Sunday, April 21, 2013

Speculative Bubbles (Part 2 of 2)

In Part 1 I talked about one of the "bubbles" David Stockman talks about in interviews about his new book.  The whole post can be found at http://sigma5.blogspot.com/2013/04/speculative-bubbles-part-1-of-2.html.  In this second post I will talk about the other two items in Stockman's "bubble" list.  The first one is the mortgage bubble.

My father was born in 1910.  The house he grew up in was built in the same year, 1910.  About 25 years later his father died at the height of the Great Depression and my father became responsible for the family finances.  The house had originally cost $5,000.  In the middle of the Great Depression it was again valued at $5,000.  And it turned out that the mortgage on the house was an "interest only" mortgage.  There was still $5,000 owing on the mortgage.  Fortunately, my father had a good job and no wife and kids.  He was able to get the house completely paid off within a couple of years and the family lived in the house for many more years.

But a lot of depression era families did not do so well.  They end up having similar mortgage situations but did not have the financial wherewithal to weather the storm.  So they ended up losing their houses.  This caused the Roosevelt administration to put in rules requiring 20% down for a conventional mortgage.  This rule resulted in a stable but boring mortgage market for more than 50 years.  It also meant that a lot of people couldn't qualify.  Do gooders, who wanted to get more people into home ownership, combined with mortgage industry lobbyists, who wanted to do more business. The result was a loosening of mortgage regulations in several stages.  The anti-regulation crowd also chipped in to weaken the powers of the agencies that oversaw the regulated part of the market.  They also made sure that efforts to extend regulation and supervision to the (by the '90s) fast growing unregulated segment of the market were thwarted.  The result was a "wild west" mortgage market where "anything goes".

I paid 20% down when I bought my house in the early '80s.  That later became a thing of the past.  The "minimum down" was reduced to 10%, 5%, 3%, and eventually in some cases to "nothing down".  And the mortgage market has always been segmented.  Besides regular mortgages there are FHA mortgages that have slightly different rules but you had to meet certain criteria to qualify.  Then there was the "jumbo" market.  The federal agencies that were market makers for mortgages FannieMae (actually the FNMA - you don't want to know what the acronym stands for) and FreddieMac (again an acronym for who cares what) would only take mortgages up to a certain size (I don't know why).  Mortgages that exceeded the limit were called Jumbos.  This all made some small kind of sense.

Then there were Alt-A mortgages.  These were mortgages to people who did not have the financial resources (down payment, income, credit history) to qualify for one of the more standard kinds of mortgages.  These mortgages were viewed as substantially more risky than the other kinds.  As a result the interest rate charged the borrower was substantially higher than for the other kinds of mortgages.  For a long time the Alt-A mortgage market was very small.  Most lenders concluded that the reward (higher interest rate) did not justify the additional risk (that the buyer would default on the mortgage) so for a long time very few Alt-A mortgages were made.

There was a slow broad societal change that happened in the second half of the twentieth century.  The number of people that lived in the typical house went down.  People had smaller families.  Older people stayed in their houses after the kids moved out.  Single people bought houses.  All these trends combined to reduce the average number of people living in a typical house.  What this meant was that more houses needed to be built than population growth would indicate.  So the housing industry saw prolonged steady growth.  This growth tended to keep the price of houses higher than they otherwise would have been.  For a long period of time the value of a house ticked up regularly as clockwork year after year.  Initially it was a small increase but it was steady.

Then speculators started moving in.  It looked like a safe investment to buy a house and then rent it out.  The worst thing that could happen is that you would be unsuccessful at finding renters.  In this case you just sold the house and perhaps took a small loss.  In other words, investing in houses looked like it was low risk.  So a lot of people did it and this too propped up housing prices.

Theoretically the risk of default was rising as the size of the down payment dropped.  But things did not turn out that way in practice.  If someone found that they could not afford their house they could always sell it.  The proceeds would cover the mortgage and there would be no default.  This even worked if the down payment was pretty small.  The increase in the value of the house plus the amount of the small down payment would almost always be enough to cover the remaining balance on the mortgage.  As a result perceptions in the mortgage industry changed.  "Risky" mortgages were no longer very risky.  This meant that Alt-A mortgages were a good deal for the investor.  The actual higher return from the high interest rate more than balanced out the now small risk of losing money in case of default.  This made Alt-A mortgages actually preferable to a regular mortgage.  Both mortgages were low risk but the Alt-A mortgage had a much higher rate of return.

So all of a sudden Alt-A mortgages are a desirable investment.  This was coupled with another trend.  This trend was internal to Wall Street.  For decades Wall Street made a nice living on stock transactions.  The fee for a particular transaction was fixed.  As computerization took over the cost of executing the transaction plummeted.  This nice racket came to an abrupt end when the fixed commission system was scrapped.  Now commission rates were negotiated.  And the large players got very good deals.  Smaller players did not do so well but fees became common knowledge and small investors shopped around.  It didn't take long for the profit in handling stock trades to dry up and blow away for all but the most efficient operators.

A second similar blow was struck a few years ago.  Most stock prices were quoted in dollars and eighths (12 and 1/2 cents).  So prices had to be rounded to the nearest eighth.  And there is the "bid" and "asked" prices.  They always differ by at least an eighth.  This allowed for profit to be made on the eighth point "spread".  It wasn't much but it added up.  Then the exchanges went to dollars and cents pricing.  This dropped the spread to a penny, less than a tenth of what it had been.  This took more money out of the "profit" column on Wall Street.  All this made Wall Street very greedy for profitable businesses.  And Alt-A mortgages seemed made to order.  Assuming that the "low risk" assessment was correct, then the high interest rate made for a compelling argument.  Wall Street's appetite for Alt-A mortgages became voracious.

The result was a giant unregulated mortgage industry specializing in Alt-A mortgages.  And did I mention that this industry was unregulated?  Fraud soon became rampant.  Many people who could qualify for a standard mortgage were pushed into an Alt-A mortgage.  Why?  Because Wall Street wanted Alt-A mortgages.  And Wall Street wanted the worst Alt-A mortgages that could be created.  Why?  The worse the mortgage, the higher the interest rate.  And they were no risk (well technically low risk but lets not sweat the details).  It didn't matter whether the borrower could afford the mortgage because it would not default.  The worst that would happen is that it would need to be "rolled over" into a new mortgage.

In this market housing prices started rising very fast, sometimes by as much as 10% per year in some markets.  This made the "there's no such thing as a bad Alt-A mortgage" argument just that much stronger.  And the best way to insure that a large stream of bad Alt-A mortgages was available was to deal with crooks.  Now it's theoretically bad to deal with crooks.  But Wall Street's solution to that problem was to erect tall strong barriers to block them from officially hearing anything about crookedness or dirty dealing.

This "tall strong barrier" strategy worked so well that large Wall Street firms bought several of the largest crooked mortgage originators.  And they paid top dollar for them.  The result has been write downs of many billions of dollars as a result of these investments.  And this is after arranging "settlements" in which no wrongdoing is admitted but fines, in some cases amounting to billions of dollars, are paid.  But the fines represent pennies on the dollar for the fraudulent activity covered by the "settlement".  The level of fraud involved in the mortgage industry, especially in 2000 and later, causes me to put the mortgage bubble into the "artificial" or fraud-based category of bubble.  The mortgage bubble and the financial meltdown bubble are closely tied up with each other.  So I am arbitrarily cut my discussion of the mortgage bubble off here and move on to the financial crisis bubble now.

I have already covered some of the relevant territory.  As noted above, Wall Street has become increasingly desperate to find lines of business that are both high volume and offer big profit margins.  As an industry it has become so large that it now needs a lot of business and the business must be very profitable. Otherwise it can't afford to pay bonuses, which in some cases amount to over a billion dollars to one individual for one year's work.  Making bonus pools that run to tens of billions of dollars per year for a single company look like just some small overhead cost requires that cash flow amount to literally trillions of dollars.  There just aren't that many markets that are big enough to make a difference.

Also, as indicated above, this led Wall Street to the Alt-A mortgage business.  A Wall Street firm would buy up a bundle of mortgages (frequently a thousand or more) and bundle them up into a large investment called an MBS (Mortgage Backed Security).  They would then split the MBS into shares and sell them to investors.  This was a nice little business.  Each MBS was like an IPO so various fees and surcharges could be assessed.  And they sold themselves because they promised a high interest rate.  But it was not enough.  There just wasn't enough volume in MBSs, like there wasn't enough volume in IPOs to generate the profit Wall Street firms now felt they needed.  And, because it will set things up for later, let me now turn to a small problem Wall Street saw.

Lots of investors are "blue chip" investors.  They want (or are required by law) very low risk.  The lowest risk investment is one that is rated AAA (usually pronounced triple-A).  The AAA comes from a "ratings agency".  The big three are Standard & Poor's, Moody's, and Fitch.  They use different codes but "AAA" is usually short hand for a top (most safe) rating from one of the "big three".  An investment in IBM stock is not an AAA rated investment.  Stocks, even the best ones, go up and down.  They involve risk.  An IBM bond, on the other hand, might be AAA rated.  This is because if there is a problem the company is required by law to pay off bond holders before stock holders and many other kinds of creditors.  IBM AAA bonds are "senior debt" because the call the bond holder has on the assets of the company is "senior" (comes first) to the call others have on the assets of the same company.  Most bonds, even those of big stable companies are not AAA.  They are some other lower but "investment grade" rating like AA.  Getting an investment rated AAA is a big deal that historically hasn't happened that frequently.

And that's the problem with an MBS.  It's not AAA.  But wouldn't it be nice if it was AAA?  Then the seller of the MBS could market it to all those finicky investors that want or need AAA securities.  Wall Street figured out a way to turn an MBS into an AAA security.  The vehicle for this is something called a CDO (Collateralized Debt Obligation). What's a CDO?  Well it is almost anything.  It's an investment you can sell like an IPO or an MBS.  But what's in it?  In the case of an IPO what's in it is the stock of a specific new company.  In the case of an MBS it's a specific list of mortgages.  The details are different but the general idea is the same with an IPO and an MBS.  The value of your investment rises or falls based on how well the company (IPO) or package of mortgages averaged together (MBS) does.  In the case of the CDO what's inside is any kind of "Debt Obligation" that is "Collateralized".  It no longer has to be a specific class of thing like a mortgage or a stock.

There has to be something deep down there that can serve as collateral and the something must be some kind of debt.   But after that, it can be pretty much anything.  It can be an MBS (or part of one), for instance.  It can be a bundle of credit card receipts or car loans or other things.  You just have to specify what the "debt obligations" are.  You can even have a CDO inside a CDO (nickname:  CDO squared).  So the first step is to decide what to put into a CDO.  And here is where something called "tranches" come in.  You take one of these investments and slice it up.  Each slice is called a tranche.  The trick is in the slicing rule.  You don't just take the first hundred of the thousand mortgages in a MBS and call it "tranche 1".  You use a rule to rate each mortgage.  Then you use the rating to determine the tranche.  Why do this?  We'll see in a minute.

Anyhow Wall Street took a bunch of say MBSs and then tranched them.  The put all of the "tranche 1" components from all of the MBSs into the first CDO.  Then they put all the "tranche 2" components into the second CDO, and so on.  Then they sold each CDO separately.  Why?  The ratings agencies.  Say what?  Remember a long time ago we were talking about AAA securities.  All this was done to create securities that the ratings agencies would rate AAA.  In actuality, as I understand it, things were typically put into only three tranches.  And the "tranche rule" for MBS securities is something like "the first 10% of the mortgages that default" goes into the "junk" tranche.  The "mezzanine" tranche consists of the next 10% of mortgages that default.  The "senior" tranche consists of all the rest of the mortgages.

Now look at what we have.  How likely is it that 20% of the mortgages are going to default?  If it's before 2006 the answer is "indistinguishable from zero".  And that's the definition of a AAA security.  So when I hire a ratings agency they will give the senior tranche (actually the CDO containing only senior tranches) a AAA rating.  I don't know the details but the interest income from the MBSs was no doubt divided up in such a way that the investors in the "junk" CDO got paid a big premium and the investors in the mezzanine CDO got a smaller premium.  The interest rate paid by the senior CDO would be a little smaller but what do you expect for a CDO rated AAA.  This trick allowed Wall Street to turn 80% or more of each MBS into AAA securities.  And these AAA securities paid a much higher interest rate than a regular AAA security.  So they flew off the shelf and Wall Street racked up monstrous profits.

Before moving on, remember I mentioned something called a CDO-squared.  What's with that?  Well let's build a CDO by consolidating a bunch of junk CDOs together.  Now given that the underlying default rate on mortgages is roughly zero, let's play the tranche game again.  What is the chance that say 20% of this combination of risky CDOs are going to go under?  It's pretty small.  (Remember this is before 2006).  So using this trick you can turn 80% of the worst 10% into AAA securities.  And you can play the same game with the mezzanine CDOs too.  So you  can turn close to 100% of these CDOs based on MBSs into AAA securities.  It's just like magic.  And like magic we eventually found out it was not real.

So let me move on to the ratings agencies.  It's a cut throat business.  And guess who selects the ratings agency.  It turns out it is the company that is putting together the security (MBS or CDO).  So let's say you put together one of these things and you take it to a ratings agency.  And let's say they say "well it's not quite AAA".  One option you have is to take the exact same package to another ratings agency and they might give it their AAA stamp of approval.  The ratings agencies are beholden to the Wall Street firms for their business.  And they know that if they are too hard nosed the business will all go to the other guy.  So the ratings agency is under tremendous pressure to rate this stuff AAA.  So they did.  But they needed a fig leaf.  And Wall Street found them one.

Let's say you were the poor soul that was supposed to rate one of these CDOs.  They are terribly complicated so going through everything carefully would take a couple of lifetimes.  That's one approach.  Now what if there existed a computer program that was certified to be wonderful.  And it would go through all that detail at warp speed and issue a single magic "risk number".  Problem solved.  A math whiz came up with just such a computer program.  It was thoroughly reviewed by other math whizzes who swear it does a good job.  What can possibly go wrong?

Most of the analysis of this issue has said "well all computer programs have shortcomings and this one's shortcomings only became apparent after the market crashed".  There is some truth to this but it misses the main problem.  The guy that developed the program was on a limited budget.  His program needed two kinds of data.  The first kind was the specifics of the securities to be analyzed.  There was no problem here.  But the second kind of data he needed was historical data.  He needed a bunch of mortgages and what eventually happened to them (e.g. defaulted or didn't).  He pulled the historical data he could easily get his hands on.  It covered not that many mortgages and it only covered a short period of time, a period of time when no one was defaulting.  He explained all this in the write-up on his program.  And it didn't make any difference to the mathematicians who reviewed his work. They had enough information to determine whether the program was doing the right thing with the data it had and that's all they needed.

But things changed (or should have) when this program went to Wall Street.  The first thing that happened was the whole "here's the limitations of the program" part got dropped.  And Wall Street never went back and added more historical data that covered a longer period, a period when mortgages did actually default.  And Wall Street plugged everything into the program without generating historical data that was matched to what was being evaluated.  So the program was used to rate credit card debt and car loans and pretty much everything under the sun.  And the "historical data" was always the original mortgage data whether that was the proper historical data or not.  Whatever inadequacies the program might have had were overwhelmed by the failure to provide appropriate historical data.  The result is not surprising.  Given that all the historical data showed low risk the program always spit out a "low risk" number no matter what security data you put in.  And that suited the sales people responsible for selling the CDOs and the ratings people at the ratings agencies right down to the ground.  They had the "magic computer blessing" on what they already wanted to do.  If something went wrong they could blame the computer and they eventually did.

So how much of the financial bubble was natural and how much of it was artificial?  Most of the damage that resulted in the financial meltdown can be tied to the mortgage meltdown.  I have noted above that Wall Street actively encouraged fraud and other illegal behavior in order to have more bad (good in Wall Street's eyes because they had high interest rates) Alt-A mortgages to work with.  The ratings agencies also were utterly derelict in their duties to appropriately rate these complex securities.  They made no effort to develop appropriate in house expertise with mortgages.  Had they it would have been manifestly obvious that many CDOs should have been rated "beneath junk" rather than AAA.  They also did not make any attempt to understand either the ratings program itself or the proper use of the ratings program (e.g. verifying that the historical data was appropriate and adequate).  Above and beyond all that they should have known that there just isn't that much AAA paper around.

There is ample evidence that Wall Street drank their own cool aid.  They were under no pressure to purchase the fraudulent mortgage originators.  They should have but obviously didn't know that these businesses were fraudulent.  Wall Street firms, some more than others, also ended up purchasing CDOs and other garbage investments for their own portfolios.  Some of this was no doubt done "in the course of business" rather than solely as a "good investment".  But they held so much of these investments and were so slow to get rid of them that they obviously thought they were sound.  The people who made these kinds of decisions were paid obscene amounts of money.  Why?  Because they were so smart.

Since Wall Street crashed many of these very same "smart" people have frequently said something along the lines of "it was so complicated I really didn't understand what was going on".  An executive that is not smart enough to understand the business line that he is overseeing should be preplaced with someone who can understand it.  A business activity that is so complicated that no executive can be found that does understand it is a good business activity to exit.  Executives overseeing activities they don't understand are violating their fiduciary duty.  Executives that employ these under qualified executives are also violating their fiduciary duty.  They should all be fired.  They certainly are not justifying their obscene compensation packages.

Given this pattern of behavior that extended up and down Wall Street firms and extended across the board to all Wall Street firms, again I go with "artificial".  So the final score is one natural bubble (dotcom) and two artificial bubbles (mortgage and financial meltdown).  Why does it matter?  Well, the "fix" for a natural bubble is different than a fix for an artificial bubble.  I don't know that there is much that can be done to prevent natural bubbles.  The general fix for artificial bubbles is "law and order".  Make sure your laws and regulatory regimes are in good working order.  Then throw the crooks in jail and throw away the key.

Stockman is right to the extent that he faults the Fed for doing a poor regulatory job.  (I haven't read the book so I don't know if he actually holds this position.)  For the rest, congress deserves a great deal of responsibility for not providing the appropriate regulatory environment (the unregulated mortgage industry as an example).  Congress also deserves blame for the long term attack on efforts by regulatory agencies to do their jobs.  Certainly the ratings agencies utterly failed to do their jobs.  But at the center is a Wall Street culture of "anything to make a buck".  This resulted in a complete lack of appropriate supervision (as long as the department is making lots of money it's all good) and aiding and abetting efforts through congress to interfere with the ability of regulatory agencies to do their job.  "Follow the money" is usually good advice.  The people who most benefitted from all the bad behavior that transpired in the mortgage industry and on Wall Street were the senior executives at the various Wall Street firms.                                

  

Speculative Bubbles (Part 1 of 2)

The impetus for this post came from David Stockman.  He has been making the rounds flogging his new book "The Great Deformation".  Stockman has always been good at self promotion.  So he has propelled his book, ostensibly a dry economics tome, to number 7 on the New York Times Hardcover nonfiction list as of the day I write this.  Unfortunately, I am familiar with Mr. Stockman.  He first came to national prominence as director of the Office of Management and Budget under President Ronald Reagan.  The OMB is the font of expertise for all things budgetary in the White House.

Stockman was famous as a proponent of what came to be referred to as the "Laffer Curve".  The Curve was a shorthand reference to a concept pushed by economist Arthur Laffer.   The main idea was that if you reduced tax rates the economy would improve so much that tax revenues would actually go up.  Most economists dismissed the idea.  But it was politically convenient so it garnered a great deal of support among conservatives.  If correct, the concept would result in lower tax rates and a smaller deficit at the same time.  Unfortunately, supporters have only been able to find one instance where this actually happened.  In the Kennedy administration tax rates were cut and the economy took off.  This strong economic growth resulted in enough increased tax revenues to keep the deficit low.

Unfortunately, what happened during the Reagan years was what most people predicted.  After a recession in the first couple of years (caused by a change in Fed policy) the economy recovered but not enough to cover for large increase in defense spending (and fairly modest cuts to social programs).  Revenues eventually went up a little but spending went up a lot during the same period.  So the federal deficit ballooned.

After he left office Stockman claimed that he had become dubious of the whole "Laffer Curve" idea early on.  But he did not say anything negative publicly about Reagan fiscal policies until well after he left the administration.  Later criminal and civil charges were filed against him in connection with his actions while running Collins & Aikman, a company that went bankrupt.  I believe none of these charges led to a conviction.  In short, Stockman's record is spotty at best.

I have not read the book.  What I want to investigate is a contention he has made in several interviews he has given in the process of promoting the book.  He contends that there have been three recent bubbles:  the dotcom bubble, the mortgage bubble, and the financial meltdown bubble.  He goes on to blame the Fed for all three bubbles (and many other sins).  At this point you will be less than surprised to learn that many experts disagree with his analysis.  But I am just going to stick with the "three bubbles" part.  Are these three events in fact bubbles?  To answer that question it is important to understand what a bubble is.

Bubbles have been around for a long time.  One of the best works on the subject, "Extraordinary Popular Delusions and the Madness of Crowds" by Charles Mackay was first published in 1841.  It has been in print nearly continuously since.  Amazon has new copies of several versions of the book (one of which was published in 2012) for sale right now.  An early bubble described by Mackay was the market for Tulip bulbs in Holland in the 1630's.  Let me use this event to describe the characteristics of a bubble, or more properly a "speculative bubble".

At the time when our story starts, the Dutch has been selling Tulip bulbs for some time.  As they became popular an informal market developed.  For explanatory purposes I am going to arbitrarily divide market participants into "investors" and "speculators".  In my explanation an investor is someone who has a stake in the underlying product.  He grows Tulips, or sells Tulips at retail, or whatever.  He is primarily in the market because he has a commercial interest in Tulips.  A speculator, on the other hand, just wants to make a buck.  He has no intrinsic interest in Tulips.  They are just a means to an end.  Now in the real world the situation is much more complex.  There are other types of participants and a particular actor may be part investor and part speculator, even if we just look at a single transaction.  But this artificial separation makes it easier to see what is going on.

I will also assume something called the "intrinsic value" of the product, in this case Tulips, exists.  Determining intrinsic value is often somewhere between difficult and impossible.  And experts can and will differ on what the correct intrinsic value of a specific commodity at a specific time is.  But we are going to ignore all these problems and pretend there is an intrinsic value and that it is knowable.  Now let's look at markets.

An investor is in a market because in some sense he has to be.  There are many forces that push a price up or down.  So in a normal market we can expect that the price of a product will sometimes get above its intrinsic value.  In this case we can expect prices to eventually go down.  Similarly, the price can get below intrinsic value and in this case we can expect the price to eventually rise.  The fundamental law of speculation is "buy low - sell high".  So, if a speculator can establish the intrinsic value, and if he can identify times where the price is below intrinsic value, he can buy low and expect to sell high later.  Similarly, he gets out of the market when the price is above the intrinsic value and waits for the market to "return to the norm".

In this situation a speculator provides an economic benefit to the economy as a whole and specifically to investors by providing "market liquidity".  An investor may need to sell when prices are low.  The speculator will buy when no one else will.  Similarly an investor may need to buy when prices are high.  The speculator will sell when no one else will.  In these kinds of situations speculators perform the service of smoothing out the market, which is usually beneficial to investors in the long run.

So we have now seen on display two of the components necessary for a speculative bubble.  We have a market and we have speculators.  And in this benign situation we have described so far, speculators are performing a useful function.  But things can go wrong.  Let's see how.

Suppose we have a market that has performed well (and in this case by "performed well" I mean gone up) for a significant period of time.  For whatever reasons prices have risen steadily and apparently inexorably.  This sounds like an opportunity to print money.  We buy and hold.  The market inexorably goes up.  We sell at a profit.  Any fool can make money in this market.  This is the third component of a bubble.  It is usually but not always present.

People may mischaracterize the actual history of a market so that it seems like it always goes up.  And probability tells us that if we monitor a lot of markets, and each market has a certain amount of random price fluctuation, then eventually some of these markets will show a pattern where prices seem to always (well, for a "long" time, whatever that is) go up.  Now the speculators we talked about before had a level of expertise.  They were able to determine the intrinsic value of the product.  But in a foolproof market that expertise is no longer necessary.  The market always goes up.  And we now have the fourth component of our bubble, the "quick buck artists".  These are people who have little or no expertise in the particular market.

This set of circumstances moves us into the inflation phase.  Prices quickly and inexorably go up.  In our Tulip example this happened when French speculators started moving in in 1634.  Prices started accelerating and by 1636 they were zooming up.  Nothing about the underlying Tulip market had changed.  But now trading was dominated by speculators.  It is also when investors and the early more experienced speculators started becoming concerned that prices had gotten out of control.  Now we move into the "greater fool" phase.

Knowledgeable people conclude that prices are at foolish heights.  But it appears that there is always a greater fool who will pay even higher prices so they stay in the market well past when they normally would. Then we finally come to the last phase.  At some point the market runs out of greater fools.  At this point prices start to weaken and perhaps fall slightly.  Knowledgeable people now start to sell aggressively before prices fall even further.  The market crashes (i.e. the bubble bursts) and prices drop, frequently to well below intrinsic levels.

To summarize, we need a market and the market needs to be open to speculators.  We need (not always but usually) a period where the market goes up apparently inexorably.  Then more speculators jump in and prices start rising rapidly.  We then have the "greater fool" period, followed by the peak and then the crash. This pattern has been well known for hundreds of years now.  When the Tulip bubble happened no one had seen anything like it before.  But it certainly wasn't the last bubble.  Mackay documents a number of other bubbles like the "South Sea Bubble" that are now known only to historians of the subject. At some point it dawned on some speculators that it was possible to manufacture bubbles.

By the late 1800's "stock operators" were routinely trying to create and profit from bubbles in the price of various individual stocks.  The 1923 novel "Reminiscences of a Stock Operator" by Edwin Lefevre documents many of the popular schemes of the time.  I bring this up because I want to differentiate between the "natural" speculative bubble, which arises naturally and is then taken advantage of, and the "artificial" bubble, which is created as a scam from the start.  With all this out of the way let me now take up the three bubbles that attracted Mr. Stockman's attention.

The dotcom bubble is accurately characterized as a classic natural bubble.  Also referred to as the "tech bubble" it started gathering steam in the 1980's.  The Apple II (introduced in 1977) and shortly thereafter the IBM PC (introduced in 1981) made home computers practical.  Many people who had heretofore not thought of themselves as tech-savvy got computers and fell in love with them.  They imputed their feelings to the market as a whole and started looking for stocks they could invest in.  An early example was Netscape.  Netscape brought the browser to the masses.  When Netscape stock went on the market it was immediately snapped up to the surprise of Wall Street experts.

In the beginning Netscape had a reasonable business model.  They were going to license their product for $25 a pop.  Then Bill Gates came along and gave Internet Explorer away for free.  Netscape soon abandoned their licensing model.  This destroyed their business model but the public continued to support the stock anyhow.  Eventually, Netscape was sold to AOL in a deal initially valued at $4.2 billion (and later at much more).  So the faith of the general public in Netscape as an investment was eventually vindicated.  It would not be the first time.

Bill Gates was of course famous for creating Microsoft.  Microsoft also went off for crazy prices.  But, unlike Netscape, Gates turned out to be a savvy businessman.  Microsoft turned in spectacular earnings and profit figures year after year.  So Microsoft stock zoomed into the stratosphere.  Apple had a much bumpier ride, at least at the beginning.  The Apple II was a big success.  But follow on products (does anyone remember the Apple III) were less so.  After a string of duds Apple introduced the Macintosh in 1984.  But it too was not a commercial success.  It only garnered significant sales after Steve Jobs left the company.  Apple then got into so much trouble that Jobs was brought back to rescue the company.  The company was rescued and went on to a run of spectacular success.  But this mostly happened after the dotcom bubble exploded in 2000.

Microsoft stands out as the great success of the dotcom era.  Mostly we have companies with track records like Netscape (never made any real profit but was eventually sold for megabucks) and Apple (up then down and eventually back up but only after the bubble burst).  Nevertheless the public was fascinated with these stocks and would buy them at nearly any price.  This was in spite of the fact that in most cases no one could figure out how the companies would ever make money.

So in the early days you had a number of these new style tech stocks.  And many had a record of inexorably rising in spite of the fact that they never made any money and it didn't look like they would make money any time soon.  The most common tool Wall Street uses to calculate the intrinsic value of a stock is PE or the price/earnings ratio.  If the price of a stock is $20 and its earnings for the year are $1 per share its PE is 20.  Historically, a PE of 15 is considered standard.  Anything between 12 and 18 considered within the normal range.  There are many reasons why the PE of a stock should be below 12 or above 18 but those are special circumstances.  Generally, in the absence of special circumstances, a stock is considered cheap if the PE is below 12 and expensive if the PE is above 18.  Many of these tech stocks had no earnings, no "E".  So their PE was off the chart on the high side.

This might be ok if there was reason to believe that earnings would skyrocket in a couple of years.  If we were certain that three years from now the stock would be earning $5 per share then the "forward" PE would be 10 if the price of the stock is currently $50.  This forward PE of 10 might justify buying the stock now.  This kind of logic is risky but not insane.  The problem with many of these stocks is that no one could figure out how they could earn say $5 per share 3, 5, or even 15 years out.  So purchasing these stocks appeared to be completely insane to Wall Street analysts.  But the public bought them anyhow.  Enter the greater fool.

From start to finish the dotcom bubble took about 15 years.  That is an extremely long time period by Wall Street standards.  And Wall Street's approach to stocks that eventually became part of the tech bubble went through three stages.  Wall Street had been investing in "tech" for a long time.  GE dates back to the 1800's.  IBM was a Wall Street darling in the '60s and the '70s.  But these were big well known companies.  They were thought of more as "industrials" than as some weird subspecies.  These new companies were an entirely different kettle of fish.  By this time Hewlett Packard was a substantial company.  But it wasn't a giant.  And it had taken decades to grow to its current (as of say 1980) medium size.  The idea that a company could go from "startup" to "largest market cap in the world" as Microsoft did was literally unimaginable.  So these tech companies started off as a side business.  Wall Street could make a couple of bucks off them by managing their IPOs (Initial Public Offering of the sale of the stock) but that was just a nice but small sideline business for a Wall Street firm.

But the public's appetite for the stock of these companies seemed to be insatiable.  So Wall Street came to see them as a serious line of business that could generate substantial profits.  And the public continued to be happy to shoulder responsibility for being the greater fool.  So Wall Street got serious about the business of shearing the sheep.

Wall Street stock brokers like Merrill Lynch maintain a group of people called "stock analysts".  They are supposed to provide informed and unbiased research on stocks.  By this third stage all the stock analysts had nothing but praise for these companies as sound investments.  Why, since the characteristics of the underlying companies had not changed?  They still had no earnings and little to no prospects for earnings in many cases.  The answer is that the same company has a retail business, servicing individual investors.  But it also has a corporate business where they do IPOs, float bond issues, advise corporations on various financial matters, etc.  It was easier to get the "corporate side" business of a particular company if the analyst was saying nice things to retail customers.  So analysts were given their marching orders to come up with reasons why these companies were good investments.  This required a great amount of skill at creative writing.  But those analysts that prospered turned out to be very good at it.

Cooking analyst's reports is a bad thing and I believe it is illegal.  But I don't think it contributed much to the dotcom bubble.  A large segment of the public had started buying these stocks long before this bad behavior became wide spread.  Were some people who might have otherwise dodged the bullet enticed into purchases they would not otherwise have made by these reports?  No doubt the answer is yes.  And I am sure that a number of pension funds and other large investors ended up putting more money into these stocks than they otherwise would have.  But the dotcom bubble was of long duration.  And it was recognized as a bubble for longer and by more people than any other bubble in history.

The stock market tends to go in one direction for three to five years, or so the experts believe.  Then it changes course, at least for a while.  The market started up from a bottom in 1982 and climbed more or less continuously until 2000,  People started saying "it's time for a correction" by about  1986 or 1987.  By the early '90s there was a "when's the correction going to start" cottage industry among investors.  But the market continued on its upward course.  There were corrections (like Black Monday, October 19, 1987) but they tended to not last more than a few months.  Then the upward trend would resume.  I was introduced to both of the books mentioned above during this period.

By the late '90s jokes about the bubble were common.  I remember a fund manager that I respected saying "I don't understand it but I have to be in these stocks to make money" during this period.  But the market kept going up even though "everybody" knew it was too high and that prices couldn't be sustained indefinitely.  Over and over in this period the "sensible people" kept being made to look like fools while the "greater fools" made bales of money.  What finally happened to derail the gravy train?

The overall economy took a slight dip.  This put some people in a modest financial squeeze.  So they sold some of the stock they had made so much money on so that they could meet their other financial obligations.  This drove the price of some tech stocks down.  Much of this stock had been bought on margin (with borrowed money).  When the stock price goes down, even a little, margin investors have to either put in more money or sell the stock.  A significant number of people, either because they wanted to or because they had to, sold the stock.  This drove prices down further.  This resulted in more margin calls, etc., a classic example of a vicious cycle.  At first only a few stocks went down.  But this caused enough investors to decide it was time to sell other tech stocks too, "just to be safe".  So the decline in tech stocks spread and soon took the whole category down.  The category of speculative tech stocks went down 90% almost over night.

The tech sector has never recovered completely.  The NASDAQ is tech heavy.  Recently both the Dow Jones Industrial Average and the Standard & Poor "500" have each hit new "all time high" records.  The record high for the NASDAQ is 5048.62.  It is currently at about the 3000 level.

(to be continued)