Calling the Bottom

by Osman Parvez

The New York Times recently ran a story on how economists calculate the state of the housing market. The article focused on methods for determining over or under valuation, in an effort to indicate where the bottom lies.
Models built on these variables show that while some markets – such as California – are on a road to recovery, others – such as south Florida – have a way to go.

These signs cannot possibly tell the whole story, especially since they point more toward where prices should be valued than where they will be. But these measures are nonetheless helpful to anyone buying, selling or borrowing against their home sweet home.

“Anybody who says they know when it’s going to end with confidence is delusional,” said Karl E. Case... “But yes, you can get a sense of where things are going.”
Despite the obvious flaws to these models, I thought it might be helpful to discuss them briefly and suggest a few sources for more information. Should you decide to do the analysis, most of the data needed can be found within this blog.

The first model takes a look at the price-to-rent ratio, the cost of owning relative to renting. This is a common method, but it's often argued that renting isn't a direct comparable to owning. Rental property conditions can often be subpar. Raising a family in an apartment is quite a bit different than raising a family in a home. There are emotional factors. Plus, the impact of large numbers of student rentals in a community like Boulder also tends to skew things. The typical rule of thumb that student landlords use is $500-550/bedroom, minimum. It goes up from there based on location and the quality of the property. Non-student rentals run an equally wide range of prices, mostly dependent on location. For polling rental prices, Craigslist is your friend. If you're simply trying to figure out whether you should buy or rent, try using our model.

The second method examines per-capita income – to see if people can afford the homes they are living in. In 2005, the median income for a 3 person household in Boulder was $78,300 (source: City of Boulder). This is substantially higher than surrounding areas. High unemployment has a large effect on this model. As of June, unemployment in the Boulder MSA was 4.7%, compared with 5.7% nationally (source: BLS).

The third model in the article is based on appreciation rates. For your research on this, I suggest looking at the research index. Both the Basic Home index and the Luxury Home index will give you the longer term trend for appreciation rates in Boulder. Take your assumed base rate of appreciation and project out from a historical date, say 2002. The gap, if any, will tell you how far the market has to fall.

Unfortunately, these models don't take into account today’s mortgage rates. The Columbia Business School’s newest, unpublished analysis, attempts to account for this. According to preliminary results, Denver comes in at 5% overvalued. Compare that with Phoenix or Miami, for example, which both come in at 13% overvalued. The study says that Detroit is 12% undervalued.

In addition, a relationship to consider is the inventory-to-sales ratio, a measure of absorption. This is possibly the most unbiased model, based directly on supply and demand. It's also the statistic most discussed in this blog. We publish an updated analysis each month, for most cities in Boulder County. It includes a chart for inventory-to-sales.

Now before you get carried away with all this analysis, let me remind you of one thing. There is no crystal ball in real estate. Attempts to call the bottom or the top are generally met with failure. Those that manage to do it with any accuracy are lucky, not necessarily smarter than the rest of us.
…some experts argued that it was silly to try to build a mathematical model for the market’s overvaluation. Too much is unknown, they say, to make any predictions.

“I try to avoid house price forecasting,” said Paul S. Willen, senior economist and policy adviser at the Federal Reserve Bank of Boston. “Let me just say this, as an economist, that asset pricing is something we’re exceptionally bad at.”
research assistance: Evan
image credit: Justin

Like this analysis?    Subscribe to our client research report.     
Want to get blog updates via email?  Click HERE.       
Ready to buy or sell?  Schedule an appointment or call 303.746.6896. 
You can also like our Facebook page or follow us on Twitter.

As always, your referrals are deeply appreciated.  

The ideas and strategies described in this blog are the opinion of the writer and subject to business, economic, and competitive uncertainties.   We strongly recommend conducting rigorous due diligence and obtaining professional advice before buying or selling real estate. 


  1. Several of your assumptions in your "Buy vs Rent Calculator and Instructions advantage purchasing over renting.

    1) A Rental increase of 5.5% vs House Purchase Price increase of 4.0% defaulted in your spreadsheet? This is circumstantial, but we live in a very nice, single-family detached house in Boulder, zero maintenance fees and pay 1/3 to 1/4th! of what the mortgage would be if we purchased the house and have 0% annual increases. I would suggest one uses at maximum the same annual increase for housing and rent and shop around in the single family detached house rental market for pricing as the value is frankly, extremely good vs buying.

    2) New homeowners err to underestimating annual maintenance costs. How many try to sell a house after living in it for 5-10 years and realize they have to put in a new bathroom, paint everything, put on a new roof, re-landscape, update a kitchen, or patio, etc... Accountants and the tax code depreciate buildings over 40 years for good reason. That's what average annual maintenance costs add up to. Subtract the land from the home price and divide by 40 as a bare minimum maintenance fee. That's really just the aging of the structure.

    3) Your "Alt Return to Down Payment" uses US Treasuries as a proxy investment vs a house.

    How can one compare the return of riskless, extremely liquid, low entry/exit cost, unleveraged US treasuries securities to a one-of-a-kind, extremely illiquid, high entry/exit cost, 4x debt/equity leveraged home?

    Perhaps, one should assume the return of the S&P 500 stock index as a proxy. The S&P 500 average annual return, since turn of the century 1900 (almost 109 years of data) is 8.4%. This is a very, very conservative assumption, since (unlike housing) an S&P Index fund is extremely liquid, has low entry/exit costs and is not leveraged to the hilt like a home. A 5% drop in home value = 25% drop in equity. A 5% drop in an index fund = 5% drop in equity.

    Supposing one changes the spreadsheet to "No appreciation curve" and Osman's Instructions suggesting 2.6% housing annual appreciation rate and changes the alternative return on downpayment to a more comparable S&P 500 8.4%, beware housing may still be too optimistic. One could argue that the S&P is more likely to outperform its 109 year average, since it is in the 9th year of its worst decade performance (including the Great Depression years of the 1930s = 0%), whereas housing is in its best decade and may be hard-pressed to achieve its average annual return of 2.6% over the forthcoming decade.

  2. Several of your assumptions in your "Buy vs Rent Calculator and Instructions advantage purchasing over renting.

    You clearly don't know me. I bet you'd be surprised how many would-be buyers I advise to slow down and rent for a year (or longer).

    As for default values, I didn't intend bias when I built that model two years ago. And as you properly note, it's easy to input whatever assumptions you want. That's the whole point. YOU decide what you think the future looks like.

    I agree on new buyers underestimating maintenance.

    As for what rate you choose to compare with owning a house, again the choice is yours. If you're savvy and know you can earn a better rate than U.S. Treasuries, go for it. Plug in what makes sense to you. But the starting point is the risk-free rate.

    When it comes to equities, I doubt the next 100 years of the U.S. stock market will look anything like the past 100. And don't forget, you can't raise your family in the S&P 500.

  3. They aren't making any more land, and you have to live somewhere. Housing will see, at worst, what looks like a permanently high plateau.

  4. Anon,
    Like any other asset, the price of a house fluctuates. Ultimately supply and demand drives the market.

    On a national level, it will take many years before house prices raise above the bubble - but ultimately it will happen through inflation. On real terms, I'm in agreement. I doubt we'll see housing rise above values seen in bubble markets in our lifetime. But again, I'm talking about prices in real terms.

    In markets like Boulder, which did not participate in the bubble, I would expect house prices to flatten somewhat on average but then once again grow at least at the rate of inflation. Demand for housing in our market is driven by the fundamental desire to live in our community, not by speculative gains. Just in case you haven't been reading my blog for long, know that within that average, we're currently seeing softness in $1MM+ luxury homes. Meanwhile other parts of the market are more healthy. See the latest Silver Fern Report for more details.

  5. Thanks for the awesome charts in the Silver Fern Report! One question: How are properties that do not sell in the contract period counted in the DOM statistic? If they are omitted, do you have statistics on the number of listings that end without selling?

    Thanks again

  6. The charts in the Silver Fern Report show DTO not DOM. Thus, only properties that go under contract are shown. Expired and Withdrawn listings is really what you want and you'll find an analysis of that in my prior post.

  7. Earlier today, I received the following comment via email:

    Hi Osman,

    I've been following your blog for a while, and I'd like to start receiving your market analysis report.

    For some reason, I keep getting errors when I try to post to your blog. Here's a comment I wanted to post related to your withdrawn/expired listing:

    Although the withdraw/expired metric looks close to the previous year, a more meaningful metric would be to divide the withdraw/expired by the total inventory. This gives you a handle on what fraction of sellers are just giving up. If you do this, you'll see that, compared to last year, the fraction of frustrated sellers is much higher than in previous years. So, the relatively low inventory is not necessarily due to absorption, but due to sellers throwing in the towel. This skews the absorption number, because it decreases the inventory. In typical years, a decrease in inventory and concomitant increase in absorption would be mostly due to sales. But now, the decrease is driven primarily by houses being pulled from the MLS.


    Scot R.

    Thanks for your comment. Sorry you're having trouble posting it yourself.

    You've got a very good suggestion for analyzing capitulation. When I get a few minutes, I'll see what I can do.

    p.s. You've been added to the SF Report distribution list.

  8. Great data! Here’s a recent statistic that may indicate interest in Boulder-Longmont real estate is building.  Online searches on in July 2008 as compared to July 2007 for Boulder-Longmont properties jumped 17 percent. Check out the full report here:


Note: Only a member of this blog may post a comment.

Please Note

This document contains forward-looking statements. You are strongly cautioned that investment results are subject to business, economic and other uncertainties. There are no guarantees associated with any forecast and the opinions stated here are subject to change at any time. Always consult your financial advisor before making an investment decision.