Even the experts missed the housing bubble

by Robert J. Shiller

Mar 05, 2008

http://www.ohio.com/editorial/commentary/16246947.html?page=all&c=y

This article starts out by summarizing how Alan Greenspan failed to predict the recent housing bubble, and how it is responsible for the collapsing of financial markets in the United States. The article explores how something like this could slip under the radar of the experts. The article raises the point that it is highly unlikely that everyone is making irrational decisions. Instead, Shiller makes reference to an article written in 1992 describing information cascades and how perfectly rational people make bad decisions because of the information presented to them by past decision makers. Shiller writes that information cascades are to blame for the housing bubble because rational investors were making good decisions that were based on poor or misleading information provided by previous investors. Finally Shiller writes how at this point a downward cascade will develop, in which investors are overly pessimistic and cautious.

This article is very relevant to our recent studies of information cascades and their presence in the world. In this case the result of an information cascade is the housing bubble, in which the pricing of houses rises above a sustainable level relative to income. This means that buyers have invested in a house that they will lose money on. Even experts such as Alan Greenspan failed to predict this housing bubble because they assumed that rational investors would be able to spot a poor investment before they buy. As the article points out, this was not the case because of information cascading. The investors were in fact rational and making the correct decisions based on the information provided. However, this information was misleading because of the nature of the information cascade. The article explains that there is a chance that a person will make the wrong decision given a signal. In this case it is possible that someone made an incorrect decision to make an initial investment. The next person to make a decision could have gotten a “low” signal but since the person before him chose to invest, they would be indifferent and could essentially flip a coin to decide whether or not to invest. If they arbitrarily chose to invest, then an information cascade begins. The next person, regardless of what signal they have, will choose to invest because the first two people invested, and he can only assume that it was because it was the correct thing to do,. In reality the first person was mistaken and the second worked off the first persons error.

As we have found out in class, information cascading is the explanation for the housing bubble situation described in this article. Although we have not yet discussed downward cascading in lecture, I believe the author of this article correctly predicts what will happen after an information cascade has been recognized. It makes sense that people will be very cautious in their investments after seeing what will happen if previous decisions are used as a primary factor in personal decision making.

An application of information cascading is groupthink.  Cohesive groups promote a series of ideas and socially ostracize those with opposing viewpoints.  Other members lose incentive to hold opposing views as this can lead to embarrassments if they are seen as foolish or against the group in some way.  Irving Janus classifies eight symptoms of groupthink:

•   illusions of invulnerability
• rationalizing warnings that challenge group assumptions
• unquestioned belief in the morality of the group, stereotyping those who oppose the group
• direct pressure to conform
• self censorship
• illusions of unanimity
• members who shield the group from dissenting information. 

 Classic examples do include the Bay of Pigs invasion and the Challenger crisis.  These turned out to be disasters in US policy and yet opposition was generally suppressed here.  Also the current day invasion of Iraq has some influences of Groupthink.  From this, it can be discerned that many historical movements such as Nazism, Facism and so forth gain their strength through aspects of groupthink.  Psychologists have offered solutions to vaccinate onself against groupthink: http://www.psysr.org/groupthinkwelcome.htm  Janus offers the following solutions: 

• leaders should make each group member to freely express objections and doubts
• higher ranked members should not express opinions when assigning tasks
• several independent groups should be set up by the same organization
• alternatives should be discussed, each members should discuss group ideas with trusted people
• outside experts should be brought into meetings
• one member should be a Devil’s advocate. 

Another video offers a safeguard against Groupthink:  http://www.crmlearning.com/groupthink-2nd-editionA further study of this can explain issues as far as adolescent peer pressure and perhaps “politician” peer pressure.  Perhaps better decisions can be made with a better study of Groupthink.   An article here (http://www.networkworld.com/community/?q=node/8836) gives groupthink another name- collective intelligence showing that “We are smarter than me”.  However while collective intelligence can give a greater generation of ideas it falls into a danger of being run by a small group.  Several current-day and historical examples point to this.  http://www.psysr.org/groupthinkwelcome.htm http://en.wikipedia.org/wiki/Groupthink 

Blog Post Number 2

Slap in the Facebook:

It’s Time for Social Networks to Open Up

Basically in this article, the author talks about how there are social networks all over the Internet that allow for the sharing of information from person to person or through various relationships. One of these websites that he cites is Facebook, which is a social network that allows for the passing of information through its network between members. The author found his problem with these social networks such as Facebook and MySpace when he noticed the information can only be passed through the members of the websites and the information is not accessible to each other. For example, a Facebook user cannot look up somebody’s profile or information on MySpace if they do not have an account and a user on MySpace cannot find out any information about a friend on Facebook if they don’t have a Facebook account. The author says that this is a problem because it does not create a free flow of information in the web and when you are putting information up that you want other people to see, you are creating a one-way street, in which only people in that network can see what you are doing/posting. The author poses an idea that there should be a social network that is universal throughout the web and allows you to access information no matter if you have an account or not. There are several websites that are being newly released that allow for this information to be passed through the Internet freely such as Pulse, which is released by Plaxo. The website will be limited at first until new ways are figured out to spread the features of these social networks across the Internet. The author also goes on to state that anybody can create a page similar to Facebook and make it better and more available to the public by using a variety of tools that are available to anybody. All in all, the author is asking for a website in which information can be passed freely through the internet and visible by people that want to search that information, no matter what website you may have or not have an account with or if you are just searching the web for an old friends information.

            The information is in a way linked to the new material we are covering in class about web page rankings and also the older material related to networks and friendships. In the beginning of the class, we learned about connections between nodes that formed a network and connections between nodes that connected two otherwise separated networks. In this article, the author talks about Facebook, which is a website that provides connections between people who belong to that network. These people or nodes can become friends with other people or nodes in the Facebook network by having similar friends or by just searching for people. However, this information is limited and the nodes or people in the Facebook network cannot link or become friends with anybody else in other networks such as MySpace. Therefore, separated networks cannot be connected because the information cannot be transferred through the social networks. In the new network that the author is talking about where there are no barriers placed on the network, information can be searched no matter if you are a member of the website, what website you are a member of, or if you are just searching information about a person. This means that these people or nodes can be connected no matter what through this network.

            Some newer material that we are learning now has to do with web page ranking and using hub pages to decide what are the best sources in searches. If this network could be made where the information that is put into this social network could be seen anywhere, then this website would be the highest ranked social network in front of other networks such as MySpace or Facebook because according to the author of the article, it would appear in all the search engines when you search for information on a person if they happened to post anything about themselves on any of these websites.

Although the VCG mechanism reproduces market-clearing prices in second-price auctions, it has several shortcomings that pose great threats to reliant sites such as eBay, Google, and Yahoo. Sellers sometimes use shill bids to artificially drive up the price of goods and provoke bidding wars among auction participants. This behavior is notoriously difficult for sites such as eBay to detect, especially when sellers collude. Kenneth Walton, author of FAKE: Forgery, Lies, and eBay, and his partners submitted shill bids in hundreds of eBay auctions before being convicted of fraud by the United States Attorney. Never the less, how should eBay combat shill bidding and preserve the integrity of its online marketplace?

eBay’s homegrown solution the shill bidding epidemic, the SMI (Safeguarding Member IDs) policy, seems to be doing more harm than good. Many members of eBay community are concerned that this anonymity may make shill bidding more difficult to detect. A recent article at AuctionByte.com believes that this policy may also discourage buyers from bidding on expensive items due to the lack of “transparency.” I stumbled across an article that I believe proposes the right solution.

The School of Electronics and Computer Science at the University of Southampton has proposed a new fee structure. eBay requires sellers to pay a fee for settings a minimum price on their goods; researchers at the ECS found that sellers often circumvent this fee by placing a shill bid instead. Under their proposed fee structure, sellers would be charged a percentage of the difference between the set minimum price and the final selling price. They believe this will simultaneously discourage shill bidding and encourage the setting of minimum prices. I believe this non-legislative solution could be highly effective at reducing the incidence of shill bidding in online auctions.

Lately, many observers and pundits are glued to the rear-view mirror, fixated on the implosion of residential real estate and excessive leverage, the last two market bubbles, wondering how in the world we could not have seen them developing. At the same time, investors are dumping financial assets they never understood in the first place – often in fire sales, regardless of the assets’ fundamentals – in fear of further market declines. Yet has anyone taken a step back and looked at where these tremendous money flows are ending up?

This article from today’s New York Times titled “Odd Crop Prices Defy Economics” describes an incipient bubble in the world’s financial markets:

http://www.nytimes.com/2008/03/28/business/28commodities.html?ei=5087&em=&en=0ec016a630af0065&ex=1206849600&adxnnl=1&adxnnlx=1206736516-AUVslv4DPIFWuHqx5aoOrw

Money is flowing rapidly out of many stocks and bonds, and into futures contracts on primary commodities such as gold, crude oil, and agricultural products like wheat and corn. The rationale is that these so-called “hard assets” are positively correlated with inflation and viewed as a safe haven for one’s wealth during a time of a deteriorating U.S. currency and volatile financial markets.

Granted, there is very compelling evidence for a continued long-term bull market in many of these commodities. But as the article makes clear, several of them are becoming dislodged from their underlying fundamental values, especially since the credit crisis began.

The answer to the question about why futures prices are expiring above cash prices is clear: investors are disregarding market fundamentals. This happens at the beginning and throughout every financial bubble.

In terms of an information cascade, this passage is telling: “The market sends a sell signal, but they don’t sell,” said Kendell W. Keith, president of the National Grain and Feed Association. “So the markets are not behaving the way they otherwise would – and the pricing formula for the industry is a lot fuzzier and a lot less efficient than we’ve ever seen.”

This is because the demand for commodity futures contracts, a liquid financial asset, is at an all-time high; and being driven by hedge funds, exchange-traded funds, sovereign wealth funds and any other esoteric fund that you could come up with. But on what information are these various actors basing their decisions? How many people really know the economic fundamentals of soybeans, palm oil, palladium and pork bellies?

Yet these arcane assets are where an unprecedented quantity of funds is flowing into. As for why arbitrageurs don’t exploit such market inefficiencies, here is an excellent article on financial bubbles by economist Alan Krueger:

http://www.iht.com/articles/2005/04/28/business/bubble.php

It is important to note that the commodities market has outperformed the stock market for nearly a decade. And sophisticated investors, such as Jim Rogers, co-founder of the Quantum Fund with George Soros, have been raking in massive profits from commodities this entire time. Warren Buffett was betting on higher energy and silver prices ten years ago, back when less-enlightened investors were dazzled by dot-com stocks. And throughout the mid-2000s, a significant portion of hedge fund and Wall Street trading profits were driven by the commodities boom.

But now, with surging prices as front-page news, those same follower speculators from the tech-stock days are piling into commodities. Their signal is not a compelling fundamental value relative to price, but an increasing price regardless of fundamental value.

What the wise do in the beginning, fools do in the end. And as we know well, financial bubbles do not end nicely.

CNN News: Southwest grounds 44 planes

One year ago, Federal Aviation Administration (FAA) found that Southwest Airlines failed to meet inspection requirements. More recently, the airlines put three employees on administrative leave and began conducting an internal investigation of allegations that it flew aircrafts without proper inspections. This triggered curiosity in the House Transportation and Infrastructure Committee and investigation of Southwest Airlines immediately commenced. Furthermore, the FAA is seeking a $10.2 million civil penalty against the airline which allegedly flew some airplanes for 30 months past government inspection deadlines. On March 12, 44 planes were pulled from their usual route and inspected, causing cancellations of about 4% of the day’s flights. Questions remain as to why an airline would put so many passengers in danger and why FAA managers, who knew about the lapse in safety at Southwest, would decide to allow the airline to conduct the safety checks on a slower schedule. However, the answer seems to lead back to the fact that taking aircraft out of service would disrupt Southwest Airlines’ flight schedule.

In class, we have been talking about information cascades, a situation in which every person makes a choice based on his own private signal (High or Low) and signals from others. Every person’s decision is expected to be rational; however, even if all actors have correct information, the first two actors may have high signals for the wrong information and cause an information cascade towards the incorrect choice. This in effect, is a similar concept to the psychological concept of “groupthink.” In groupthink, members of a group try to minimize conflict by reaching consensus, often without critically analyzing and testing their ideas. People may feel intimidated to share their thoughts or may just want to “get the job done”- often resulting in hazardous decision making.

FAA’s action involving Southwest Airline’s breach of security is a perfect example of an information cascade that turned into groupthink and eventually ended up with very dangerous consequences.

Flight cancellations are an airlines biggest fear. Each cancellation means angry fliers, overwhelmed staff, and of course, money. It is therefore no surprise that Southwest Airlines would forgo flight inspections to avoid flight cancellations and delays. There is no doubt that at the headquarters there was a struggle between meeting inspection needs and keeping up with business needs. Three employees have been put on administrative leave as a result of this struggle. However, psychology and groupthink show that whether or not these employees actually agreed with the decision to keep flying the planes, in the interest of minimizing conflict, they would feel that the best decision is to allow the aircrafts to fly and, each of them would avoid speaking up, even in the interest of safety. This is where information cascade comes in. As each employee begins to agree with the idea to let the airlines keep flying the planes, by the third employee, an information cascade would begin in which all employees would agree with the idea. In effect, all planes could be flown without any employee ever saying anything – and a simple psychological trend could cause a cascade of events that jeopardize the safety of thousands of airplane passengers everyday.

In class we’ve talked about information cascade. One of the aspects is that when the majority of participates chose to accept/reject, it’s rational for you to mimic what they did no matter what signal you’ve got. The reason is that if more people chose to accept/reject, it is more probable that accept/reject is the right thing to do; in other words, majority rules. However, one question arises, any rational people would ask: “Is following what other people have done without thinking themselves truly rational?”

In the article “Herd Behavior and Public Information” by Pierre Lemieux, the writer talked about information cascades and the different variations. He specifically talked about the wrong cascades and their inefficiency. “Lots of people can be wrong for a long time.” Some of the examples include global environmental scares, jihad against smokers, etc. If everybody doesn’t think themselves and just follows what other people do, it’ll do a lot of bad things to the society.

I claim that wrong cascades are the ultimate causes for economical bubbles in the history.  For example, the housing bubble recently. All Americans believed that houses prices would forever rise, because of information cascade. More and more people could not enter the market, they took more and more risks. I believe some people noticed the already too-high prices of houses, but they, according to the “rational” solution in information cascades, followed what other people did.

The good thing is that wrong cascades will be corrected at some point. “To reverse wrong cascades, credible individuals must obtain, and act on, correct private information.” says Pierre, “There are good reasons to believe that wrong cascades, even supported by special interests, can be reversed by free speech, individual liberty, and the dispersion of power in society. Individuals and groups who defend such ideals are contributing to an efficient economy.”

Warren Buffet has written a great deal on value investing, a technique to investing that he has used to consistently outperform the S&P 500 over the past 20 years.  A short summary of his work can be found here  .  Value investing is not a new concept, it was most famously written about by Ben Graham in his investing classic, the Intelligent Investor, and it consists of a process of buying stock in solid companies that can be bought at a discount to their intrinsic value.  A solid company is one that has paid consistent dividends, has had great returns on capital, and has produced significant amounts of cash profit as a percentage of revenue. While this may not sound revolutionary, its implications are outstanding.  Firstly, value investing does not concern itself at all with what other investors think about a stock, it only instructs the investor to buy based upon a detailed assessment of the company.  Some of Buffet’s best performing stocks, like American Express, were sluggish to rise in price because it took a while for other investors to realize that a mainstream, often overlooked company can be a great company to own.

Another way of viewing value investing, and stock picking in general, is through an information cascade.  A simple model for the value of a stock might be what an investor thinks it is worth, through financial analysis, plus its popularity.  News shows, newspapers, and news reporters make their living telling investors which stocks are “hot” or are popular at a particular time.  Most investors will make no money over time by picking “hot” stocks.  The biggest flaw in investing based on popularity, or what other people are doing, is that it is impossible to predict the popularity of a stock in the future.  Popularity is not quantifiable, it is speculative.  Investing by following trends has been proven to be at the very least market performing, and as lethal as being caught in a bubble, like the 1990’s tech bubble.  However, value investing disregards what other investors think of a particular stock, instead postulating that the true value of a stock will be revealed in its price over time.  One chapter in Graham’s book, the Intelligent Investor, speaks about the need to control the natural urge to invest in what is popular.  For example, a profitable company, like Caterpillar, which might not have been as popular to invest in during the tech boom, turned out to far surpass most tech stocks in the following years because it was a solid company, making solid earnings, and its intrinsic value was realized in its price over the following few years. 

Most of us have a list of friends at that numbers in the hundreds or even thousands, but I’m sure that most of us only consider some proper subset of them friends. Whether the complement of that set is the remnant of people who randomly friended you before coming to Cornell or just people you don’t talk to so often, the fact is that not all of our friendships have quite the same meaning.

Of course, we know that we can’t label our ties with a W or a S, like we did in approximating the strength of links in certain sorts of social networks early in the term. This sort of rudimentary labeling was surely useful in formalizing the intuition of the difference between a friend and an acquaintance for the purposes of an introductory class about networks. However, I think it would be pretty uncontroversial to assert that the Facebook’s lack of explicit edge-labeling is a very good thing. Surely it would be a nightmare to try to group our friendships into categories. It would probably introduce all sorts of other conflicts as well–I’d probably be at least slightly bothered in cases where the person with whom I shared an edge disagreed about the character of that tie.

With that said, logging into The Facebook today, I noticed a new feature: based on your existing friendships, the service will enumerate people (who aren’t yet your friends) who are likely to be your friends. I laughed when I saw it, because I think this is clearly a no brainer that should have been done at least two years ago. Check out a screen shot:

Surely, this feature is intended to capture the same sort of intuition as we did in our discussion of triadic closure that we discussed earlier in the course: if a person x shares a friendship with a person y and a person z, then it is likely that y and z are friends. Of course, this generalizes it somewhat in some ways that might be unknown to us and are likely very interesting.

But beyond that, I think there is an even more interesting observation to be made by thinking about the application’s suggestion to use the feature so that the results can become better for you. Clearly the clever engineers at the Facebook have some cute algorithmic tricks up their sleeves here. If it is indeed possible to improve your results by offering feedback, then perhaps the application is capable of learning which sorts of ties are stronger than others.

Suppose that you are friends, on the Facebook, with A, B, C, and D. Suppose also that you are friends (but not on the Facebook) with E as well. Suppose A and B are Facebook friends with E, and suppose that C and D are Facebook friends with F, who you might know but wouldn’t consider a friend. Now, suppose that the Facebook suggests that you might be friends with E based on your friendship with A and B and that you might be friends with F based on your friendship with C and D. Naturally, you might add E as a friend based on this suggestion, but not F. This gives reason to believe, perhaps, that A and B are closer friends to you than C and D are.

So imagine that it does really turn out to be the case that the Facebook can gather enough data from this feature such that it can apply fancy algorithms to try to approximate the strength of ties. Yeah, there are a lot of ifs buried in there. But seriously: sweet.

Is the magic of Google’s ad network wearing off?

This blog post from BloggingStocks raises that exact question. Google built its empire on the idea of contextual, text-based ads that are relevant to the content that the reader is interested in. This simple system, described fully in chapter 11 of the test, allowed the search company to go from a PhD project at Stanford, to a $100 IPO, all the way to a stock price high of over $700.However, Google – along with a good portion of the market – has fallen on hard times recently, with its stock price falling to under $500. (As of the market close on Thursday, Google was at $444.08.) Conventional reasoning writes this loss off as a consequence of the economic difficulties the market is facing right now. However, what if there is a deeper problem; what if people are starting to wise up to Google’s method of capturing attention?

In class, we have learned that Google’s system is at least partly based on the generalized second price auction model. The whole system, however, is predicated on the assumption of independent, private values. If those values are indeed being influenced by a perceived devaluation in a placement on Google AdSense, then Google may be in a difficult position; if I don’t have confidence that I will see a return in my advertising investment, then I will be unlikely to make that investment in the first place. The billion-dollar question is going to be whether this represents a temporary blip or a permanent downswing.