How Close to the Top? - SF Bay Area Housing Affordability Analysis - (w/ Charts & Graphs!) by me

[J. Martin]

Executive Summary:

Because of the real estate I own, and prospective potential purchases, I wanted to know more about where we are in the real estate cycle in the SF Bay Area. Inspired by great students and researchers of historical data like Bruce Norris, Shawn O’Toole, and @Account Closed, I decided to dig into the Housing Affordability Index (HAI) in the Bay Area a bit more. So I downloaded the data and got to work! I found some interesting observations in my analysis, so I dissected it some more, made some charts and tables, and here it is for your informational pleasure.. Feel free to share it. A summary of my observations:

1. 1) We have marched the majority of the way back from the most affordable housing in 2011 towards the least affordable housing watermark in 2005 – 2007. In fact, most Bay Area Counties are more than 70% of the way back to peak unaffordability! Should they ever get that unaffordable again anyway?
2. 2) Historical lows and highs in the HAI appear to be good predictors of a market correction. And we’re getting closer..
3. 3) The difference, or “spread” between HAI’s in lower-end and higher-end counties decreases (convergence) as the market gets hotter. This statistic is also decreasing today towards 2003-2004 levels.
4. 4) This spread between lower-end and higher-end county HAI’s may be another good, new predictor of market corrections.
5. 5) Alameda and Marin counties are closest to being unaffordable, while Sonoma, Contra Costa, & even SF are relative deals.
   
   Disclaimer: This data is believed to be accurate, but may not be, and should not be relied upon. Do your own due diligence on any matter concerning your decisions.
6. Chart 1: Historical HAI in Bay Area Counties; Data Set 1991 Q1 - 2014 Q3.
7. Notice how close we're getting towards historical unaffordability. And see how unaffordability across counties converges during a "hot" market, and diverges during a "crisis" market.
8. ----------------------------------------
9. I started with wanting to know where each Bay Area county’s HAI is today, relative to its historical highs and lows. In addition, I wanted to see how far the HAI’s have come from their most affordable during the bottom of the crisis towards their least affordable during the peak of the last boom.
10. Well.. We’re most of the way there! – we’ve retraced more than 70% of our way to being as unaffordable as 2005-2007, from the most affordable in 2011, in the majority of Bay Area counties. And should it ever really get to being as unaffordable as it was back then!? There’s a lot of info in the table below. So let’s dig into the details a little more..
11. --------------------------------------------
12. Alameda, Marin, San Mateo, & Santa Cruz counties are closest to their least affordable watermark. Alameda County is the most significantly under its median HAI of 27%, at 19% today. Alameda unaffordability has also climbed 76% of the way back towards its least affordable, from its most affordable. Marin is similar, at 15% HAI today, and having marched 77% of the way back from peak unaffordability.
    Sonoma has the strongest value indicators, exactly at its median HAI. At 29%, it is 22 percentage points (ppts) from its all-time low of 7%. On the other hand, it was also the most volatile, and probably “overshot” most significantly during the boom. So it could correct significantly sooner than the 7% low HAI, which was the lowest of any Bay Area county at any time.
    Similarly, Contra Costa County has better relative value indicators, only having climbed 62% of the way from its most affordable towards its least affordable ever. At a current HAI of 19%, it is still 11 percentage points from its peak unaffordability of 8%.

Chart 2: High, Median, Current, and Low Historical Affordability Across Bay Area

See how far we've come back from peak affordability towards peak unaffordability? All counties are more unaffordable than their historical median, except for Sonoma, right at its median. Should they ever get to peak unaffordability again?

Surprisingly to me, Santa Clara County had mixed indicators, only marching up 70% of the way to historical unaffordability from the most affordable watermark. Currently at 21% HAI, it is the most affordable besides Sonoma County of the 8 counties here. It also has 10 percentage points to go before hitting its historical low of 11% HAI.

And perhaps the biggest surprise of all, San Francisco! Despite constant complaints about it being one of the most expensive places in the US, it has only climbed to 67% of its historical unaffordability, just below the middle of the pack. The HAI is only 2 ppts below the median, and has 7 ppts to go before hitting the min. It also has the lowest range of HAI, so these ppts to the worst represent more relative movement in SF.

Table 1: Summary Data & "% of the way to peak unaffordability"

You'll notice from Charts 1 and 2, and looking at the top right corner of Table 1 that "higher-end" and "lower-end" counties tend to converge in unaffordability during a hot market. In fact, almost surprisingly so. This correlation tends to hold true over cycles also. I have various theories I've written, but I'd like to hear others' opinions first..

So here's a chart measuring the difference between the least affordable and most affordable HAI among Bay Area counties (convergence), as a potential indicator of risk building in the system - especially if most counties are approaching low historical affordability...

------------------------

Look Below at Chart 3.

If you’re thinking, “What the hell is J talking about in this xmas-colored chart?! Convergence? Who cares?” Stay with me for a second. This could be expanded to inland areas vs. coastal, Class A vs. Class C, etc. The idea is about how frothy markets “bid up” lesser-quality assets, they converge, then everything blows up and they deviate significantly again (then buy!). I see it as sort of an “overbought/oversold” indicator. You can see this convergence as all Bay Area Counties become utterly affordable approaching 2005 and 2007, and back in 2000. How close does that convergence appear to be taking place today? Getting closer!

Chart 3 - High / Low HAI Spread - 8 Bay Area Counties

I think that if the Spread is getting into the danger zone, especially as HAI’s march closer towards their record unaffordability, there are potentially big risks building in real estate. As of 9/30/14, the Spread is at 14 ppts, after bouncing from 7-16 ppts since the beginning of 2013, within the warning zone of 15 ppts, and crossing into the danger zone of 10 ppts briefly. It has not yet approached the “RUN!” zone of 5 ppts.. yet.. But again, should it every approach the 2005-2007 levels again..? How close are we? (See red question mark below)

We’re back towards 2003/2004 High / Low Spread, but took a brief dip towards 2006-2007 levels. Lower-end and higher-end counties are converging in unaffordability towards what they did in the prior boom years. Couple that with the fact that most Bay Area Counties have marched more than 70% of the way back to the least affordable HAI in history, and we have a couple strong indicators that risk is building and we’re close to the top.

How close are we to reversion / disaster?

With the understanding that prices should probably not get as unaffordable as they did in the last boom – because it was driven by so much nonsense financing – should we ever get to such little affordability as we did in the last boom, especially from 2005-2007?

So how much less should the HAI be stretched in this upward cycle, relative to the last? Only 80% of the prior distance from peak to trough of HAI from the 2006 cycle? 70%? 90% of it?

Will we need looser lending to maintain current price levels, or any additional gains?

How will the interest rate impact on affordability be offset, if not by lower prices, or looser lending?

Can the market still increase without the benefit of continually lower rates to improve affordability, as happened in prior cycles?

Should the HAI spread between low-end and high-end counties ever go as low as it did during the last boom? (In other words, should low-end areas become just as unaffordable as high-end areas in the future, again?)

Knowing these things, what are you going to do? Does it impact your strategy? Does it impact your outlook?

For myself, I’m re-thinking some of my plans, and staying cautious. There are signs we’re approaching new record unaffordability. And that can impact the whole market..

What do you think?