Thoughts on Statistics?

so based on your numbers 30% decline in prices to match the 1998 numbers
interesting -- one can look at this as a bound on how much prices could drop from this point

of course 1998 is an arbitrary choice -- and prices have come down 22% from the peak in 2006 based on your numbers

If I choose 2005 to start my calcs and apply your formulae, the current average price should be $510k, so relative to the trend since then prices have dropped 27%

What happens if you go back to 1980 or so and then repeat the same analysis?

IN any case if we accept 1998 as a valid benchmark for today then your analysis suggests that the correction may be limited to about 30% more, which is interesting as an upper bound unless you factor in the collapse of the economy at large. Then there are no limits here or elsewhere

Looks like a pretty naive and simple forecasting model. Don't know who's it is.

For example, it ignores the change in financing costs during the period. T-bill yields started at 4.05% and were at 0.86% by June 30, 2003 as the Fed moved to aggressively pump up the economy. It ignores lending spreads to Treasuries - or LIBOR. It ignores new construction activity (supply). Et cetera.

I've seen similar charts on the stock market which suggests stock prices should also move steadily higher with very little fluctuation. But it doesn't happen. The pendulum(s) always swing too far. That said, sophisticated models can be useful in coming up with "long term" levels about which markets will fluctuate. Buy low, sell high.

you are right
the T-bill yields and other factors are likely more important than the naive factors from the OP.

Any thoughts on what comes up in the near term?
assume interest rates do not change much in the next 3 years and new construction stagnates + credit remains relatively tight despite low rates?

I wasn't sure where to find numbers back to 1980. The NAR statistics only go back to 1989.

I agree the vast increase in supply would tend to call for even less appreciation. The increased availability of credit would tend to call for more appreciation, though that is dicey, since many potential buyers still can't get loans at all no matter how good the rates are.

I don't know, all it tells us is what the rate of inflation has been just by itself....

http://www.usinflationcalculator.com/inflation/current-inflation-rates/

...which is roughly 25%.

Though one could say buying in 1998 at $500 per square foot is the same as buying at $625 per square foot today just on inflation?

Housing sales not directly skewing this number because if memory serves, inflation takes cost of renting in housing and not owning?

"I've seen similar charts on the stock market which suggests stock prices should also move steadily higher with very little fluctuation. "

I don't really think so.

I don't think any of these are supposed to be how they "should" move, they are just the historical norms plotted out to, and given that markets generally revert to the mean....

Bogle actually does a great long-term view on stock gains relative to earnings gains. I can't think of one of these guys who has ever said "which probably means you'll get x% each year", its almost complete universal to be talking about long term CAGR.

They exist to give some long term direction, and as such are usually pretty decent.

As year to year fluctuations, of course not, because Americans can get particularly stupid in short time frames.

truthskr: I also factored in population growth, since that's where any demand would come from. Of course as Topper points out supply has also increased so the relationship between population and value is probably more complicated than a simple proportion. With that caveat, if you compound inflation and population growth, you end up closer to 50% instead of 25%, so $500/sf in 1998 would be more like $750/sf in 2009.

nyc10022: I agree with your general point but I think it's unfair to single out Americans. Speculative bubbles existed long before the Mayflower.