The Story of a Bubble – and its Aftermath

December 13, 1999

By definition, a market bubble is not a very obvious affair. If the man on the street begins to see what by all accounts is a bubbly market, he will act rationally, or so the academics believe, and find a safer haven for his savings. Which means that in principle, if markets were truly rational and relied on tried and trusted measures of investment value, then bubbles could never develop.

Yet we find that commentators on Wall Street and on the Nasdaq market, as well as the people who play these markets, can be divided into two distinct camps.

One of the camps, currently having much the smaller population, believes we might be witnessing the market Bubble of The Millenium. To these people it looks like a bubble, smells like a bubble and even tastes like a bubble. Therefore it has to be a market bubble, and sooner or later – so far, mostly later – some piece of economic or financial news or some event will come along to act like the proverbial pin and the bubble will be no more.

It will not deflate slowly, ending with a whimper, but go the way all bubbles depart for bubble heaven – with a bang! (Apologies to TSE). This bubble perceiving camp as a rule refers to the very high PE ratios in historical terms that are reflected by current and recent stock prices as a major reason for their belief that a market bubble really exists.

The other camp have their eyes firmly fixed on recent the past – in particular the history of ever rising earnings reported by US companies, supported by strong and sustained growth in the US GDP. They say, "We are living in Goldilocks land. Its even better than the magic land of OZ. Inflation has gone away for ever and labour is content not to push for higher wages – they don't have to, because they are making a fortune on their investments on Wall Street. Mark our words; today's apparently sky-high PE ratios will tomorrow prove to have been outright bargains."

The question of who is right cannot be resolved by one camp merely repeating, "It is a bubble" to be answered by, "You are only jealous, because you went short of the market and got burnt, while we are coining money!" In this first part of what at the moment seems to a two or three-part series, I will hazard my way onto the field of economics and try to present evidence that could in a later part be used to justify the presence or not of a genuine bubble on Wall Street and, if this happens to be true, enable some exploration of what is likely to happen later.

My approach will be much as in 'A Japanese Tale", published by GOLD-EAGLE under the following link: 99/joubert011899.html

In that analysis I tried a systems method, looking at the situation more as an ex-physicist, to determine what the key forces in the market are likely to be. My objective is to identify any major imbalances that might be developing and, if they happen to get out of control, could later exert a strong influence on market behaviour.

I also do not consider the markets to be adequately – perhaps not even remotely – rational. People are largely driven by emotion and their respective world views determine how they cope with and react to news that are fundamentally positive or negative for the market. Irrespective of their own world view, they have little difficulty in disregarding news that do not conform to what they believe, while supportive news is seen to vindicate and reinforce their viewpoint. And remembered long after any negative news has been disregarded and forgotten.

Driving forces in the economy

It is axiomatic to say that consumer spending lies at the bottom of the economic food-chain. If consumers fail to spend on anything except essentials, then the economy takes a nose dive. If consumers spend freely to acquire luxuries and durable goods in addition to their purchases of essentials, the economy booms.

Japan has first hand experience of this basic law over the past decade. With their banking system in shreds and tatters and savings that have all but literally disappeared from reality, subject to the guarantees of the Japanese Government – which may or may not materialise when push comes to shove – Japanese consumers have pulled in the belt and are spending as little as possible on anything that is not essential.

While various ambitious programs to kick-start the economy have been announced and some already implemented, their effect has been negligible, despite recent excitement that the Japanese economy has bottomed and is ready to start improving. The bottoming part one could begin to believe – after almost 10 years of decreasing spending, the Japanese are now likely to be at the level where any further decrease in consumer spending will mean they have to forego necessities and essentials. That does not, however, mean they are now ready to replace the family car or splurge on the latest model TV or Hi-Fi, or a holiday at a fancy and expensive resort.

When consumers spend, the economy booms; when they effectively restrict spending to the purchase of essentials, the economy bleeds. Very basic. Equally basic is the fact that the amount of money they have available to spend plays a major role in their decision to look for an item that is not an essential. If the consumer has no money for anything except necessities, then it is no good for your sales target to throw the latest model car or the most fancy holiday destination or a faster PC at him via the internet, the glossy magazine or the TV. He might have his tongue hanging out and be drooling at the nice pictures, but his purse will remain shut.

And it does not matter whether the decision not to spend has its origin in a strong drive to save whatever can be found to save, or whether it is forced on the consumer by the fact that he has made commitments that leave him with little left to spend on non-essentials.

Money to spend comes from mainly two sources – income earned as wages or salaries and through making new debt. Income of course is subject to primary deductions, which leaves the wage or salary earner with what the economists call 'disposable income' – whatever remains after primary deductions have been made. It is this amount that has to pay the rent or the mortgage, the installment on the car and/or the furniture and also buy the food. After paying for these essentials, the consumer is free to spend as he/she sees fit on what the heart desires – things that are not intended to fill the stomach or keep the roof from leaking.. Even to save a little if they are so inclined.

Of course, debt increases purchasing power while the amount of debt is growing, but at the same time the need to service that debt later acts as an inhibiting factor on disposable income – the greater the amount of debt and the higher the interest rate, the wider and deeper is the hole it makes in disposable income. This means that an increase in overall debt at first acts as a boost to consumer spending, but once the ceiling is reached where consumers find it difficult to afford any further increase in debt, the cost to service the existing debt reduces the amount of money that can be spent on non-essentials.

Particularly if at this point in the cycle interest rates happen to move higher.

If, after feeling the pinch of too high a level of debt for some time, the consumer decides to repay some of his obligations to reach a more comfortable level, ready cash available for non-essentials is reduced even further. Should this action become widespread, overall spending is bound to decrease, with a cooling effect on the whole economy.

This is all nice basic theory, with a focus on just a few of the factors that determine the degree of economic activity in a country – disposable income, the amount of household debt and interest rates. Yet, from the perspective of the consumer these are the more important factors that affect his decisions to spend or not, either to ask for credit or to go to the bank for another loan or, alternatively, to repay at least part of current loans and reduce outstanding credit.

What does the real world out there tell us is happening with respect to these factors? In the following charts, based on data obtained from the website of the St. Louis Federal Reserve Board at total US disposable income was used as the reference against which trends in other variables are measured.

The absolute growth in US total disposable income is what releases the primary driving force of the economy. Increases in this variable enables American consumers to spend more on goods and services, thus adding to the GDP. Any increase in consumer debt, in absolute terms, also adds to the funds available to the consumer and thus further boosts growth in GDP, which helps to increase company earnings and thus also justifies higher than usual PE ratios on Wall Street.

And if part of the new debt is funneled to investments on Wall Street, already high PE ratios advance that little bit more.

However, once consumer debt reaches its maximum affordable level, in comparison to disposable income, it no longer serves as a source of funds to fuel the kind of spending that keeps the economy in overdrive. The effective contribution to GDP growth made by new debt is reduced or falls away completely. When that happens, even without an increase in interest rates, a high level of outstanding debt starts to act as a brake on any consumer spending that might otherwise have been funded by sustained increases in disposable income.

Under those conditions, when the requirements of servicing the debt reaches painful proportions, consumers can be expected to rather use available funds for a reduction in their exposure to debt – the more so if they had already had opportunity to purchase most of the luxuries and non-essentials that had ignited their desires initially.

The validity of these speculations is tested against the data presented in the sections that follow. The results are interesting.

Disposable income and consumer expenditure

The chart below shows the growth in total disposable personal income in the US together with the amount of that income that finds its way into the markets for goods and services as consumer expenditure. The third line on the chart, relative to the scale on the right hand side, is the ratio of consumer expenditure to disposable income.

Figure 1

Data are monthly values from the beginning of 1959, so that the history spans 40 years. During this time the US had its periods of growth and stagflation and, since about 1982, Wall Street enjoyed one of the greatest bull markets ever – a bull market that accelerated since 1993 when the US economy set off on a period of rapid and sustained growth. This extended bull market made up for the 16 years from 1966 to 1982 when the Dow Jones moved essentially sideways just below the level of 1000 points, including the bear market of 1973-74 when the Dow declined by 40%.

Sustained growth in total disposable income and expenditure of course goes hand in hand with an expanding economy as well as a growing working population. While the two main variables present a view of this growth, the graph of their ratio is much more interesting. It tells us how comfortable the population of the US have been at different times with different levels of consumer spending, relative to the amount of money they have in their pockets. The degree to which American consumers are keen to spend, or not, are of course a major determinant of US economic growth.

Observe for example that during the first half of the chart, from 1959 through to 1982, the fraction of disposable income spent on consumer goods and services showed a persistent if volatile decline. Consumers were not spending as much as before from their incomes during this period of rising and high inflation. Then, from 1982 onwards, the ratio starts to increase showing that consumers were more keen to own some of the newfangled goods that started to come on the market.

This was the time when the Dow Jones languished below 1000 points as company performance was nothing to get excited about. The cause-effect relationship may not be the dominant one for the lack of interest in Wall Street over this period, but a reduced propensity to spend surely must have had some effect, even if indirect, on the poor performance of the Dow Jones.

Note also the sideways trend in the US total disposable income from the late 80's to the early 90's. This was when the US experienced a period of recession – of low growth in the GDP and an uncomfortably high rate of unemployment, two factors that have an effect on total disposable income.

The decline in the fraction of income spent on consumer goods over the first half of the chart reached about 89% by the early 80's, where it leveled off until the early 90's. Starting in about 1993, the ratio rockets upward in a sustained rise that has consumers at the moment spending 94% of their incomes on goods and services.

Now 94% may not sound like much of an increase compared to 89%, but the chart shows the increase relative to traditional levels of spending is sustained and quite substantial.

In fact, when people spend 94% of their disposable income on consumer goods and services one could think there is not much left over for other purposes, such as repaying loans or servicing debt. One could even ask the question how households are balancing their books at all if they are on such a major and historical spending spree.

The role of consumer credit in feeding the spending spree

Consumer credit is one of the easiest ways of using debt to fund spending. It is the one avenue open to fund extra spending for people who have insufficient financial standing to request and be granted anything but a minimal loan from a bank and is therefore often and widely used by a significant proportion of the population.

However, this practice also finds favour with middle income households and is not limited to the less affluent part of the population.

As before, the chart shows the rate of growth in disposable income in combination, this time, with the increase in consumer credit. As is to be expected, the amount of relatively short term credit someone can obtain before becoming a risk to the issuer of the credit is quite small compared to the amount of income, but with total consumer credit now over $1,3 trillion, the amount is not to be sneezed at.

Figure 2

Here too the ratio of credit to income tells the more interesting story. Firstly, the relative steep increase in the use of credit, from near 14% of income to at times over 18%, during the early 60's, can perhaps be explained by more widespread use of credit and also the various options for purchasing durable good on time that became very popular then.

Note that whenever credit started to exceed 18% of income, households would apparently feel the stress of servicing what is typically quite expensive short-medium term debt. In response they soon begin to reduce the amount of credit relative to income to a more comfortable level. The spike in the ratio that shows a historically high level of credit in early 1987, may bear no relationship to the run up into the Wall Street crash of October 1987. However, it does point to reduced ability of consumers to spend themselves out of trouble at that time. The reduction in relative credit levels in the years following 1987 may well have played a role in the recession that set in during this period.

This reduction in credit may be a reaction firstly to the pain of having to service a high level of credit, as speculated earlier, but secondly also to increased uncertainty felt in the wake of the events of October 1987.

By 1993 consumers had set off on an explosion in credit, reaching a fraction of income not seen before. Then, suddenly, during the first half of 1997 the relatively steep growth in new credit reached a plateau and settled down to increase from there on at the same rate of increase in disposable income. It would seem that households had found their ceiling of comfort at that level, beyond which the amount of credit relative to income becomes too painful to contemplate.

This increase in the relative credit level coincides with the steep rise in consumer spending, also in relative terms, seen in Figure 1. Households were suddenly inclined to make greater use of consumer credit, almost as if they had other, more urgent destinations or applications for their income. We know that 1993/94 was also the start of the big boom on Wall Street and it does not require much pondering to realise that two things were going on.

Firstly, people were more confident of their improving wealth and thus also more comfortable with a higher level of credit than before. Secondly, they were becoming ever more keen to be fully invested on Wall Street and the use of credit – up to a limit – was probably perceived as one way to free a greater amount of funds for investment.

Yet in Figure 1 we see the proportion of income being spent on consumer items continuing to rise well beyond the time when the expansion of credit had already leveled off, relative to income.

Where is that additional money coming from? Given that disposable income seems stretched to a greater extent than ever before, there must be some other source of funds that now find their way into the shops.

Other forms of debt

For all but the less affluent, bank loans or a mortgage is a cheaper way of obtaining funds than to ask for credit. A bank loan against some collateral, or a real estate loan against a house, are typically much cheaper than credit and has the advantage that a greater amount of money can be obtained on a much longer term arrangement.

In Figure 3 below, the chart of the disposable income is joined by a chart of what may well be the total debt load of the typical household – a bank real estate loan against the house, other bank loans against some other form of collateral and thirdly, consumer credit. We now see that total debt is a significant, even large, proportion of disposable income and this might be the reason why a ceiling on consumer credit is reached rather quickly – it is only a relatively small portion of the total debt of an average household.

Figure 3

It is not clear from the data source whether these three factors indeed account for the total debt load of all households or whether there are categories of debt not included here, for example, mortgages at other places than banks and margin debt at brokers.

However, unless the proportion contributed to household debt by these unaccounted for categories vary greatly over time, they are of little consequence. We are interested in how the ratio of debt to income varies, and not so much the absolute values. This third chart of the ratio of debt to income therefore should be little affected by unaccounted for types of debt provided their contribution to total debt has been quite stable over time.

Note that this chart only covers the period form 1973 onwards.

Debt as a proportion of income rises substantially from 1985. Whereas the chart shows a limit of about 78% prior to 1985, the level soon rose to well over 85% of income. While levels of consumer credit relative to income declined quite rapidly and steeply after 1987, as shown in Figure 2, total debt relative to income declined much less markedly and then bottomed out by the early 90's at an historically high level. Yet this decline, too, even though less marked than that of consumer credit, must have contributed significantly to the slower economic growth of the late 80's and early 90's.

Total debt as fraction of total income increased substantially along with the steep rise in credit as from about 1993 – just when Wall Street also found its second wind in what became a very steep and sustained bull market.

However, distinct from consumer credit, which seems to have found its ceiling quite early on, as seen in Figure 2, total debt just kept on increasing right to the end of the chart (October values). Since consumer credit is included in the total debt used here, this means that loans and real estate loans from banks have continued to increase markedly right through the first half of 1999, after the growth in consumer credit had slowed to merely keep pace with the increase in disposable income.


The first important conclusion is that households seem to go through cycles of increasing amounts of credit and debt, relative to income, until a level is reached where the pain and discomfort of servicing the debt load triggers a reaction that is followed by a substantial period of time during which effort is made to reduce the amount of debt, again relative to household disposable income.

In principle, given the steep rise in disposable income, this objective might even be achieved by merely keeping the level of debt static and thus allowing its proportion of income to decrease over time.

The second observation is perhaps more pertinent to economic activity. Without doing a really detailed analysis, it still seems that the periods when households are reducing their debt are also periods when the economy moves mostly sideways with a low rate of growth. On the other hand, when the proportion of debt increases at its typically quite steep rate, the economy does very well indeed.

The creation of new credit and debt therefore appears to play a significant role in the boom periods of the US economy, while times when households scale down their relative exposure to debt coincide with less robust economic growth or even recessions.

Since 1993 America experienced a sustained rise in debt relative to disposable income and total household debt is now at historical record levels, substantially higher than ever before during the past 40 years or so. This recent increase in debt levels coincide almost exactly with the often referred to "Goldilocks years", that also saw the US economy in a major growth phase.

It was also the period of the Big Bull Market on Wall Street and there can be little or no doubt that the rising levels of debt had at least a small part to play in that development.

Looking back, it all seems so rosy. The question though is how long can the American household continue to increase levels of household debt – something that seems to be a prerequisite for good growth in the economy and perhaps even for the sustained performance of Wall Street. While the answer to that question might be difficult to find in the analyses presented here, it is evident that the end of the line for the debt-funded spending spree cannot be far off. And also the boom period for the NYSE and Nasdaq.

A one-ounce gold nugget is rarer than a five-carat diamond.