Making sense of market indicators, part 2: days on market.

“Days on market.” Sounds like a real stem-winder of a market indicator, doesn’t it?

“Market indicator.” Sounds like a real stem-winder of a topic, doesn’t it?

Don’t worry. Because days on market, which I’ll occasionally shorten to DOM, is a fairly simple and sometimes useful market indicator. As market indicators go.

“Days on market” has a straightforward definition: it’s the number of days a home is on the market before it sells. But already this straightforward definition has one important variant.

That’s “cumulative days on market”, the total number of days a home is on the market unsold. Why “cumulative”? Because unless the local Multiple Listing Service regulates this, it’s not uncommon for homes to be taken off the market temporarily in order to restart the DOM count. It’s like rolling back the listing’s odometer. Home has been on the market, say, 55 days with no offers. Since homes in that market typically sell in two weeks or so, 55 days on market makes it clear that this listing is overpriced. Seller responds correctly and lowers the price. To make the 55-day-old listing appear new, listing agent has seller cancel or withdraw the listing briefly, then put it back on at the new low price and with the DOM reset to 0. Home gets to forget its past and start life anew.

Our local MLS now prohibits manipulations of DOM, and in addition, it gives anyone with access to the real, subscription-based MLS (as opposed to the Web sites that only claim to offer access to the MLS) the total number of days the home has been on the market, both cumulative and in its current iteration. An agent who looks up the relisted property described above will see not only 0 DOM but also 56 CDOM.

Why is it important that an agent or a buyer know how many days a home has been on the market? For one thing, a home that hasn’t sold by its “sell by” date is obviously a home that the market in its collective wisdom has decided is overpriced. But don’t make the common mistake of wondering “what’s wrong with it?” Something is wrong with every home except maybe the Taj Mahal, and even that masterpiece must be a real bear to maintain. No, the only thing wrong with a stale listing is that its price doesn’t reflect its condition (and a stale listing can be either a teardown or a new home), its location (peaceful cul-de-sac or major commute artery), local schools (internationally renowned or under state administration) and the other factors that determine value.

Another reason is that DOM is—or should be—or could be—your indicator that the home can be had for less than list price. This is tricky, because some sellers won’t budge from their price for a second, no matter how much the market disrespects it. We’re seeing this now in the mid-range and top-end price ranges, even where sales are declining, as sellers without the pressure of an impending foreclosure or layoff successfully resist buyer attempts to grind on the price.

From all this talk about the relationship between DOM and sales price, you may think that there’s an inevitable correlation between these two market indicators: the longer homes take to sell, the less they’ll sell for. This correlation would certainly simplify things if it existed. It’s almost certainly true of individual homes. Price a home too high and it’ll linger on the market, drop off every buyer’s and agent’s radar screen, lose all momentum and eventually, if it does sell, sell for less than if it had been priced correctly. This is easy for an experienced market watcher to believe but difficult for him or her to prove empirically: since each house is unique, direct comparisons are difficult if not impossible.

But in the overall marketplace, the correlation between prices and DOM is far from clear. “How can this be?”, you ask. “Didn’t you just say that the longer a home is on the market, the less it’ll sell for? If this is true of one home, why not of the market as a whole?”

Great question! While I call Security, let’s look at what if any effect more days on market has had in three distinct local markets, representing the top end, the low end, and the midrange: Palo Alto, East Palo Alto, and the medium-priced neighborhoods of southern San Mateo County respectively. Let’s first compare days on market, then sales price, between March 2005, generally considered the peak of the market, and March 2008, well into the downturn, to see if there’s any discernible connection between these market indicators.

First, days on market:

DOM1

As you can see, two of these markets have seen DOM go up dramatically between March 2005 and March 2008; the time it takes to sell a home in midrange San Mateo County and East Palo Alto has more than doubled. But note that Palo Alto DOM is identical, at a sparkling 15 days. Then why is Palo Alto included? You’ll see in a moment.

With days on market more than twice what it was during the boom, you’d naturally expect sales prices in these two markets (and perhaps even all three) to go down substantially. Let’s see if they have:

DOM2

Not exactly. Yes, East Palo Alto prices have nosedived, by almost 40 percent, but midrange San Mateo County prices are up about 5 percent. Remember, DOM for each market is more than twice what it was. And Palo Alto prices are up over 20 percent—without Palo Alto DOM declining.

What’s going on? Partly it’s that the pace of a real estate market doesn’t have to be lighting quick to sustain prices. I keep coming back to the real estate truism that “prices go up quickly and go down slowly”. Like any truism, this one isn’t always true, but it’s more likely to be true in affluent neighborhoods and in strong local economies. A homeowner not under undue duress to sell will have a far more relaxed attitude toward how long his home is taking to sell, or even if it does. A strong local economy that brings people to the area gives sellers the fallback of turning their homes into rentals “until the market gets better”. (My experience in the similar market of the early ’90s tells me that this is probably a great time to be leasing and managing rentals.)

But that’s only part of it. We’re also seeing real estate’s version of survival of the fittest. Increasingly, only the better homes are selling and, all things being equal, better homes sell at higher prices. In many neighborhoods in this downturn, lesser homes don’t get the chance to inflate DOM or drag down sales price because they don’t sell. You may see far more homes taken off the market unsold than sold, often by a ratio of 2 or 3 to 1, but you see little if any effect on average sales price.

On the other hand, sellers in less-affluent East Palo Alto typically have less of a safety net than sellers further up the price chain, or no net at all, especially as equity vanishes (and for many recent homeowners it never existed) in low-end neighborhoods. And increasingly, EPA sellers are banks with only one goal: get out, and quickly. Banks haven’t managed to bring down EPA’s DOM, but they’ve sure helped it find its new price level.

One final example of how DOM can mask a market downturn: over the years I’ve seen a few instances where the handful of homes that did sell in a slowdown sold so quickly that if you looked only at the average DOM you’d think you were in the middle of a boom.

So much for “data don’t lie”.

Which brings me to one last instance where, if the data aren’t exactly dissembling, they’re not making clear declarative statements either.

Again, if DOM goes up, you’d expect to see the market slowing, right? Often that’s true, but not always. Invariably one early sign of a market recovery is rising DOM. It’s a subtle hint that buyers are reaching further back on the shelf. Remember what I said about hot properties selling quickly even in a bust? In a recovery, demand rises to the point where not only the first rank of homes sells but also the second rank that’s hung around a while, and even the third rank that’s loitered even longer on the market, and so on. As buyers start working their way from the front of the shelf to the back, older listings sell and temporarily inflate the market’s DOM.

It’s important to realize that DOM is relative. A recent article in a national publication cautioned sellers that in the current downturn homes that once sold in “only” six months were now taking a year or more. Here on the mid-Peninsula, homes that once took an average of fifteen days to sell are, in some cases, still selling in an average of fifteen days. That’s only in sought-after cities such as Palo Alto, and in a handful of top-end neighborhoods in other cities. Most inventory in the less-expensive midrange cities takes longer to sell these days, now averaging about thirty days on market. And at the bottom end of the price range, where most of the subprime damage is focused, homes are averaging about seventy days on market. Yes, occasionally I run across a low-end home that’s been on the market a year, but it’s always a short sale, which is to say an unsalable home.

If all market indicators have only one purpose—to predict where prices will go—then DOM is an interesting but potentially misleading signpost along the way. It’s just another sign that, despite what the simplifiers say, the real estate market is anything but simple and easily decoded.

In two weeks: Making sense of market indicators, part 3: sales prices.

copyright © John Fyten 2008

Leave a Reply