I’m glad to have better substance to what had been simply an impression of mine that 1998 prices (or the chart indicates 1997 but not with much difference) probably represented about historical average value, with market values since then being overpriced relative to long-term historical or even grossly overpriced.
The chart goes only to 2006, so it doesn’t show where we are now, but we have not fallen back yet to 1998 valuations.
In other words, when prices drop to 55% of 2006 values, then that will be in line with typical historical valuation. (Or to 50%, if you want to go more than 50 years back in time in determining an average, which I don’t think ought to be done.)
EDIT: Correction would be necessary for inflation between 2006 and present, which I haven’t done.
I’m glad to have better substance to what had been simply an impression of mine that 1998 prices (or the chart indicates 1997 but not with much difference) probably represented about historical average value, with market values since then being overpriced relative to long-term historical or even grossly overpriced.
The chart goes only to 2006, so it doesn’t show where we are now, but we have not fallen back yet to 1998 valuations.
In other words, when prices drop to 55% of 2006 values, then that will be in line with typical historical valuation. (Or to 50%, if you want to go more than 50 years back in time in determining an average, which I don’t think ought to be done.)
[/quote]
Looking at the new chart, home prices have already fallen to 1990 levels.
Also, look at sheet 2 of the excel spreadsheet and note the indices that were used in determining prices. The index changed in 1934, 1953, 1975, and 1987. In 1997 it was changed to the Case-Shiller index. Shiller is the one that made this chart, and also put it in his book “Irrational Exuberance”.
I just signed the purchase agreement on a house, tomorrow is inspection and financing, closing at the end of March.
What do you all think of that?
[/quote]
I think if you like the house it was probably a good move. Sure, you may not have hit the exact bottom, but it’s impossible to time it perfectly. I think the general consensus is that we have basically bottomed out and may even be headed back up, despite what the sky is falling crowd around here may think.[/quote]
I do like the place. I plan on being there at least 10 years, in which time I will be paying down the principal at a much higher rate than just the monthly payment plan.
The price/value on it is very good- 20K below its 2005 tax assessment, plus it is a good sized lot and modestly sized but custom built house (read: not a McMansion). The township in which it is located was dead set against the housing development boom, and has very strictly controlled new construction for the past 20 years.
The interest is looking like 5.25%. We will find out for sure tomorrow. It seems to be the fortunate product of patience and diligent hunting.
[/quote]
10 yr. time frame for a house you love…sounds like a good purchase for you. Also sounds like you got a pretty good deal.
[quote]Bill Roberts wrote:
I’m glad to have better substance to what had been simply an impression of mine that 1998 prices (or the chart indicates 1997 but not with much difference) probably represented about historical average value, with market values since then being overpriced relative to long-term historical or even grossly overpriced.
The chart goes only to 2006, so it doesn’t show where we are now, but we have not fallen back yet to 1998 valuations.
[/quote]
[quote]Bill Roberts wrote:
I’m glad to have better substance to what had been simply an impression of mine that 1998 prices (or the chart indicates 1997 but not with much difference) probably represented about historical average value, with market values since then being overpriced relative to long-term historical or even grossly overpriced.
The chart goes only to 2006, so it doesn’t show where we are now, but we have not fallen back yet to 1998 valuations.
[/quote]
I am going by your “try again” chart as I get nothing with the other.
And in that chart, the end point, which appears to be approximately the present, is higher than any point near 1990 including 1990 itself.
And aside from the fact that the current price has not dropped down to that point, why would you pick an extremely narrow peak that wasn’t maintained for any length of time? Isn’t that cherry-picking?
Doesn’t 1997 or 1998 seem more representative of typical value for the past several decades?
[quote]Bill Roberts wrote:
I am going by your “try again” chart as I get nothing with the other.
And in that chart, the end point, which appears to be approximately the present, is higher than any point near 1990 including 1990 itself.
And aside from the fact that the current price has not dropped down to that point, why would you pick an extremely narrow peak that wasn’t maintained for any length of time? Isn’t that cherry-picking?
Doesn’t 1997 or 1998 seem more representative of typical value for the past several decades?[/quote]
And spikes up tend to overdo going in the opposite direction when they do pop. If we get an overcorrection, real estate (in normal times) drops another 30% at least.
[quote]Bill Roberts wrote:
I am going by your “try again” chart as I get nothing with the other.
And in that chart, the end point, which appears to be approximately the present, is higher than any point near 1990 including 1990 itself.
And aside from the fact that the current price has not dropped down to that point, why would you pick an extremely narrow peak that wasn’t maintained for any length of time? Isn’t that cherry-picking?
[/quote]
So sorry for not being precise with my comments. Take a look at the data in the spreadsheet:
1989: 127.51
2009.125: 130.55
I am willing to bet 3 is well within the margin of error of the data.
I could have easily chosen 1979 or 2000, before housing prices really took off. This is no more cherry-picking than picking 1998 or 1984, which are at the bottom of valleys. I also think it is fair to say that housing prices will increase in real dollars over time, given the limited supply of prime locations.
For some more interesting data, go into the spreadsheet and divide column b by column d to get a ratio of housing prices to the cost to build houses. This gives a general idea of a houses worth when you factor out the cost of a prime location and also let’s you see how speculation and appreciation outside of that which increases utility have affected prices. That chart is attached, we are now at a pretty low level there, historically.
Well, if you think that comparing to a momentary peak isn’t cherry picking and is just as valid as choosing a calculated mean value or a fairly typical value seen within a range of years, then that is your opinion.
However a person interested in whether prices are at a level such as to be commensurate with typical historical values they would be ill-served by choosing such a point for his comparison.
You are, by the way, now posting a different graph and switching the comparison. Your “try again” chart, which is the one we were discussing and which you claimed I needed to look at more closely, shows more than the 3 point difference of your new graph.
Tedro, for clarity, here’s the “try again” chart you posted before and which was the basis for my response saying that prices had not fallen to the same as 1990, to which you disagreed and told me I needed to look closer.
Do you now dispute the validity of this chart, in favor of the one you’ve now posted? If so, why?
From this chart, I think a person would be quite in error to conclude that prices have fallen to levels commensurate with typical values in say the 1950-1998 period.
Now if he believes that 1990 has some great truth to it and so long as prices are near there then they are unlikely to drop much, then that could be personal opinion and a reason to pick 1990 as a reference date. Myself, I wouldn’t recommend doing that. I think it is cherry-picking a peak that showed no tendency to sustain.
[quote]Bill Roberts wrote:
Well, if you think that comparing to a momentary peak isn’t cherry picking and is just as valid as choosing a calculated mean value or a fairly typical value seen within a range of years, then that is your opinion.
[/quote]
Ok, let’s make a trendline. I don’t think a mean value would be representative, as it does seem we both agree that prices should have trended slightly higher. A simple linear trendline gives an equation of y=0.51x-889.04. So for the year 2010 this gives a value of 136.06, which is actually higher than the data value at 2009.875 of 134.21, which could lead one to conclude that prices actually dropped to far and are now correcting themselves, as seen by the slight uptick over the last year. Feel free to play with the chart with different trendlines, but I think a linear trend is a good assumption.
Just as they would be using 1998, a low point, as a reference.
[quote]
You are, by the way, now posting a different graph and switching the comparison. Your “try again” chart, which is the one we were discussing and which you claimed I needed to look at more closely, shows more than the 3 point difference of your new graph. [/quote]
The “try again” chart is still the one we were discussing. I simply posted the other one because it provides a different statistic to forecast the same trend, that is housing prices. Looking at building cost lets us ignore some of the variables that are common in housing prices and can provide for a better baseline. This is a different statistic so there is no reason for it to show the same 3 point difference.
So people are still unemployed, their outlook is bleak, real estate prices have yet too fall, but government statistics say everything is fine?
Mmmmmmh…[/quote]
I don’t necessarily disagree with you, my aim was simply too show that a catastrophic recession is not imminent. As said in the article, the economy is still very unstable, but i have found that industry in this country has picked up immensely over the last two quarters, and the latest government statistics support this observation.
[quote]Bill Roberts wrote:
Tedro, for clarity, here’s the “try again” chart you posted before and which was the basis for my response saying that prices had not fallen to the same as 1990, to which you disagreed and told me I needed to look closer.
Do you now dispute the validity of this chart, in favor of the one you’ve now posted? If so, why?
[/quote]
The only thing I dispute about the original chart is how high the 2006 peak was. I am very skeptical of the numbers Shiller claims. I think a few markets skewed the results greatly. Most parts of this country did not see housing prices double. It is very fitting for him to put this graph (as of 2006, ie headhunters version) in a book titled “Irrational Exuberance”. Aside from that, I do think that most of the graph is probably fairly accurate. The runup starts at the right time and the end result puts us at the right price point, I just doubt the peak was truly that high throughout the nation.
I disagree. The trendline also disagrees. Perhaps with no government intervention you would be right, but we still have a high level of interference, plus the limited supply of prime locations, that leads me to believe we will never again see levels like we did from 1950-1980.
[quote]
Now if he believes that 1990 has some great truth to it and so long as prices are near there then they are unlikely to drop much, then that could be personal opinion and a reason to pick 1990 as a reference date. Myself, I wouldn’t recommend doing that. I think it is cherry-picking a peak that showed no tendency to sustain.[/quote]
I only chose 1990 as a reference date because there was not any panic associated with the housing prices in 1990. With a quick glance, it also appeared to me that the trend would put prices somewhere near this level. I was actually surprised that the trendline put levels in 2010 as high as they did, but this is likely becuase of the peak we have recently seen. For the sake of this argument, I’ll exclude everything after 2000 and find a new linear trendline:
I get y=.2712x-428.53. This gives a value of of 116.58 for 2010, which would suggest prices still have ~10% too drop (far less than the 40% suggested earlier), but if anything is cherry-picking this is definitely it. In order to disregard these numbers, you would have to come up with a good reason not too also disregard the low numbers of the actual great depression and world wars.
I stand by my hypothesis that housing prices have dropped to their expected levels and we are unlikely to see them fall anymore.
Ok, I’m kind of embarrassed that I didn’t think of this earlier, but if we want to dispute the graph and talk about cherry picking, maybe we should be questioning the starting year. Any guesses why the author picked 1890 instead of 1900, or even 1894 (123.98)? A quick look at the chart gives a good guess. Funny that he picks a starting year that’s part of a downward trend.
[quote]tedro wrote:
Ok, I’m kind of embarrassed that I didn’t think of this earlier, but if we want to dispute the graph and talk about cherry picking, maybe we should be questioning the starting year. Any guesses why the author picked 1890 instead of 1900, or even 1894 (123.98)? A quick look at the chart gives a good guess. Funny that he picks a starting year that’s part of a downward trend.[/quote]
Not necessarily. Real estate is traditionally a bad investment. In a bubble, it can often take decades just to get back to the nominal value pre-bubble. Examples abound.
[quote]tedro wrote:
Ok, I’m kind of embarrassed that I didn’t think of this earlier, but if we want to dispute the graph and talk about cherry picking, maybe we should be questioning the starting year. Any guesses why the author picked 1890 instead of 1900, or even 1894 (123.98)? A quick look at the chart gives a good guess. Funny that he picks a starting year that’s part of a downward trend.[/quote]
[quote]tedro wrote:
I only chose 1990 as a reference date because there was not any panic associated with the housing prices in 1990. With a quick glance, it also appeared to me that the trend would put prices somewhere near this level. I was actually surprised that the trendline put levels in 2010 as high as they did, but this is likely becuase of the peak we have recently seen. For the sake of this argument, I’ll exclude everything after 2000 and find a new linear trendline:
I get y=.2712x-428.53. This gives a value of of 116.58 for 2010, which would suggest prices still have ~10% too drop (far less than the 40% suggested earlier), but if anything is cherry-picking this is definitely it. In order to disregard these numbers, you would have to come up with a good reason not too also disregard the low numbers of the actual great depression and world wars.
I stand by my hypothesis that housing prices have dropped to their expected levels and we are unlikely to see them fall anymore.[/quote]
You still wound up choosing a peak that, last time around, didn’t sustain itself at all.
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.
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.
