New Home Sales at Record Low

Off Topic Politics & World Issues
81

[quote]Bill Roberts wrote:
You still wound up choosing a peak that, last time around, didn’t sustain itself at all.
[/quote]

If you consider only 1950-2000, yes it appears the trendline would be pretty flat. However, just like what you say about the peaks, over the last 30 years, neither of the two valleys could sustain themselves either, and these low points (1984, 1997) would appear to fall directly on the trendline. This would lead me to believe that if we only want to consider these 50 years, then we do not have a linear trend. (Checking the chart really quick)

I guess I would be wrong, the line does definitely trend up, and gives a formula y=.1488x-181.79, which gives a value of 117.30 for 2010, still no where near the 40% suggested, but lower than I would suggest. However, this trend required throwing out all of the most recent data and ignores all appreciation that happened during the last decade that was indeed sustainable, ie, most of what happened between the coastal states, and does assume a linear trend.

Actually, the more of history I cut out and the more relevance I put on just the last 30 years supports my argument even more, I do agree that prices 100 years ago have little relevance to todays. So let’s look again at the trend for just the last 30 years. y=2.0236x-3907.6 gives (2010, 159.84)

[quote]
I don’t care whether you go back only as far as say 1960, or you go to 1950, or you go to 1965, or what have you.

I’d suggest re-running your analysis so as to not go back so extremely far, and try running it both with obvious bubbles and obvious short-term crashes removed, as we are not interested in predicting bubble or crash prices, but prices that might be considered long-term typical.

If you do this, I’m pretty sure your trendline will be quite different than you’re calculating now, and present prices will be substantially above it.[/quote]

The problem with doing this is that housing prices have not shown an obvious trend since 1975. Disregarding the bubbles and crashes disregards government programs that were behind these occurences that are still in place today.

This is why I displayed the second graph (the ratio of prices to costs), to give a better baseline of the value of the building and factor out location and as much other appreciation as possible. If we start with the actual cost to build and just look at what the markup is over that, we see a graph with no distinct trend, but it also shows a markup well below most historical values.

82

There most certainly is a trendline that can be computed by using only data within say the last 50-60 years or whatever value of your choice, and if desired by omitting periods of obvious bubbles or temporary price-depressions.

The fact that that trendline is fairly flat in no way invalidates it.

One problem with your price/cost ratio is that it assumes that construction cost is the major cost item involved. In rural areas with small lots this is true, but in very very many cases the price of the lot exceeds the building cost of the home.

Factoring out land valuation – except for rural small lots – results in an almost meaningless calculation.

83

[quote]tedro wrote:

[quote]Bill Roberts wrote:

It seems likely to me that your trendline is largely shaped by your decision to use ancient history. Personally, it seems to me that the further in the past one goes the less relevance the data has. What housing prices were in 1910 really has no predictive value, I would say, on the worth people could best be expected to place on houses in the future.
[/quote]
Actually, the more of history I cut out and the more relevance I put on just the last 30 years supports my argument even more, I do agree that prices 100 years ago have little relevance to todays. So let’s look again at the trend for just the last 30 years. y=2.0236x-3907.6 gives (2010, 159.84)[/quote]

A huge problem with limiting to just the last 30 years and using all these years is that the last 10 years of it is a period which is now widely recognized as a huge bubble.

That will of course produce exactly the effect on a trendline that you got.

84

[quote]Bill Roberts wrote:
There most certainly is a trendline that can be computed by using only data within say the last 50-60 years or whatever value of your choice, and if desired by omitting periods of obvious bubbles or temporary price-depressions.

The fact that that trendline is fairly flat in no way invalidates it.

One problem with your price/cost ratio is that it assumes that construction cost is the major cost item involved. In rural areas with small lots this is true, but in very very many cases the price of the lot exceeds the building cost of the home.

Factoring out land valuation – except for rural small lots – results in an almost meaningless calculation.[/quote]

If we were trying to determine an exact price, I would agree with you. However, we are only trying to determine where the trend will be in the next few years, in which case this ratio helps determine this trend by showing (attempting to anyways, the volatility of the graph makes it somewhat difficult) a worthwhile baseline that costs are unlikely to fall below.

85

[quote]Bill Roberts wrote:
A huge problem with limiting to just the last 30 years and using all these years is that the last 10 years of it is a period which is now widely recognized as a huge bubble.

That will of course produce exactly the effect on a trendline that you got.[/quote]

Obviously. I don’t think anybody honestly trying to determine the trend would do this, and I certainly wasn’t suggesting that prices need to rebound 20%. I still don’t think the last 10-12 years should be thrown out completely though, so how do you propose we weight them?

86

The question of what data should be discarded is in general a problematic one because bias can so readily be introduced by choice of method.

When it comes to measurements – let’s say the physical properties of some thing and the issue is that it’s known that certain measurements were wrong for technical reasons, it’s completely valid to throw them out for that reason.

On the other hand, changing methods because one doesn’t like the outcome is invalid. E.g., it is intellectually dishonest to, after counting votes according to standard and agreed method and getting a given result, then doing a recount according to standard and agreed method and getting the same result again, to hen decide well let’s change the method of counting in a way that predictably would favor one side.

And even worse, let’s say on STILL getting the same result, to decide yet again that there should be yet a further count on yet different rules again predictably favoring one side, with the only reason to stop the endless series of counts being, finally getting the desired outcome.

That being an example of what not to do.

Simply from the standpoint of coming to personal opinion for my own personal use, I’d eyeball the thing and excise sections that the eyeball says are aberrations.

From a more technical and valid standpoint, removing an equal number of low data points, e.g. average yearly value, as high data points until anomalies are removed (there probably is a developed and validated method for establishing this, but I don’t know it) is an acceptable method.

As to this specific set of data, there are very serious problems that cannot be compensated for except that one may be willing to stipulate certain things and then to consider analysis valid under those stipulations.

For example, it’s certainly arguable that home prices are related to long term mortgage interest rates. When rates are low, people of a given income can afford, and thus may pay, higher prices for a home.

So do we want to stipulate that interest rates will remain at current low levels?

It’s also arguable that home prices have a relation to amount of down-payment typically required for a mortgage. If so, do we want to stipulate that required money-down will stay as low as it typically is now?

If we had an analysis that gave reasonably well-supported correction factors for these and other things, then we could compute a trendline, for any given set of stipulations, much better than where the price data is affected by such things, but we compute trend without accounting for that.

For all these reasons, and probably more, it is more rule-of-thumb here than precise analysis.

Just going by rule of thumb, I see a bubble from 1999 to present that hasn’t quite settled to what had been a fairly flat historical valuation from the 1950’s or so onwards, excluding this lengthy bubble and a few rather brief deviations.

87

Btw, on the vote counting example and why I used it:

It actually is a general problem, not a narrow one, and vote counting gives a good example of it.

All the time in analysis, unless one is just really really smart and somehow is of the sort that an analysis is not run until SO much thought and time has gone into it that there is no room for the same person to improve it, one comes up with an idea for how to analyze some data, runs it, and then realizes reason why perhaps the analysis should have been done differently.

And it is ALL too easy for bias to creep in here.

For example, let’s say two people are wedded to ideas (bad situation already) on home valuation. One is wedded to the idea that homes are now overvalued, and the other is wedded to the idea that prices have returned to proper (in the historical sense) valuation.

Each is wedded to his view because his gurus said this and each has written a number of things on it and would have to publicly contradict himself if he now changed his position.

Each person had the same professors in college, have very similar thinking mathematically and analytically, and happen to come up with the exact same method, independently.

Each on running his analysis, which he really put a lot of care into and is quite sure has had all possible sources of error properly accounted for, upon crunching the numbers computes… homes aren’t overvalued now.

Well what may well happen?

The guy who is wedded to that idea is done.

The guy who is wedded to the idea that homes are overvalued is convinced that obviously he must have screwed up, and goes and finds a “problem” in his analytical method. He keeps finding such “problems,” and correcting them, until he gets the outcome that homes are overvalued. At which point he’s convinced that the analysis has had all its problems corrected and is a completely valid one.

The thing is, that was a lot harder to explain than the vote-counting example.

But the principle is the same: decision on whether to change method of analysis cannot be allowed to be influenced by the outcome of the analysis. It takes care to avoid this.

88

The only way to buy a house today is on the Courthouse steps. Real estate is going to crash big time. Commercial RE is going to hell very soon and residential will follow.

Mansions in CT that sold for $1,000,000 in 1929 were selling for $25,000 in 1933.

89

Just as an odd, useless “what if, though it wasn’t” trivium:

Henry Ford was offered $1 billion for the Ford Motor Company in either 1928 or early 1929. He turned it down.

Imagine if he’d taken the cash.

Similarly to what you are saying on mansion prices, if he’d had that billion in cash in the early 30’s, he probably could have bought just about all of American industry, or utterly vast amounts of real estate.

Of course, doing as he did still wound up doing not exactly badly for his heirs, so it’s not as if it proved disastrous to turn down the billion bucks. (Which with inflation, would be comparable to, I haven’t looked it up, about $20 or $30 billion today.)

90

Not to mention the fed has no more ammo to buy mortgagae backed paper - meaning rates will go up without further govt intervention.

The tax credit effect is wearing off. They have hit diminishing returns to stimulus. Thats where the shit starts to hit the fan - theyhave no control and deflation takes over. What isnt known is if we get Japan or Weimar when they crank up the quantitative easing.

And the adjustable rate reset schedule has 2 peaks in late 2010 and mid 2011 IIRC that are larger than the first subprime reset in total, though presumably the initial data has been skewed higher than it is now by people refinancing while the fed holds rates down. Plus, its better paper - more alt-a and less subprime. A little silver lining. I imagine the fed will do ANYTHING to keep rates down through this period.

Also, there is HUGE shadow inventory. Meaning homes over (IIRC) 90 days past due (which have a 98% chance of foreclosure) that the banks will not begin foreclosure on, Homes owned by banks that they havent put on the market yet to not take the write down and not crush the weak housing bid, and homes in the foreclosure process. Some estimates are up to 5 years of total inventory (traditional and shadow) exists.

Housing is going to double dip hard and long. I would stay out of housing speculatively.

91

I am looking to buy foreclosed homes and fix them up for renting in my city…lots of colleges and professionals moving in. Also though the housing market in my city, pittsburgh, has been relatively unscathed compared to most areas… If anyones done the above I would definatley love a first hand perspective. Seems like a nice cash flow opportunity

92

[quote]666Rich wrote:
I am looking to buy foreclosed homes and fix them up for renting in my city…lots of colleges and professionals moving in. Also though the housing market in my city, pittsburgh, has been relatively unscathed compared to most areas… If anyones done the above I would definatley love a first hand perspective. Seems like a nice cash flow opportunity [/quote]

Make sure that the title is clear. I looked into this as my first approach to buying a house. There were places in Fox Chapel, Mt. Lebanon, and Upper St. Claire going for less than 10K, but had assloads- 200K-1.5mill. in debt attached to them. The transactions are about as complicated as chinese math, and even a real estate attorney that I consulted told me to pass on this idea.

I figured after that it would be best to let a bank take them and clear the title, then let them sell it at what ever they were trying to recoup with a deed free of liens in a competitive market.

A friend of mine owns a bunch of places in Oakland and rents them out to college students. He does well, but that has been an accumulation of property and wealth over time.

I’ve only been looking in the south hills, but good places at a good price are out there. Contractors are snatching them up for cash very quickly though.