bostonbubble.com

[balor123]

http://www.khanacademy.org/

You all seen this site? Supposedly Bill Gates is a big fan of it. Lots of housing and finance tutorials here.

[Greater Bos]

Cool site. Very useful.

[balor123]

I feel like posting a link to the housing conundrum video on every site that mentions that it's good news that housing prices are rising. It's like it's 2005 again and rising is simply a good thing, completely ignoring a whole bunch of other factors like supply vs demand. Media presents a very shallow analysis but maybe that's also what people want to see.

[GenXer]

Bubbles are actually very complex. The factors that give rise to bubbles can be different, but the mechanism is always the same. Its called 'multiplicative cascade'. This mechanism works in systems with multiple complex feedback loops which are able to give rise to extreme types of probability distributions (i.e. the resulting gains and losses can be approximated by a certain class of fat-tailed distributions). Thus, by definition, bubbles are very far from equilibrium, hence, there are very large imbalances that are generated which have no relation to the so-called fundamentals. We can retrofit any bubble with an explanation after the fact, but complex feedback loops are weighted quite randomly (that is, the causes are deterministic, but the interplay between the variables can appear random because of the complex feedback loops), resulting in very chaotic behavior. Pretty much in a nutshell, systems exhibiting 'multiplicative cascading' properties can align (correlate) in such a way as to produce very sharp jumps in prices, which are usually deemed extremely rare, yet the systems with these properties are able to generate such outliers quite often, a lot more often than fundamentals would suggest. Thus, we can not predict bubbles because we have no way to map or assign weights to all of the available feedback loops. Often the government can skew the output of such a system because of unintended consequences of its meddling. There is research being done to try to see if bubbles can be predicted numerically by looking at the charts (basically, if the charts rise too fast over a certain short period of time - we got a bubble on our hands), but otherwise, aside from 'broken clock' approach or simply somebody making a lucky guess (fundamentals can always be interpreted either way, for or against), we have no way of preventing bubbles (predicting vs. preventing is a big deal). There are many reasons for that - some of them have to do with the availability of easy money (this may actually be the primary cause of the last real estate bubble), but there are many more complex causes (i.e. the dot-bomb growth generated easy money that went into real estate when the interest rates were still relatively high), yet those may or may not have pushed us over the hill towards a real big bubble (easing of the rates in 2002 may have started us on a cascading course, though).

To sum up, while all bubbles are mostly unique (with some misleading commonalities) in how they are created and how they unravel, all bubbles have several common aspects to them, having to do with the emergent properties of the markets, such as the ability of the markets to cascade multiplicatively, or to align in such a way as to cause a very large price jump or fall (which can be described by power law distributions - the same distributions that govern earthquakes).

[balor123]

I don't fully agree. I think we can never fully prevent bubbles but you can hit the common ones at least, leaving you with the difficult to recognize problems (ie idiopathic). My opinion is the best solution to this problem is to try and contain the ones you know about and limit the impact of those you can't identify. The way you do that is with a high cost of borrowing. The chief problem with that approach that I hear is that then there is no growth. You can make up for that with strategic subsidies paid for by taxes and savings. Even if I could get everyone on board with this philosophy, we still have this huge gaping hole that we've been digging for 30 years to deal with. I don't have a good solution to that except tough love or default. I also don't know if a democracy like ours could resist the temptation not to game those rules, as the potential will always exist to reverse or circumvent them.

[GenXer]

Thank you for making my point for me. The only way to prevent bubbles is to not to engage in the behavior that leads to bubbles. But like I said before, easy money does it, and there is nothing that can be done about it. I mean, it CAN be done, but it WON'T be. Besides, a free market (or even reasonably regulated one) will give rise to bubbles, and we can not stop them from happening. To do that we'd actually need to go forward in time, as not every bubble that's forming is going to become a crisis, and some that become a crisis are not deemed a problem simply because they take so long to form and they become part of the economy.