Real Clear Analysis 6: The Great Financial Build-Out 2003-2007

As discussed in prior postings, Panics and Depressions are fairly straightforward to understand: People cash in savings to invest in a new thing (Railroads, the electricity grid, the internet, etc), the build-out of the new thing requires all sorts of new work throughout the economy (steel rails, copper wire, fiber optic cable, etc,) and the economy becomes temporarily “over large”. Then the build-out ends. When it ends, the economy goes back to its natural size. Until the economy’s natural size grows to exceed the “over large” peak, there exists a period of Depression. The Depression is felt most keenly in the areas that expanded the most during the boom.

It is unfortunately a recurring pattern.

As I write in December 2008, a build-out period has clearly come to an end. What did we build-out? Most pundits seem to believe that the build-out was in housing. My analysis (from Real Clear Analysis #5) indicates that the housing build-out was too small by itself to have killed the economy. Perhaps housing didn’t act alone. Could it be that, as in 1929, the recent housing boom coincided with another, larger build-out?

As with earlier periods, data on New Security Issues provides critical clues. Issues of new securities are the gateway through which savings become new investments. In the pre-1873 period, Railroads raised billions of dollars by issuing securities (stocks and bonds), far more money than Americans had in liquid assets. In the pre-1929 period, Electric Utilities, Telephone Service Providers, Automotive Manufacturers and Construction projects also raised and invested capital in excess of national resources.

What money was raised in the pre-2008 period? It is clear that the security of choice in recent years has been the bond. Between 2000 and 2007, $10.6 trillion of bonds were issued. That was certainly a lot, but was it “unnatural”?

“Natural” growth of bond issuance has been high in the post-war period (about 10%, well above the rate of inflation). The chart below illustrates the issuance of Corporate Bonds between 1947 and 1999. With the exception of “Junk Bond Mania” in the late-1980s and the telecom-driven spike in 1993, Bond issuance remained near the long-term trendline:

Beginning in 2003, United States Corporations began a four-year run of bond issuance sharply above the trendline. Over four years, the “excess borrowing” amounted to more than $1.5 trillion (of the $10.6 trillion issued, about 14% “excess”).


Unlike earlier periods, however, the capital raised in the 2003-2007 period did not go to Railroads, Utilities or other industries directly. The bulk of the capital went to Financial Companies. The chart below illustrates that Non-Financial Corporation issuance of bonds has been relatively flat since 1990 (the green line) and that virtually 100% of the growth in bond issuance has been by Financial Corporations (the pink line). As recently as 1992, nearly half of bond issuance had been by Non-Financial Corporations. In contrast, between 2003 and 2007, 85% of bond issuance was by Financial Corporations -- $9 trillion in total.



Where did the money go? The specific facts require further analysis, but it is safe to assume that some went to mortgages, some to Private Equity and a fair chunk of it was loaned to Hedge Funds. Because Financial Corporations raised the capital, it is also safe to assume that virtually all of it was entrusted to a Financial Institution and re-loaned to others with the hope of earning a profit margin (or “spread”).

These financial assets, in their own way, are like the steel rails or copper wires of previous generations: they were installed on the expectation that they would generate a profit for the investors. When the expectation of profit is replaced by the fear of loss, a Panic ensues as investors clamor to get their money back. In 1873 it was the failure of the Northern Pacific that initiated the selling, in 1929 it was the rate ruling against Boston Edison and in 2008 it was probably the failure of Lehman Brothers.

The parallels are pretty clear. For the Railroads of 1873, the amount of debt (or “leverage ratios”) of Financial Corporations ballooned to 27/1. At the time of its demise, Bear Stearns’ leverage ratio had reached 35.5/1. Following the crash of 1929, the shares of build-out-related industries such as Electric Utilities dropped by more than two-thirds, Financial stocks have dropped by a similar amount in 2008.

The good news is that, on an economy of $13 trillion, $1.5 trillion of “excess” over four years is nowhere near 1873 or 1929 levels. Indeed, in the second quarter of 2008, the time deposits of Americans exceeded $7.4 trillion and total financial assets totaled $44.3 trillion. Historically speaking, $1.5 trillion is a drop in the bucket.

That is why the chart below is frightening: it paints a different picture. By adding the activity in 2008 to the graph, we can see that bond issuance by Financial Corporations has come to an abrupt near-halt. In a historical heartbeat, Financial Corporations have gone from raising $2-2.2 trillion per year to a run-rate of $700 billion per year.


Is this an overreaction destined to clean up four years of excess in short order – or does this imply that, like the Railroads and Electric Utilities of the past, the build-out of the Financial Sector is somehow “finished”. Were that to be the case the “hole” in the economy would not be a relatively manageable $375 billion per year (about 3% of the economy – “recession-sized”), but rather about $1.5 trillion per year (more than 10% of the economy) -- big enough to cause a depression.

It is a big question, perhaps the big question.

To be continued

Real Clear Analysis 5: How Bad is Housing?

The epicenter of the current financial crisis is widely believed to be the Housing sector. Therefore it seems useful to examine dynamics in the Housing sector to see if things are as bad as they may seem.

Americans are builders. Dating back to 1915, the value of new construction per year in the United States economy has grown – very steadily – at a 6.2% annual pace. About 45% of all construction over that time period has been Residential Housing. The Residential Housing sub-sector is prone to booms and busts. The chart below shows the progress of Construction, both Residential and Non-residential, since 1915. The data is the dollar value (in millions) of construction put in place each year:


The booms and busts in Residential Housing are difficult to see on the big chart, but can be extreme. For example, the Great Depression period illustrates that a housing boom typically produces a painful echo in the years that follow. Statistical analysis shows that the United States overbuilt housing by about $22 billion between 1923 and 1929. The boom was followed by a period of under-building between 1930 and the outbreak of war in 1941 of about $20 billion. At the peak in 1929, Residential Housing put in place was $4 billion. By 1933, the total had fallen to $0.5 billion, leaving an $3.5 billion “hole” in the economy.

The chart below illustrates the swings in Residential Housing Construction in the Depression period:

A similar analysis of the period since 1970 illustrates the booms and busts that marked recent recessions. Each of the recessions – except 2001 – were preceded by a period of overbuilding in housing. The periods of under-building have not been proportional to the overbuilding in dollar terms due to underlying growth. Nonetheless, the pullbacks have been dramatic.

The peak of Residential Housing in 1972 of $60 billion was followed by the 1975 nadir of $46 billion – a $14 billion nominal decline in activity but a $25 billion decline on a growth-adjusted basis.

The next peak, in 1986, was $55 billion above the trendline and part of an eight-year housing boom where $183 billion per year of new housing was constructed. The 1991 housing recession saw only $166 billion of new housing, $25 billion below the previous year and $77 billion below the peak on a growth-adjusted basis.

Between 1992 and 2002, the United States experienced a long period of steady growth right along the trendline.

From 2003 to 2007, however, the country experienced a sizable boom in Residential Housing totaling $329 billion above trendline expectations.

The chart below illustrates the build-outs and contractions from 1970 through 2007:


$329 billion is a lot of money. But how does it compare with past periods of overbuilding? The $3.5 billion “hole” in 1933 was nearly 6.2% of GDP. The $25 billion hole in 1975 was 1.5% of GDP. In 1991, the $77 billion hole was 1.2% of GDP. The overbuilding of 2005, which totaled $126 billion, was about 1.1% of GDP.

The overbuilding of 2003-2006 is clearly not on the scale of the Great Depression. It isn't even on par with the 1970s or 1980s. The boom of the late 1970s was $65 billion too large over four years and the bust erased $18 billion over two years on a growth-adjusted basis. The boom of the late 1980s was $300 billion too large over seven years and the bust erased $25 billion in two years.

The United States’ reaction this year has been far more severe. The annual rate of Residential Construction at the end of 2007 dropped to $414 billion from $641 billion in 2005. The rate to-date in 2008 is only $337 billion., or $304 billion below the 2005 peak and $302 billion below trendline expectations. The current rate is enough to correct the entire three-year overbuilding boom in one year. The $304 billion “hole” in the economy equates to 2.2% of 2007 GDP.

The chart below illustrates what has happened. The 2007 and 2008 data points are annual rates rather than actual totals:

It appears likely that the United States is currently over-correcting by a wide margin. Are we headed to another Great Depression? An economic "hole" of 2.2% of GDP is very, very far from 20%. Housing alone would not seem able to create the sort of long-lasting economic crisis last seen in the 1930s. Like the 1920s, though, a lot of other capital build-outs have been happening, it is the size and shape of those build-outs in aggregate that is important to understand.

To be continued

Real Clear Analysis 4: The (Second) Great Depression

The Great Depression is a hot topic of conversation these days as we all wonder “how bad this could get.” It is therefore useful to revisit the Great Depression and other Panics in American history to find or dismiss parallels with our own time.

Most histories of the Great Depression involve complex discussions of government policies, monetary policies and stock market dynamics. Indeed, many people think of the 1929 stock market crash as the cause, or even the entirety, of the Great Depression itself.

Data analysis suggests that the cause of Panics and Depressions may be simpler than most historians suggest.

To recap, the First Great Depression (of 1873) was the result of a massive capital build-out that doubled the Railroad capacity of the country and connected it coast-to-coast by rail. The build-out consumed far more capital than the country had in liquid assets and therefore required unprecedented levels of new debt and equity. The build-out ended naturally when the project was complete following the driving of the Golden Spike. Without the impact of the build-out, the economy was doomed to shrink back to “normal”, though the productivity impact of the new technology rendered the prospects for pricing and employment even lower than they had been before. As the poor outlook for profits and threats to invested capital became clear, investors reasonably tried to get their money back before it was lost forever. Though it seemed like a panic, the Panic of 1873 was rational and a symptom of the wider economic situation rather than a cause. The country was left with a dramatic slowdown that the Government was essentially powerless to combat even if it had tried. It was only as the United States grew into the added capacity and increased productivity that the Depression could end.

A similar narrative applies to the Second Great Depression (of 1929). Instead of Railroads, the 1920s saw the concurrent build-out of the electricity, telephone and motor vehicle industries that consumed about $50 billion in capital. In addition, the 1920s saw an unprecedented level of investment in residential and commercial construction that consumed another $49 billion. All told, the $100 billion of capital-build-outs made the economy $10-$13 billion “too big” in each year of the latter part of the 1920s.

With frightening synchronicity, each of the build-outs reached completion in 1929-1930: At a moment when there were about 25 million non-rural households, electricity, telephones and automobiles came to reach virtually all of them. As with the Railroads of 1873, when the build-out was finished it simply stopped. Millions of miles of wire had been run and millions of cars and tractors had been built in a headlong rush to equip Americans with the latest technology. Once equipped, the build-out ended.

The end of the build-out left a $13 billion annual hole in the economy that was made even worse by the enhanced productivity made possible by technology. Electricity, in particular, made possible dramatic gains in manufacturing productivity that displaced thousands upon thousands of workers.

At the same time, the new technologies carried the seeds of their own destruction: once capacity was in place, the downward pressure on electricity rates, telephony rates and automobile prices was relentless. As in 1873, the immediate cause of the 1929 Panic had been the sudden realization that profits would not be anywhere near what investors had dreamed. In 1873 it had been the failure of the Northern Pacific, in 1929 it was the news that Boston Edison profitability would be limited by lowered rates. The diminished outlook for profits had a predictable, albeit violent, effect as investors tried to get their money back by selling shares, cashing in bonds and making runs on their banks.

By 1929 a significant slowdown was inevitable – the next build-out was not on the horizon (and there may not have been enough capital to fund it if it had been). The scale and shape of the Depression was an echo of the build-out that had caused it. Indeed, while the Great Depression is seen as a generalized disaster, it was actually quite concentrated: 2.2 million of the 2.8 million lost manufacturing jobs in the early 1930s were in industries directly related to the build-out. Likewise, in the stock market, Industrial stocks had tripled from their 1932 lows by 1937 while Utility stocks had recovered only modestly.

The Depression may also have been inevitable, though much effort was made to fight it. Beginning in 1931, the Federal Government made significant efforts to mitigate the impact of the slowdown. Between 1931 and 1937 the Government took on $20 billion of new debt to finance social and infrastructure spending, more than doubling the National Debt. Unfortunately, it was $20 billion of money thrown into a $100 billion hole.

There are four simple lessons to draw from the two Great Depressions:

1. Financial Panics are a symptom, not a cause, of inevitable economic slowdowns following massive capital build-outs.
2. The size and shape of a Depression is an echo of the build-out that preceded it. The ability of new economic activity (or Government spending), to neutralize the effects of the Depression depends entirely on the size of the “hole” that needs to be filled.
3. Productivity makes things temporarily worse – the greater the productivity gains, the worse the Depression.
4. If the size of the “hole” in the economy is large, the Government will not have sufficient resources to meaningfully mitigate its effects – especially if existing Government debts are large.

The relevance to our own times is equally straightforward:

1. Are the financial markets to blame? Despite all the attention on banks and bankers, they are not the cause of the current economic slowdown. The cause is a capital build-out that has already happened.
2. What build-outs drove the economy beyond its natural size and by how much? Home Building, Technology and Retail Infrastructure have certainly been important build-outs, but more important may be the infrastructure build-outs in China, India and other countries where Americans and others have invested in pursuit of greater productivity.
3. How great are the permanent productivity gains from recent investments in technology and infrastructure? Production costs per unit have dropped meaningfully by the use of cheap overseas production. The permanence of these productivity gains will have a great impact on the length of the current slowdown.
4. Does the Government have the resources to mitigate the effects of the slowdown? Since the Great Depression (of 1929), the Government definitively has not salted away resources in order to fight the next slowdown. Indeed, the opposite has happened, with borrowing becoming only greater. The Government doubled the National Debt in a largely failed bid to mitigate the Great Depression. A similar doubling of the current debt is certainly impossible. The failure to “make hay while the sun shone” will limit flexibility to fight the current slowdown.

To be continued

Real Clear Analysis 3: The (First) Great Depression

The Great Depression is a hot topic of conversation these days as we all wonder “how bad this could get.” It is therefore useful to revisit the Great Depression and other Panics in American history to find or dismiss parallels with our own time.

Most histories of the Great Depression involve complex discussions of government policies, monetary policies and stock market dynamics. Indeed, many people think of the 1929 stock market crash as the cause, or even the entirety, of the Great Depression itself.

Data analysis suggests that the cause of Panics and Depressions may be simpler than most historians suggest.

There have actually been two “Great Depressions” in U.S. history, the first one followed the Panic of 1873 and the second one followed the Panic of 1929. Looking at the 1873 Panic first is useful because it came at a “simpler time” and is easier to understand analytically.

Between 1867 and 1873, the post-Civil War United States connected itself coast-to-coast by rail. In six years the U.S. added 33,275 miles of new railroad to a base that, in 1866, had been 36,801 miles. This overall doubling of the rail infrastructure cost about $1.8 billion to install. In context, the investment in railroads equaled 2.3 times the entire deposit base of all the banks in the United States in 1870 ($775 million). It was a “bet the country” sort of investment.

It was also clearly not the sort of investment that could be made by reinvesting profits or by saving a bit here or there. The 1867-1873 investment was a “capital build-out” where national (and international) assets were turned into cash and reinvested behind a new and exciting technology (fast travel and fast-shipped goods). From an economic perspective, it was “unnatural”: a nonrecurring event that created nonrecurring demand.

Unnatural, maybe, but it must have felt good. Demand rippled through the entire economy, requiring iron and steel for rails, wood for ties, dynamite for blasting tunnels, coal for steam engines; workers to make the steel and mine the coal, workers to install the rails, foremen and managers to oversee the operations. Production of steel and iron for rails doubled from 413,000 tons in 1867 to 893,000 tons in 1872. But that wasn’t enough – an additional 532,000 tons were imported, creating demand for international financiers, shipping and land transport. The increased demand for everything in a young economy created new levels of activity and generated vast amounts of wealth in a country that had never known it before.

Then it stopped.

It stopped for a good reason: the capital build-out ended. It could have stopped the day the Golden Spike was driven in 1869, but the rush to complete the link-up had resulted in a lot of sloppy work that had to be re-done. There were also other huge railroad projects that began before the Golden Spike and took a while to be completed or to fail. In fact, it was the collapse of the Northern Pacific, an also-ran in the race to connect the coasts, which was the immediate trigger of the 1873 Panic.

By 1873, a forward-looking person had a lot of things to panic about:
- The next five years saw railroad construction fall to roughly 2,300 miles per year. That is $274 million less per year than the peak in 1871. If build-out costs were, say, 30% labor, that would be $82 million worth of jobs lost in a time when a Carpenter on the Erie Canal made $2.50 per day.
- Imports of steel rails fell from 532,000 tons in 1872 to 149,000 tons in 1874 on their way to 11 (not 11 thousand, just 11 tons) by 1878.
- Iron horses don’t eat Hay: the value of a ton of Hay dropped from $14.52 in 1872 to $8.59 by 1877.
- Fast shipping opened vast new areas of land for agriculture, resulting in an agricultural price collapse: Corn prices fell from 65 cents a bushel in 1874 to 42 cents in 1875 on their way to 35 cents in 1877.

Ironically, it was a primary objective of the build-out that created many problems: productivity. Improved shipping resulted in higher yields that resulted in lower prices and economic hardship. 1873 was a double whammy: the end of the build-out and the beginning of a productivity boom.

To the financial markets of 1873 the only companies that mattered were the Railroads. As late as 1907, Railroads accounted for over two-thirds of what we would call “Wall Street Activity”. In the 1870s, it would have been virtually 100%.

To the Railroads, the capital build-out carried within itself the seeds of its own destruction. It put capacity in place ahead of demand, resulting in a rate collapse. Rates to ship a bushel of Wheat fell from 34 cents in 1872 to 20 cents in 1877. Railroad profits peaked at $189 million in 1874 before eventually falling to $170 million in 1877. Massive growth in infrastructure, equity and debt had generated no new profit and had seriously destabilized the Railroad companies.

Railroads by 1874 carried a crushing debt burden of over $4.2 billion. In 1874, the leverage (total debt/net earnings) of the Railroads was 22 times. By 1877 it was 27 times earnings – Railroads had become “junk” well before the term had been invented. When this outlook became clear, investors of capital justifiably sought to get back what remnants of their capital that they could by selling stock and calling in loans.

From a historical perspective, it seems clear that the Panic of 1873 was probably not a “panic” at all, but rather a sudden recognition that economic activity was about to rapidly decline and that a meaningful portion of the stock market was in danger of bankruptcy.

Government land grants tied to the speed of the build-out acted as the “easy credit” that helped fuel the boom, but they weren’t the reason the railroads were built. “White collar crime” as exposed in the Credit Mobilier scandal made the build-out more expensive, but not meaningfully so in aggregate.

If the Panic had become inevitable by 1873, could the Depression that followed have been avoided? The decision to build Railroads coast-to-coast had been a strategic one based on sound principles and good intentions. Building out half a network to limit the costs would not have achieved the project’s goals. Slowing the build-out might have been helpful, but natural forces of competition and the “unnatural” force of government incentives conspired to accelerate the process.

Could the impact have been softened? Build-out related spending of about $274 million per year had evaporated overnight. Total government revenues at the time were only $333 million. Spending the small surplus of $43 million would have covered a small fraction of the $274 million hole in the economy. Railroad debt exceeded $4 billion, about twice the federal debt of just over $2 billion (itself swollen by Civil War costs). Bailing out the Railroads would have meaningfully increased the National Debt that was already six times revenue.

It is possible that the Government could have “taxed the boom” early on, essentially forcing savings to be drawn later, but that would have taken a degree of foresight that Governments rarely achieve. By 1873, it was too late, the hole was too big and the shovel too small.

Ultimately, the impact of the Panic and subsequent six-year Depression was an echo of the enormity of the activity undertaken in the period 1867-1873. The time taken to work through the Depression was the time it took to grow out of the “hole” ($274 million per year) plus the time it took to mop up the productivity improvements generated by the build-out.

The next Great Depression was to be eerily similar.

To be continued

Data from the Statistical Abstract of the United States, various years

Real Clear Analysis 2: Is It The 1970s Again?

The specter of the 1970s is being raised frequently to warn investors that the recent stock market collapse could prove to be long lasting. In that light, it seems useful to revisit the Bear Market that ran from the market peak of January 1973 to the breakthrough in the fall of 1980.

The popular story of the Bear Market is familiar: in 1973 OPEC launched an oil embargo in order to drive up oil prices. The oil shock resulted in long lines, fistfights and economic devastation. In 1974, the pressure of the Watergate scandal forced the resignation of President Richard Nixon. Gerald Ford became the first President of the United States never to be elected. In 1976, the United States elected a “tax and spend” President in Jimmy Carter. The aging industrial base of the United States was failing under heavy competition from abroad, especially Japan. Top marginal tax rates exceeded 70%, discouraging entrepreneurs. President Carter complained of a “malaise” in the country and “stagflation” in the economy. The final cut occurred when Iran, flush with oil profits and fresh from Islamic revolution, took 52 hostages at the U.S. Embassy in Teheran, subjecting Americans to a morale-crushing “America held hostage” vigil of 444 days that seemed to underscore American helplessness. The footage of the smoking hulks of helicopters that couldn’t even get off the ground to attempt a rescue summed up the decade. In 1980, Ronald Reagan was elected and it was “Morning in America”. The hostages were home within days. Shortly thereafter, the United States suffered a sharp, but cleansing recession and then all was pretty much well. Stocks, so often a barometer of American national health, mirrored the country’s ordeal.

That’s the story, anyway.

From an analytical perspective, the story is a bit simpler.

In January 1973, adjusted for inflation, the earnings of the major U.S. Corporations were $34 per share in aggregate. The stock market stood at 118.4. By the cold December of 1974, the stock market stood at 67.1, down 43% from less than two years before. Economic devastation? Hardly. In December 1974, earnings were $38 per share – up more than 10% from 1973. At no point in the period did earnings fall below $31.75 (in September of 1975), less than 10% below 1973. Malaise? A crushing lack of confidence in the future? No, it was interest rates.

The chart below shows the market valuation (expressed in Price-to-Earnings terms where Earnings are the average of the preceding 10 years) and the interest rate on long-term government bonds. Earnings times the number expressed by the pink line yields the market average. With the “multiple” in 1975 at half the 1973 levels even decent increases in Earnings could not produce stock market gains. The market “multiple” did not reach 1973 levels again until 1992, when interest rates were once again below 7%.

The relationship is clear, but requires some imagination. In the next chart the same data is shown, but I have inverted the Price/Earnings ratio into an Earnings/Price ratio (and multiplied it by 100):


In late 1974/early 1975 there may have been a loss of confidence (the pink spike), but otherwise the fate of the stock market was in the hands of Interest Rates.

By May 1977, corporate earnings broke above $38 per share and would not fall below that level for the rest of the decade. The stock market remained below 1973 levels because interest rates held down the valuation of stocks relative to earnings. Indeed, when stocks finally broke above the 1973 highs (in 1980), earnings were about to go into a tailspin that saw them fall as low as $28 by 1983. Yet the stock market average at that point in May 1983 was 164 (up from 118 in 1973). Interest rates tell virtually the entire story.

At first glance, the current crisis appears to be completely unlike the 1970s because Federal Reserve interest rates are very low. On the surface, this is true.

In reality, the scarcity of credit, not so much from banks but from bond investors, has driven borrowing costs for companies up dramatically. In recent weeks interest rates and yields on investment grade corporate bonds have jumped to about 7% and rates and yields on non-investment grade debt are now near 13%. These terribly high interest rates on loans that are actually available to borrowers – even good credits – are a contributing cause to the current market declines. The chart below compares the yield, effectively the interest rate, on a major High Yield bond fund to the market valuation (inverted, an E/P ratio as above) for 2008:



It appears to be happening again, though the trouble is in the private sector. When the Government says we “need to get lending going again”, this is no doubt a big part of what they mean. Based on the experience of the 1970s, when these effective interest rates start to come down, the stock market should start to go up.

[An easy barometer to watch is the yield on the Powershares High Yield ETF, ticker symbol PHB, which has a 12-month yield of 14.51% as of 11/20/2008.]

Thanks to Richard Smith for getting me started on this analysis

Historical raw data from the Shiller Irrational Exuberance database

Real Clear Analysis 1: Gas Prices and Mortgages

In all the talk and all the panic of the last months there has been a lot of blaming, a lot of political posturing and quite a bit of fabricating… but few facts.

The basic story being told is straightforward: The system is unraveling because of bad mortgages, mostly mortgages issued to lower-income households. When bank loans go bad, bad things happen (that is the topic for another day).

But why did the mortgages go bad? Looking at the data, such as it is, the answer appears to be pretty simple, but it requires doing some math. The facts come from disparate sources and from different years, but they hang together pretty well.

There are about 60 million households in the United States that earn less than $50,000 per year. That is more than half of all households. These households generate average income of about $24,800 annually, or about $477 per week.

About 59% of households in this segment are homeowners. About 32% have a mortgage. That totals about 19 million mortgages among lower-income homeowners. The average mortgage in this segment is about $68,500. On average, there is also about $3,200 of home equity debt, resulting in total debt of $71,700 per household. In total, the 19 million households in this segment represent nearly $1.4 trillion of borrowing.

Back in 1995, Adjustable Mortgage (ARM) Rates averaged 4.1%. At that point, the payment on the average mortgage was $4,200 per year or $81 per week. $81 per week was 17% of pretax income. This is clearly a reasonable level of home debt. Of the 60 million lower-income households, about 54.5 million own a car. Those lower-income, car-owning households burned more than 51 billion gallons of fuel annually in 2001. Let’s guess that those households burn, say, 60 billion gallons per year today. That works out to 1,100 gallons of gasoline per household, or about 21 gallons per week. That is about a tank of gas per week, clearly not extreme usage.

Back in early 2005, when the dream of home ownership burned brightly, the average price of a gallon of gasoline in the United States was $1.82. 21 gallons of gas times $1.82 equals $38, or about 8% of pretax income. Together, gas and mortgage were 25% of pretax income, leaving about $350 per week for taxes, food, clothing, heat and other things.

Then came the oil shock.

By mid-2008, the price of an average gallon of gasoline had peaked at $4.10 per gallon. 21 gallons of gasoline times $4.10 equals $86 per week versus the $38 back in 2005. $86 represents 18% of pretax income for households in this segment for gasoline alone. In short, nearly $50 in higher gasoline costs per week made perhaps millions of households simply unable to meet the necessities of life and still pay the mortgage. To make things worse, ARM rates had increased to 5.5%, raising the mortgage payment to $95 per week. Gas and housing that had once cost $119 per week increased to $181, or 38% of pretax income.

Nobody comes off as a criminal in this analysis. Homeowners took on reasonable debt burdens and lenders made reasonable loans based on the economic situation at the time.

Were there excesses? Undoubtedly. Were the excesses what brought down the system? The facts would argue “no”: Gasoline prices are the root cause.

In June, just over 1.5 million mortgages had gone or were going bad and another million were falling behind. Considering the economic headwinds facing 19 million lower-income households with mortgages, we are lucky that the numbers aren’t worse. The majority still current on their mortgages had to have been cutting corners or dipping into savings to meet their payments.

By August of 2008 the old rule of thumb had been proven correct: a 10% reduction in demand has resulted in a 50% reduction in oil prices. The national average gas price of $2.35 is at 11% of pretax income for lower-income households and may be nearing an affordable level.

An extra $30 per week might even mean Christmas is on again for the children of these households, though rising unemployment may derail a recovery.

Assuming we survive, what do we take away? We must not fail to learn the lessons of 2007-2008. The danger has not passed. For Americans, reducing demand for oil – particularly imported oil – may be the preeminent strategic issue of our time.

Higher-income households hold the key. They burned 62 billion gallons of gas in 2001, perhaps 75 billion today. A strategic reduction in gasoline use by these households would be felt around the world.

Two-thirds of higher-income household gasoline usage is in cars getting fewer than 22 miles per gallon. Even a 10% increase in fuel efficiency in this segment would reduce demand by 5 billion gallons of gas annually.

Permanently lower demand for gasoline will lower prices dramatically, relieve pressure on lower-income households, deny our enemies the cash that funds their schemes against us and perhaps even allow our financial system the chance to get healthy again.