in which Manhattan Loft Guy bravely calls BS on the Market Trend Meme Of The Day

fools rush in …
Refill your coffee cup if you intend to get beyond the first few paragraphs here. My executive summary is that the much-repeated meme accounting for the huge sales volume at the end of 2012 is probably BS and certainly can’t be proven by any of the numbers I have seen publicly reported. Number-driven skeptic that I am, I don’t think there were many sellers with a big enough gain to trigger much anxiety over being taxed at the fearsome top marginal rate in 2013 rather than the capital gains rate in 2012, as the candidate pool for such skittish sellers is ones likely have had to have owned for ten years or more, and be within the top 30% of sales prices, and have sold a coop rather than a condo. As you will see, the search for the ‘extra’ sellers in the last quarter who (per The Meme) hurried sales to avoid 2013 is difficult, labyrinthine in a (sadly) typical Manhattan Loft Guy way, and unrequited; they may exist in some numbers, but they can’t be found in macro data.

You’ve been warned …. Anyone who has any interest in the state of the overall Manhattan residential real estate market has already seen one or another of the articles from the Manhattan Media Division of the Real Estate Industrial Complex about the state of The Market at the end of the year and/or seen one or another of the 2012 Fourth Quarter sales reports from the from the Manhattan Brokerage Division of the REIC. So you know that this lede from the New York Times yesterday (Manhattan’s Housing Market Ends 2012 With a Sales Rush) is typical:

The Manhattan real estate market went out with a bang in 2012, with the number of sales rising by as much as 40 percent in the last three months of the year — mainly because sellers were in a hurry to close deals before new tax laws kicked in with the new year.

You can find other versions of “mainly because sellers were in a hurry to close deals before new tax laws kicked in” elsewhere (including the BHS guy quoted in The Old Grey Lady, and this little datum: the “number of sales over $10 million rose 44 percent, to 23, from 16 a year ago”), but for reasons you will see below, I prefer the more restrained overview from The Miller in his 4Q Report for the firm formerly known as PruDE:

There were a record 2,598 sale in the fourth quarter as looming changes to federal tax laws and general economic improvement elevated activity in an already improving housing market.

That “mainly because …” is my target. Although (because?) it is simple, it is (probably) BS. Sadly, we may never know (in the sense of “see from data”) whether the late 2012 sales results were “mainly because” of the tax issues, and if we can know (etc) it won’t be until another quarter or two of data arrive.

I will explain why I am a Tax Uncertainty Skeptic, and then what future data may tell us.

show me the numbers
I tend to be skeptical about any Mass Herd Theory that asserts that a few thousand Manhattan buyers and a few thousand Manhattan sellers made the decisions about their money and their properties that they did. But that does not mean that I don’t like (appreciate) trends, just that we too often reason backwards from results to “reasons”. (And there is never a shortage of people willing to give reasons, no matter how little fact-based their personal reasons may be.) In the world according to Manhattan Loft Guy, you would see more commentary on facts (“in 4Q12 many more sellers sold [and buyers bought] than in any prior 4Q”, with analysis of which market segments accounted for the surge) and less on supposition (“out of fear that taxes would rise in 2013”). But we do not live in a Manhattan Loft Guy world, alas.

My skepticism extends to much larger, more transparent, more efficient markets, like the stock market. I will grant you that a financial reporter can survey traders who work with huge institutional investors who can actually move the Dow Jones average on a day, and get some sort of consensus that these huge stockholders became sellers in response to, for example, a new economic report. But even then there are still individual portfolio-balancing and other motivations for individual huge stockholders, and every trade has, of course, two sides. For every huge seller who didn’t like whatever news is the consensus market-driver-of-the-day there is a huge buyer who saw an opportunity to buy. (I don’t want to take this stock market analogy too far, as I will quickly get over my head and there are certainly tax motivated stock transactions at year-end.)

I wonder about who the Tax Uncertainty Sellers might be who could be so numerous as to move the 4Q Manhattan residential real estate market. I have no doubt that some of those ‘extra’ $10mm sellers might have been motivated to avoid 2013 income and capital gains rates, but look again at how many (rather, how few) there are in that rarified atmosphere:

The number of sales over $10 million rose 44 percent, to 23, from 16 a year ago.

Frankly, I think the REIC Media license of anyone who calls a difference of 23 sales compared to 16 sales “44%” in a tiny niche of more than 2,500 data points is guilty of a numeracy infraction, even though the computation is correct, because we are only talking about 7 extra sellers. If I were writing about a market move, I’d try to resist the temptation to generalize from the decisions made by 7 extra sellers, even if the BHS guy knows each of them personally.

Of course, the Times reporter is using those 7 as examples, but I doubt the efficacy of those hyper-rich sellers as a proxy for the real group that comprises The Newsworthy Group: (using the BHS / Halstead numbers in the Times) the 40% extra deals in 4Q12 over 4Q11 number 656. (Using The Miller’s numbers, the relevant population is 587 extra sellers, compared to the prior year quarter.) Let’s talk round numbers about those 600 or so folks who, if the Market Meme Of The Day is correct, ‘hurried’ to sell at the end of 2012 instead of waiting more patiently to sell in 2013.

That is what we are talking about, right? That a significant enough number of sellers to move The Market sold more quickly because of Tax Uncertainty than they would have without that Tax Uncertainty. That is, in fact, exactly what the BHS guy said, without defining what he means by “a lot”:

“Without the tax law changes, a lot of that would have gone on into January or February.”

how much money are we talking about?
If we start with the largest market segment in The Miller’s report we can get some concrete numbers to work with. Per The Miller, the coop market was larger than the condo market in the last quarter (1,558 coop sales, 1,040 condo sales) and the largest coop segment was 1-bedrooms at 41.2% of the 1,558 sales; if we throw in the coop studio segment at 16.7%, they comprise 57.9% of the coop market. The median price for that 1-bedroom segment was only $560,000, which hardly seems worthwhile to talk about as Tax-Uncertainty-motivated transactions. Here’s why:

The first $250,000 of gain for any individual tax-filing owner who sold at $560,000 is not recognized at all. So before you get into any Tax Uncertainty, we are talking about a hypothetical 4Q12 seller who sold at, say, the $560,000 median for 1-bedroom coops, after having bought at significantly less than $310,000. That’s probably a 2012 seller who was a buyer from 2002-04. How many folks in Manhattan stay in studios or 1-bedrooms for 8 to 10 years? Not very many. But they would need a buy-in starting point even than $310,000 to have any taxable gain, so the potential for Tax-Uncertainty-motivated transactions at this level will be limited to people who held much longer even than 10 years.

Try a population with higher values: using The Miller’s quintiles, 60% of the condo sales in 4Q12 were at a median of $1,287,500 or lower, with the fourth quintile with a median of $2.09mm. Let’s focus on that fourth quintile as a numbers playground for potential Tax-Uncertainty-motivated transactions.

Assume most of those 208 condo sellers file joint tax returns, where the non-recognition limit is $500,000 in gain. Assume they are worried about being in the group whose marginal federal tax rate would go all the way up to 39.6%, with any recognized gain from the condo sale being taxed at that marginal rate instead of at the 2012 capital gains rate of 15%.

The median seller in this quintile sold for $2.09mm; had they bought at $1.6mm or higher, that median seller in the ⅘ quintile would have no taxable gain at all. But if you assume a $1mm gain, based on an original purchase of $1.09mm (let’s not get into how their adjusted tax basis would likely be higher), their taxable gain would be $500,000. Real money, right?

Looking only at federal taxes (simplicity is a virtue) that $500,000 gain would be taxed in 2012 at 15% ($75,000) or (the supposed fear was) in 2013 at 39.6% ($198,000). Now we are talking about motivating money! But remember that the premise at this point is a $1mm gain; how many of the ⅘ quintile sellers held long enough (10 years??) to get a $1mm gain?

Cut the hypothetical gain to $750,000, and we are still talking about a long holding period for that $2.09mm seller, and these numbers: at prevailing 2012 capital gains rates the recognized portion of gain ($250,000) would be taxed $37,500; the feared 2013 marginal rate would bite at $99,000, still a possibly motivating difference.

Yet remember, we are trying to account for ‘extra’ sellers. In The Miller’s 4Q11 report, the ⅘ condo quintile had almost exactly the same number of sellers as the 4Q12 ⅘ quintile: 202 v. 208. In other words, at the condo quintile in which the Tax Motivated sellers might begin to reside (assuming long holding periods, etc) there are no extra sellers whose motivation we need to explain. (Unless the BHS guy knows those 6 extra sellers, also.)

On the larger, coop, segment of the market (those 1,558 sales) you know how to do the playground math. The ⅘ quintile for coop sales had a median sales price of only $980,000, so fully 70% of coop sales in the quarter (1,091 out of 1,558) were sold below $980,000.

The top coop quintile is where the action is, if there is any action at all. Because, as you see from The Miller’s condo numbers being essentially equal 4Q12 to 4Q11, all of the ‘extra’ sellers were coop owners. That top 30% is 467 coop sellers, roughly equal to the entire population of extra sellers we are putting under the microscope, though the top 30% in 4Q11 were 299. One-third of that 2012 group sold above the coop top quintile median of $2.2mm. As above, couples with a gain of $750,000 from their original purchase might worry about possibly paying an extra $61,500 in taxes. There are 268 ‘extra’ sellers in this Top 30%,but not many of them would have held long enough to have that level of gain to worry about.

the numbers don’t show any ‘hurried’ deals
The proposition to be tested is that of the BHS guy in blue, above: that “a lot” of 4Q12 activity was hurried into 2012 instead of selling (more naturally, without hurry, one assumes) in January or February 2013.

You’d’ think then, that there would be fewer days on the market for these extra 587 sellers (otherwise, there is no ‘hurry’). The data are not precise enough to show days on market by the quintiles that might show Tax Uncertainty motivation, but the overall Miller numbers show much longer time to contract for 4Q12 sales compared to 4Q11: 177 days in the last quarter, 130 in the prior year quarter. No evidence of ‘hurry’ there.

(By the way, here is what The Miller suggested about the increase in coop sales in the last quarter over the prior year quarter: that 4Q11 coop sales total was “seen as an anomaly”.)

You’d think, then, that a seller who wanted to rush to contract and close in 2012 would have taken a deeper discount to asking price. Again, the data are not precise enough to discounts for the quintiles that might show Tax Uncertainty motivation, but the overall Miller numbers show that 4Q12 sellers were slightly less inclined to negotiate than 4Q11 sellers: the ask-to-contract discount was 3.7% in 4Q12, 4.9% in 4Q11. No evidence of ‘hurry’ there.

Finally (!), if there were a significant number of 2012 sellers with Tax Uncertainty motivation, you’d think there would have been … you know … more 2012 sellers. There were, in fact, an unusually small number of 2012 sellers, if by “sellers” we include both people who signed contracts and people who offered their apartments for sale. The big story of the 4Q12 reports, in my mind, at least, is the continuing low level of inventory.

At the end of 2011, in addition to the 2,011 “selling sellers” there were 7,221 “offering-to-sell sellers” (inventory). That total of 9,232 is dramatically more than the 2,598 “selling sellers” and the 4,749 “offering-to-sell sellers” (inventory) at the end of this past year (7,347). No evidence of ‘hurry’ there.

some were, but some always are
Don’t get me wrong: some of the folks who sold in the last quarter sold then because they did not want to run the risk that they would pay higher taxes if they waited into 2013. I just don’t think there were enough to move the market. And I think there are always people who do year-end deals instead of January deals for idiosyncratic financial reasons; they are just (you’ve heard me say this before) not a big enough group of similarly-minded sellers to move the market.

future data might, or it might not
I am not going to flag the reports at this point, but I assume that anyone who has gotten this far is knowledgeable enough to take as a given that there were market footprints of tax-advantaged thinking when the first-time home buyer tax credit (of 2009?) expired. In those days, the entry level sales were demonstrably weighted in the quarter in which the credit expired, and sales in that segment were demonstrably reduced in the following quarter. In other words, niche volume showed a reallocation of buyers into one quarter rather than the following one.

So one might look at overall Manhattan residential sales volume in the first quarter of this year for evidence of a reallocation of sellers into 4Q12 rather than 1Q13; in theory, if 1Q13 volume is low compared to past first quarters that would support a calendar-based shift of “should be 2013” sales back into 2012.

I will look, I promise. But the problem I anticipate is that we can already confidently predict that 1Q13 sales volume will be relatively low because we already know that inventory is remarkably low. As noted above, at year-end The Miller counts 4,749 units for sale in Manhattan; last year there were 7,221. So I don’t expect the low sales number we will be talking about in 3 months to be evidence that sellers hurried 2013 sales into 2012. Whether there are other hints in the next quarter’s report remains to be seen.

Not worth speculating about. Certainly not worth extending this post about.

© Sandy Mattingly 2013