facebook advertising for a new fractional ownership real-estate venture such as Fundrise, RealtyMogul, Realcrowd, or Iintoo. All of them promising blistering returns with very limited risk. Generally, it’s solid wisdom to avoid anything that promises this combination.

All of this points to, in my mind, another speculative bubble in real-estate. Many people will make some money. A lot more people will lose money. No one knows exactly when this bubble will eventually pop, but the Grizzlies decided to leave with our cash before the music stops.