bostonbubble.com

[GenXer]

I can not find the actual study, which apparently came out in 1992 in a publication of some sort. I couldn't find an electronic copy, though I'm sure its possible to locate this study, since Nassim Taleb was able to do so.

[admin]

[Teavo]

GenXer,

I think you misunderstood my point about the struck-by-lightning issue and how it's provably a low probability. To use your example of the "extreme random 20 sigma event" I never once said that I or anyone could "predict this event". What I said is that such events are rare and this is provable. To bring this back to the world of making predictions in real estate, the point is that some information is better than no information, which means we can make sensible predictions (that are better than a random coin flip) when we have information upon which to base those predictions.

[GenXer]

You are right about the response type, but I was simply trying to illustrate that if you ask an 'expert', say, whether to buy a house or not, you will get a binary answer. The 'expert' may then come up with an explanation which is totally bogus based on a ton of assumptions, all of which he never bothered to test. So, while the knowledge to produce an answer is not binary, the result of applying this knowlege is: you either buy a house or you don't.

My argument with regards to models follows from above. If we know for sure that the model is not correct (and I mean, a model which is trying to predict results subjected to extreme randomness, not the 'nice' random walk type randomnes), then we may simply discard this model, because USING it will cause more damage than not using a model at all. If we prices followed a random walk, then by all means, we can try to predict where they were headed, with a certain probability of being right (if prices were higher than some mean, then they'd more likely come down than not). HOWEVER, if the distribution of prices is a power law, which allows for extreme events with a much higher probability than a random walk would, THEN if we try to use intuition, or past results, we can end up badly surprised by outcomes which can not be derived from our models. In the worst case, we can end up losing a lot of money. Because we underestimate the degree and the magnitude of the randomness we are dealing with, it may be better to use only the 'affordability' test to decide whether to buy a house or not. I know this is counterintuitive, and I know we like to make order out of very chaotic information, but this is reality, whether we accept it or not.

[GenXer]

Teavo: Let me reiterate what I said before: making predictions about rare events (by using past real estate data, for example) introduces errors which are so large that you have no chance (except by luck) to be right. If you are trying to guess what will happen based on past data, you may as well throw darts at the chart (if you could aim along the time axis), and you have more chances of being right doing that than you have by trying to use the 'knowledge' that you gained from 'information' which you think you possess.

Yes, you have it right - if an expert did have a poor 'batting average', then you would do well to do the opposite. Or flip a coin. Or throw a dart at the wall. In fact, if you want to pick individual stocks, you may actually do quite well if you use darts. At least, you will beat plenty of 'stock analysts'. There are many examples where using what we think as knowlege fails us miserably. What if you knew the earnings of every company ahead of time? Would you be able to make money off of it? I think the answer may surprise you. We are constantly fooled by randomness by underestimating its effect in our life. Our perception is altered by thinking of sanitized gambling at casinos, where the odds are known beforehand. When you do NOT know the odds, making estimates is much harder than it seems.

[john p]

As far as experts in the market now, I think that some people have really good insight about certain things so you've got to talk to a range of people. When I was looking, it was kind of frustrating to talk to people that weren't up with current issues.

I'm curious to see what types of first time buyer incentives the government is planning and what they might do to the mortgage rates. I have a suspicion that they might offer incentives to buy cars as well, but the whole "Buy American" thing is kind of a third rail politically. On this one, Obama has to bridge the fact that he beat Clinton by his criticism of NAFTA, but he's trying to be this global guy who speaks to hundreds of thousands of people in Germany, so for the people following closely it was interesting to see him be a Union populist/ protectionist at the same time the overwhelming majority of the international population wanted him over McCain. Now is where the politician can't hide, he has to pick sides and either be pro-trade or isolationist protectionist.

I think this applies because we dont' know who the real Barack Obama is and if it is possible to usher in an "Era of Reponsibility" while bailing out people that got illegal and fraudulent mortgages. I give him some credit because it seems he's trying to build an ideological basis from moderates; trouble is Pelosi got out in front of him and got her spending. Because the whole "Bailout" or "Recovery and Reinvestment" was so large and we knew so little about what was in it, I kind of thought that it encompassed the real estate issues and health care issues. I guess what we're talking about is a series of trillion dollar spending sprees. It's kind of like playing the card game "Hearts" and trying to shoot the moon.

[Mark]

[StallionMang]

To my points, posted many pages ago

1) Unemployment generally leads housing prices - losing a job adds to supply, takes away demand. Housing and bad mortgages led, but unemployment will add further downward pressure.

((wish I remember where I saw this...historical overlay of employment vs. housing. Anyone know?))

2) The historical graph makes it very clear, I think, how much of a bubble it was. The BIG PICTURE is hard to see. Take any stock market Pump and Dump (do they still happen? I was shorting stocks during dotcom days, on the sidelines now). Prices fall back to baseline after the dump and the latecomers realize what's happened. Rents have been basically flat.... indicating there no structural reason for housing prices to rise, but a tulipfest abetted by low interest rates.

[GenXer]

Mark: Are you looking for the original study? I do not think it is available online, but the links I provided above have references to this study. Please see earlier in the thread.

[admin]

[john p]

You know in business you've either got something, some skill or some good that people want or you don't.

I think the likelihood of sucess depends on the magnetism of the thing that's out there.

If you want to start a pizza place, you can run the numbers and assess the catchment area, the profile of the customers etc. What matters is that the pizza tastes good. I wish I lived closer to Bianchi's Pizza (Revere Beach). I've driven 50 something miles to get it, so it has a pretty strong gravitational force. I have a cousin in Newburyport that gets it as well...

Good Pizza...

[barry]

hell you have to buy now-- even in some of the not so immune towns home prices are still going up over 10% over the last 2 years... well at least according to this guy and the realtor... This is not even in an immune town...

29 Winter Street, Natick MA 01760
9 Rooms, 3 Beds, 2 Full and 1 Half Baths
List Price $494,500
Sold: 7/21/2006 for $447,000
MLS # 70870427
Living Area 2102 sqft
Lot Size 20,038 sqft
Taxes 5053.00
Year Built 1989


there are dozens of these out there so prices must be going up... or we have a long an painful road ahead before we can actually get some equilibrium in home prices