In the general field of securities analysis (predicting stock prices) there are two basic schools of thought, fundamental analysis and technical analysis. Fundamental analysis can be summarized as saying that stocks are worth a price based on their financial statements and that investors can profit by deeply understanding the details of said information. Benjamin Graham with his book "Securities Analysis" is an example of a fundamental approach to stock valuation.
On the other hand, a different school of thought belongs to the technical analysis camp, which states that stock prices have patterns and can be bought or sold for profit based on these patterns. Technical analysts frequently chart stocks and are responsible for the myriad types of charts available at any financial web site (such as Yahoo!) including "Bollinger Bands" and the like.
Without going into the relative validity of both theories (an endless topic in and of itself) I will glibly summarize the two models as "buy it because your detailed financial analysis says it will go up in value" vs. "buy it because it has been going up and other items similar to it are going up in price".
In real estate, the first thing that an agent will show you when you are looking for are the "comps", or prices of recently sold real estate in your target area with the basic characteristics you are looking for (# of bedrooms, square feet, closeness to schools, etc...). This is basically a technical analysis technique - "buy it because this is what other people are paying for it" and usually (until the recent past) followed with "and it has gone up x% from when it was first purchased".
To glibly summarize again, technical analysis basically "works until it doesn't work anymore". As long as prices are rising, these techniques look like genius. But when prices start falling, then what do you have to hang your hat on?
On the other side, you could view your housing investment through the eyes of fundamental analysis. For fundamental analysis, you could view your house purchase against the cost of renting - if it costs $1500 / month to rent a 3 bedroom house in your neighborhood that meets your needs, this could be viewed as a baseline for comparison against your potential housing purchase. To this model you would add in the "tax shield" of the fact that your interest expense is tax deductible while your rent expense is not deductible - this might mean that paying $2000 / month for a mortgage is equivalent to a $1500 rent payment. On top of this you'd have to add in property taxes of $300 / month, which may or may not be deductible depending on whether you are in the AMT or not. Then you also need to add in the cost of repairs since you will be footing these bills and not your landlord. On the other side, there is the potential upside of the fact that IF this asset (the house) increases in value, you will be able to capture this value, and not your landlord. Don't forget to factor in transaction costs, which include fees to your realtor and fees on your mortgage.
While I am vastly simplifying the "fundamental analysis" path for house buying, it is safe to say that most people are in the "technical analysis" camp although the act of signing up for a mortgage causes people to have to build a budget that would enable them to do the fundamental analysis based on rent if they wanted to do so.
Under the "imputed rent" model, valuations are sky-high. Try to calculate how much you'd have to rent out your house for in order to make money under this model; in many areas of the country such as New York or California where a middle class house costs between $500k and $1M, the results would be an outrageous monthly rent of many thousands of dollars a month. Justifying a house purchase based on fundamental factors is often impossible; only the feeling (characteristic of technical analysis) that "you'll sell it for more than you bought it" drives the process forward.
Another fundamental analysis model would be to compare the average wage in an area against house prices; while this method has a lot of range in its model it is safe to say that house prices beyond three times the average wage for a family are in the "danger zone" - if the average household income is $100,000 and house prices for a reasonable home are $500,000, then you are at a 5-1 ratio and under this analysis method housing prices are too high.
While I am vastly simplifying fundamental and technical analysis on securities and houses, I think that the base generalization that most houses in the bubble (deflating right now) were sold on technical factors is correct. Like the internet stock boom around 2000, people bought stocks because they were going up, and fundamental stock analysts at the time looked like fools saying that these internet companies couldn't create positive cash flow to justify the valuations, because the market kept showing that the technical factors were carrying the day. In the end, the fundamental analysts were right (in this case).
I think that there are a lot of reasons that asset prices move up or down, like securities. Both the fundamental and technical analysis tools have their uses, even though as a financially trained person (rather than a trader) I lean towards the fundamental tools instinctively (else I am just wasting my time, to use the full weight of the technical analyst's model). The domination of relatively simplistic technical tools contributed towards the housing bubble, in my opinion, and realtors are going to have to work harder than to just show the equivalent of a price chart the next time they want to talk a buyer into a large purchase.
Cross posted at Chicago Boyz