Credit Expansion, Crisis, and the Myth of the Saving Glut
Credit Expansion, Standard Money, and Fiduciary Media
The Stock Market and Real Estate Bubbles
Evasion of Responsibility for the Bubbles
The Saving Glut Argument
The Non-Existence of a Saving Glut
Current Account Deficits as a By-Product of the Increase in the Quantity of Money
Net Saving as a By-Product of the Increase in the Quantity of Money
Summary and Conclusion
Readers who are already familiar with the nature of credit expansion and the concepts of standard money and fiduciary media should skip the first section. Readers who are also already familiar with the role of credit expansion and fiduciary media in generating the stock market and real estate bubbles should skip the second section as well and proceed directly to the third section “Evasion of Responsibility for the Bubbles.”
Credit Expansion, Standard Money, and Fiduciary Media
Since the mid-1990s, the United States has experienced two major financial bubbles: a stock market bubble and a housing bubble. In both instances, the bubble was inaugurated and sustained by a process of massive credit expansion, i.e., the lending out of newly created money by the banking system, operating with the sanction and support of the country’s central bank, the Federal Reserve System.
The concept of credit expansion rests on two further concepts: standard money and fiduciary media. Standard money is money that is not a claim to anything beyond itself. It is that which, when received, constitutes payment. Under a gold standard, standard money is gold coin or bullion. Under a gold standard, paper notes, which were claims to gold, payable on demand, were not standard money. They were merely a claim to standard money, which was physical gold. The dollar was defined as a physical quantity of gold of a definite fineness, i.e., approximately one-twentieth of an ounce of gold nine-tenths fine.
Today in the United States, standard money is the irredeemable paper currency issued by the United States government. That money is not a claim to anything beyond itself. Receipt of such money today constitutes final payment.
The total of standard money today is the sum of the outstanding quantity of paper currency plus the checking deposit liabilities of the Federal Reserve System. Since the Federal Reserve has the power to print as much currency as it likes, and thus is always in a position to redeem its outstanding checking deposits in currency, these checking deposit liabilities can properly be viewed as a kind of different denomination of the paper currency, much like hundred dollar bills that are to be redeemed for notes of smaller denomination, or one-dollar bills that are to be redeemed for notes of larger denomination. Thus the total supply of standard money is to be understood as the sum of the supply of paper currency in the narrower sense plus the checking deposit liabilities of the central bank.
These two magnitudes, currency plus checking deposit liabilities of the central bank, when taken together, are known as the “monetary base.”
In December of 1994, the monetary base was $427.3 billion. In December of 1999, it was $608 billion. In December of 2007, it was $836.4 billion. In all years prior to 2008, the overwhelming portion of the monetary base consisted of currency. For example, in December of 2007, currency was $763.8 billion, while, as just noted, the monetary base as a whole was $836.4 billion.
A portion of the currency outstanding and a portion of the checking deposit liabilities of the Federal Reserve constitute the reserves of the banking system. These reserves are the standard money that the banks possess and can use to meet the withdrawals of depositors requesting currency. The reserves are also used to meet the demand of other banks seeking to redeem net balances accruing in their favor in the process of the clearing of checks.
In December of 1994 such reserves were $61.36 billion; in 1999, they $41.7 billion; in December of 2007, they were $42.7 billion.
Normally, as the overall quantity of money in the economic system increases, bank reserves increase more or less in proportion. The fact that reserves were almost one-third lower in December of 1999 than in December of 1994, and then barely higher in December of 2007 than they were in December of 1999, despite major increases in the quantity of money over these years, is a major anomaly. It reflects the long-standing, deliberate policy of the Federal Reserve System of reducing and even altogether eliminating reserve requirements.
As a recent scholarly paper noted,
The Depository Institutions Deregulation and Monetary Control Act of 1980 had begun phasing out interest-rate ceilings on deposits andThe concept of standard money underlies the concepts of fiduciary media and credit expansion. As I wrote in Capitalism, “Fiduciary media are transferable claims to standard money, payable by the issuer on demand, and accepted in commerce as the equivalent of standard money, but for which no standard money actually exists.”
modified reserve requirements in complex ways. Combined with subsequent administrative deregulation under Greenspan through January 1994, these changes left all the financial liabilities that M2 adds to M1—savings deposits, small time deposits, money market deposit accounts, and retail money market mutual fund shares—utterly free of reserve requirements and allowed banks to reclassify many M1 checking accounts as M2 savings deposits. M2 and the broader measures became quasi-deregulated aggregates with no legal link to the size of the monetary base.
The overwhelmingly greater part of our money supply today consists of fiduciary media in the form of checking deposits of one kind or another. For example, as of December 2007, the total money supply of the United States, i.e., currency plus bank deposits of all kinds that are subject to the writing of checks, including the making of payments by debit card, was $6901.9 billion; at the same time, the monetary base was $836.4 billion. Accordingly, the amount of fiduciary media in the United States was equal to the difference, which was $6065.5 billion. This was the sum of money representing transferable claims to standard money, payable on demand by the various banks that issued them, accepted in commerce as the equivalent of standard money, but for which no standard money actually existed.
The only standard money that the banks had available with which to redeem their checking deposits was $42.7 billion in standard money reserves. These $42.7 billion of reserves were the standard-money backing for a total of $6108.2 billion checking deposits, i.e., deposits equal to the sum of $42.7 billion + $6065.5 billion. To say the same thing in different words, there was full, 100 percent standard-money backing for $42.7 billion of deposits, and no standard-money backing whatever for $6065.5 billion of deposits, which latter constituted fiduciary media.
The quantity of fiduciary media in existence at any time represents the cumulative total of all of the credit expansion that has taken place in the country’s money supply up to that time. It represents the sum of all of the loans and investments that the banking system has made based on the foundation of the creation of money out of thin air. The difference between the amount of outstanding fiduciary media at two points in time represents the credit expansion that has taken place in the interval.
The simplest way in which to understand the process of the creation of fiduciary media and credit expansion is to imagine a deposit of standard money in the form of currency into a checking account. After making the deposit, the depositor has just as much spendable money in his possession as he did before making it. Instead of a roll of currency, he has a checking balance of equal amount. Either way, he can spend the same amount of money. Before making his deposit, he would have had to peel off bills from his roll in order to make payments. Now, instead, he writes checks and makes payment by check. Instead of his roll of currency diminishing each time he peels off a bill, his checking balance diminishes each time he writes a check. In the one case, the spendable money in his possession is his roll of currency; in the other it is his checking balance.
Up to this point in our imaginary scenario, there has been no creation of fiduciary media and no credit expansion. The money supply does not exceed the quantity of standard money. In the one case, before making his deposit, the standard money is in the possession of an individual. After the individual makes his deposit and holds money in the form of a checking balance, the same quantity of standard money is in the possession of his bank. Under such conditions, the quantity of money in the economic system is equal to the quantity of standard money held either by individuals as holdings of currency, or by banks as reserves against the checking deposits of those individuals and equal in amount to the size of those checking deposits.
Fiduciary media and credit expansion enter the picture insofar as the banks in which standard money has been deposited proceed to lend out the standard money that has been deposited with them. To the extent they do this, borrowers from the banks now have spendable money in their possession which is in addition to the spendable money in the hands of the banks’ checking depositors. There has been a creation of new and additional money, which new and additional money represents fiduciary media and an equivalent expansion of credit.
The currency which the banks lend out can easily, and almost certainly will, be deposited. When it is deposited, the same process of the creation of fiduciary media and credit expansion can be repeated. Indeed, under the conditions largely created by Greenspan, checking deposits came to stand in a multiple of more than 160 times the standard money reserves of the banks. In December of 2007, there were $6901.9 billion of checking deposits backed by a mere $42.7 billion of standard money reserves.
In modern conditions, of course, banks do not lend currency. Rather, they simply create new and additional checking deposits for their borrowers. When the borrowers spend those checking deposits by writing checks of their own, the people who receive the checks in turn deposit them in their banks. Those banks then call upon the banks that have created the deposits, for payment. This entails a shifting of standard money reserves from the one set of banks to the other.
To the extent that all banks have engaged in the process of checking deposit creation, the reserve balances due from any bank may be more or less closely matched by the reserve balances due it from other banks. This is because the checks written by its customers to the customers of other banks will be more or less closely matched by checks written by the customers of other banks to customers of this bank. In such a case the only movement of reserves will be the net amount due in the clearing.
From December of 1994, prior to the start of the stock market bubble, to December of 2005, shortly before the end of the housing bubble, the quantity of fiduciary media increased from $1.91 trillion to $4.93 trillion. This represented a compound annual rate of increase in excess of 9 percent over the eleven-year period. From December of 1999, shortly before the start of the housing bubble, to December of 2005, the amount of fiduciary media increased from $3.25 trillion to $4.93 trillion, which represented a compound annual rate of increase of 7.21 percent.
The increase in the quantity of fiduciary media over the period as a whole is significant, not just the increase that took place over the period of the housing bubble itself. This is because fiduciary media created in the years prior to the housing bubble played an important role in financing that bubble. And the same was true of the role of fiduciary media created in the years prior to the stock market bubble in financing that bubble.
As interest rates rose in the latter parts of these two bubbles, vast checking balances created earlier, that had been held as though they were savings accounts, and on which a modest rate of interest was being earned, were drawn into the financing of stock market purchases in the one case and housing loans in the other. The transformation of these deposits from de facto savings accounts into de facto checking accounts was based on the combination of their having had the potential for check writing all along, together with a rise in the rates of return that could be earned by switching their use from a vehicle for savings into a vehicle for buying investments. The rise in rates of return in the one case was in the gains to be had from stock market investment; in the other, in rates of interest on various vehicles for financing housing and real estate purchases.
It might be thought that what I have said of the transformation of deposits on which checks could be written would largely apply also to genuine savings deposits, on which checks could not be written. For the rise in rates of return would provide the same incentive to move funds from them into more lucrative investments. This is true. But nevertheless, there is a crucial difference.
Before the savings deposits can be spent, they must first be converted into checking deposits. All of the checking deposits that come under the heading of M1, most notably those held at commercial banks, require that those banks hold significant reserves, typically in an amount equal to 10 percent of a bank’s total deposits in excess of $44 million. Savings deposits in contrast have not required the holding of any reserves whatever for many years, and even when they did require the holding of reserves, it was at a far lower percentage than applied to checking deposits.
As a result, any movement of funds from savings into checking accounts entails an increase in required reserves. To obtain these additional reserves, banks must sell various assets, the effect of which would be to reduce their prices and to raise their effective yields to the new buyers. Unless the Federal Reserve intervened to provide new and additional reserves equal to the increase in the need for reserves, the effect would be not only a rise in interest rates but a general tendency toward a contraction of credit. This last would result from the loss of reserves by banks whose reserves were already at the bare minimum necessary to conduct operations.
In contrast, the use of savings held in accounts with already existing check-writing privileges to make purchases does not require any additional reserves. The problem of a need for additional reserves arises only insofar as a net movement of funds might occur, through the clearing, from checking accounts of a kind requiring no reserves to checking deposits of a kind that do require reserves. Checking deposits with no legal reserve requirements are money-market deposit accounts and retail and institutional money market funds. Checks drawn on such accounts and then deposited in other such accounts do not require any additional reserves. Additional reserves are required only when and to the extent that checks drawn on such accounts and deposited in conventional checking accounts exceed the volume of checks coming from conventional checking accounts and deposited in such accounts.
To the extent that the Federal Reserve is willing to supply the necessary additional reserves to meet the greater need for reserves arising from such a movement of funds, all checking deposits come to stand on an equal footing as sources of spendable money. And so too do savings deposits that end up being convertible into checking deposits with no net increase in the scarcity of reserves because the Fed has enlarged the supply of reserves to the same or even greater extent than the increase in the amount of reserves required as the result of such conversion.
Consistent with the fact cited earlier that total reserves were substantially lower in December of 1999 than they had been in December of 1994 and grew only slightly from December of 1999 to December of 2007, it must be pointed out that additional reserves can be supplied by the Fed by means of its reducing or eliminating reserve requirements at various points in the banking system. Thus, for example, when the Fed eliminated the requirement that once existed that a 3 percent reserve be held against savings deposits, all of the reserves previously held to meet that requirement became equivalent to a supply of new and additional reserves of that same amount.
The same was true when the Fed allowed commercial banks on weekends and holidays to “sweep” substantial parts of their outstanding checking deposits into types of accounts that did not require reserves. This too made a substantial portion of already existing reserves the equivalent of new and additional reserves. Indeed, the amount of such new and additional reserves constituted such an excess of reserves above the now diminished reserve requirements, that the Fed was obliged to reduce the outstanding amount of reserves by means of resorting to “open-market operations” in which it sold some of its holdings of government securities in exchange for newly excess reserves.
The Stock Market and Real Estate Bubbles
Credit expansion was the source of the funds that fueled both the stock market and the real estate bubbles. In the case of the stock market bubble, credit expansion provided funds for the purchase of stocks. The sellers of the stocks then used the far greater part of their proceeds to purchase other stocks, whose sellers did likewise. In this way, the new and additional money created by credit expansion traveled from one set of stocks to another, raising the prices of the great majority of them. It continued to do this so long as the credit expansion went on at a sufficient rate.
Ultimately, a sufficient rate would have had to be an accelerating rate. This is because rising share prices resulted in people feeling richer and thus believing themselves able to afford more luxury goods. It also led to a stepped up demand for physical capital goods by firms coming into possession of the new and additional money by virtue of sales of stock of their own. The issuance of such stock and use of the proceeds to finance the purchase of physical capital goods was encouraged by the fact that the rise in stock prices made it more and more attractive in comparison with acquiring capital goods through the purchase of stocks in other companies.
Thus, an important later effect of the credit expansion was a tendency for funds to be withdrawn from the stock market, for the purchase of luxury consumers’ goods and also of physical capital goods. To offset this withdrawal of funds, more rapid credit expansion would have been necessary.
When, instead of an acceleration of the credit expansion, there was a diminution in its rate, the basis of the market’s rise was doubly undercut. Since the funds provided by credit expansion had come to represent an important part of the demand for stocks, the reduction in credit expansion constituted a reduction in that demand. Coupled with the outflows of funds just described, the result was that share prices began to plummet. Their fall was compounded by the unloading of shares by people who had purchased them for no other reason than their expectation of a continuing rise in stock prices.
The more recent, real estate bubble originated in the Fed’s panic-response to the collapse of the stock market bubble it had caused earlier. To overcome the effects of that collapse, it progressively reduced its target federal-funds rate, i.e., the rate of interest banks pay one another on the lending and borrowing of funds that qualify as reserves against commercial-bank checking deposits. In this way, it launched a new and more momentous credit expansion.
For the three years 2001-2004, the Federal Reserve created as much new and additional money in the form of additional bank reserves as was necessary to drive and then keep the federal-funds rate below 2 percent. And from July of 2003 to June of 2004, it drove and kept it even further down, at approximately 1 percent.
The new and additional money created by the banking system on the foundation of these new and additional reserves appeared in the loan market as a new and additional supply of loanable funds. The effect was a reduction in interest rates across the board.
Because interest is a major determinant of monthly mortgage payments, the fall in interest rates made home ownership appear substantially less expensive. As a result, a great surge in the demand for mortgage loans and the in the purchase of homes took place. Instead of pouring into the stock market as in the previous bubble, the funds created by credit expansion now poured into the real estate market and drove up the prices of homes and commercial real estate rather than the prices of common stocks.
In the stock market bubble and even more so in the real estate bubble there was both large scale overconsumption and malinvestment. These are the two leading features of booms as explained by the monetary theory of the trade cycle developed by Ludwig von Mises. In both cases, the rise in the price of major assets—most notably, stocks and homes respectively—led people to believe that they were richer and could thus afford to consume more. In both cases, particular branches of industry were greatly overexpanded relative to the rest of the economic system, resulting in a subsequent major loss of capital. In the stock market bubble, the malinvestment was mainly in such things as the “dot.com” enterprises that later went broke. In the real estate bubble, it was in housing and commercial real estate.
Evasion of Responsibility for the Bubbles
Credit expansion is what was responsible for both the stock market and the real estate bubbles. Since its establishment in 1913 and certainly since the expansion of its powers in World War I, responsibility for credit expansion itself has rested with the Federal Reserve System. The Fed is the source of new and additional reserves for the banking system and determines how much in checking deposits the reserves can support. It has the power to inaugurate and sustain booms and to cut them short. It launched and sustained the stock market and real estate bubbles. It had the power to avoid both of these bubbles and then to stop them at any time. It chose to launch and sustain them rather than to avoid or stop them.
To be responsible for a bubble and its aftermath is to be responsible for a mass illusion of wealth, accompanied by the misdirection of investment, overconsumption, and loss of capital, and the poverty and suffering of millions that follows. This is what can be traced to the doorstep of the Federal Reserve System and those in charge of it. It is destruction on a scale many times greater than that wrought by Bernard Madoff, the swindler who first made his clients believe they were growing rich, only to cause them ultimately a loss of more than $50 billion. Madoff is one of the most justly hated individuals in the United States.
In contrast to the $50 billion of losses caused by Madoff, the losses caused by the Federal Reserve System and those in charge of it amount to trillions of dollars, probably to more than $10 trillion if the stock and real estate bubbles are taken together. Instead of affecting thousands of people as in the case of Madoff, tens of millions have been made to suffer hardship. Indeed, practically everyone has been harmed to some extent by what the Federal Reserve has done: the owners of stocks that have plunged, pensioners, the unemployed and their families, towns and cities suffering the consequences of business failures and plant closings.
It is difficult to imagine living with the knowledge that one is personally responsible for such massive destruction. Such knowledge might easily drive someone to suicide or at least to some means, such as drink or drugs, of not having to allow it into consciousness.
Alan Greenspan, who was Chairman of the Federal Reserve’s Board of Governors from 1987 to 2006, the period encompassing both bubbles, is clearly the single individual most responsible for the bubbles. The present Chairman, Ben Bernanke, also bears substantial responsibility, though not to the same extent as Greenspan. While Chairman only since January of 2006, Bernanke has been a member of the Federal Reserve Board since 2002. Thus he was present in a major policy making position during most of the housing bubble and crucial years leading up to it.
Neither Greenspan nor Bernanke have resorted to drink or drugs to conceal their responsibility from themselves. Instead they have resorted to specious claims about the cause of the bubbles, the housing bubble in particular.
One can read through their widely disseminated public statements and not find a single explicit reference to credit expansion and fiduciary media, nor to malinvestment and overconsumption. To avoid recognition of any need to discuss these phenomena, Greenspan seems to have wiped his mind clean of all knowledge of how Federal Reserve interest-rate policy affects interest rates in the economic system.
In what appears to be his closest reference to credit expansion, he wrote, in an article in The Wall Street Journal of March 11, 2009:
In these passages Greenspan invents a version of the opposition to Federal Reserve sponsored credit expansion that no opponent of credit expansion or “easy money” has ever held. No opponent of credit expansion has ever claimed that reductions in the federal-funds rate need directly affect long-term interest rates. To the contrary, the significance of reductions in the federal-funds rate is that what is required to bring them about in the actual market for those funds is an increase in member-bank reserves. The increase in those reserves is then the foundation of credit expansion to a vast multiple of the additional reserves. That credit expansion is what then serves to lower long-term interest rates, such as mortgage rates.
There are at least two broad and competing explanations of the origins of this crisis. The first is that the "easy money" policies of the Federal Reserve produced the U.S. housing bubble that is at the core of today's financial mess.
The second, and far more credible, explanation agrees that it was indeed lower interest rates that spawned the speculative euphoria. However, the interest rate that mattered was not the federal-funds rate, but the rate on long-term, fixed-rate mortgages. Between 2002 and 2005, home mortgage rates led U.S. home price change by 11 months. This correlation between home prices and mortgage rates was highly significant, and a far better indicator of rising home prices than the fed-funds rate.
This should not come as a surprise.
After all, the prices of long-lived assets have always been determined by discounting the flow of income or imputed services by interest rates of the same maturities as the life of the asset. No one, to my knowledge, employs overnight interest rates—such as the fed-funds rate—to determine the capitalization rate of real estate, whether it be an office building or a single-family residence.
The way the process works is as follows. To actually achieve the lower federal-funds rate that it announces as its target, the Federal Reserve goes into the market and buys government securities from banks or the customers of banks. It pays for those securities by means of the creation of new and additional standard money. When the Fed purchases securities from banks, the banks directly and immediately have equivalently more reserves in their possession. When it purchases securities from the customers of banks, the banks gain equivalently more reserves as soon as those customers deposit the checks they have received that are drawn by the Fed on the Fed. These checks are then forwarded to the Fed and the reserve accounts of the banks in question are equivalently increased.
Depending on the amount of their increase, the immediate effect of the additional reserves is to reduce or eliminate deficiencies in the required reserves of some, many, or all of the banks that have had such deficiencies, to replace deficiencies of reserves with excesses of reserves, and to increase the excess reserves of some, many, or all of the banks that have had excess reserves. The effect of this in turn is to reduce the demand for federal funds, i.e., funds that qualify as reserves, while increasing their supply. This combination is what brings down the federal-funds rate in the market for federal funds.
What is far more significant is that the creation of new and additional excess reserves by the Fed—reserves beyond the amount legally required to be held—places the banking system in a position in which it can expand the supply of checking deposits and thus fiduciary media to a multiple of the additional reserves. And thanks largely to Mr. Greenspan that multiple came to be enormous. By December of 2005, it exceeded 126 times. Two years later, it exceeded 160 times.
Thus for each dollar of additional excess reserves created, a credit expansion was made possible on the order of a vast multiple. The new and additional fiduciary media corresponding to the credit expansion were the source of the funds for stock purchases in the stock market bubble and for housing and commercial real estate purchases in the housing bubble. Their pouring into the home mortgage market was what drove down mortgage interest rates. Between December of 1999 and December of 2005, almost $1.7 trillion of new and additional fiduciary media were created and lent out.
As market interest rates started rising in the second half of 2004 and then through 2005, increasing amounts of deposits earning a modest rate of interest and on which checks could be written, came to be used more and more as checking accounts rather than savings accounts. They were drawn into the spending stream in response to the higher comparative rates of return that could be earned through investment in securities. This allowed the life of the housing bubble to be extended until 2006.
The Saving Glut Argument
Along with denying the causal role of Federal Reserve expansionary monetary policy in the housing bubble, Greenspan advances the claim, greatly elaborated by Bernanke, that what was actually responsible for the bubble was an excess of global saving. He argues in his Wall Street Journal article that
[T]he presumptive cause of the world-wide decline in long-term rates was the tectonic shift in the early 1990s by much of the developing world from heavy emphasis on central planning to increasingly dynamic, export-led market competition. The result was a surge in growth in China and a large number of other emerging market economies that led to an excess of global intended savings relative to intended capital investment. That ex ante excess of savings propelled global long-term interest rates progressively lower between early 2000 and 2005.In a series of lectures beginning in March of 2005 and continuing into the current year, Bernanke elaborates on this claim. At a lecture given at the Bundesbank in Berlin, Germany, on September 11, 2007, titled “Global Imbalances: Recent Developments and Prospects,” he argued that stepped up saving in developing countries was largely responsible for “the substantial expansion of the current account deficit in the United States, the equally impressive rise in the current account surpluses of many emerging-market economies, and a worldwide decline in long-term real interest rates.” (For the benefit of non-technical readers, the “current account” balance encompasses the difference between exports and imports both of goods and services, the difference between incomes earned abroad and incomes paid to abroad, plus the difference between remittances from and to abroad.)
These developments, he held, “could be explained, in part, by the emergence of a global saving glut, driven by the transformation of many emerging-market economies—notably, rapidly growing East Asian economies and oil-producing countries—from net borrowers to large net lenders on international capital markets.”
In a speech delivered on April 9 of this year at Morehouse College in Atlanta, Bernanke stressed that “the net inflow of foreign saving to the United States, which was about 1-1/2 percent of our national output in 1995, reached about 6 percent of national output in 2006 an amount equal to about $825 billion in today's dollars.” He then proceeded to blame the housing boom on this inflow of foreign savings. “Financial institutions,” he declared, “reacted to the surplus of available funds by competing aggressively for borrowers, and, in the years leading up to the crisis, credit to both households and businesses became relatively cheap and easy to obtain. One important consequence was a housing boom in the United States, a boom that was fueled in large part by a rapid expansion of mortgage lending.”
Thus, according to Bernanke, it was not credit expansion or anything that he and the Federal Reserve System and Mr. Greenspan were responsible for, but the inflow of foreign savings. That inflow, representing a “global saving glut,” was responsible for the bubble and its aftermath.
Bernanke uses the expression “saving glut” repeatedly: 9 times in his lecture at the Bundesbank in September of 2007, 11 times in his lecture at the Virginia Association of Economics in March of 2005, and 10 times in his Homer Jones Lecture in St. Louis in April of 2005. Despite his constant repetition of the claim, it turns out to have absolutely no substance. Nowhere is the existence of anything remotely approaching a saving glut in any way substantiated.
The Non-Existence of a Saving Glut
The very notion of a saving glut is absurd, practically on its face. As I wrote in Capitalism:
Before the scarcity of capital … could be overcome, capital would have to be accumulated sufficient to enable the 85 percent of the world that is not presently industrialized to come up to the degree of capital intensiveness of the 15 percent of the world that is industrialized. Within the industrialized countries, capital would have to be accumulated sufficient to enable every factory, farm, mine, and store to increase its degree of capital intensiveness to the point presently enjoyed only by the most capital-intensive establishments, and, at the same time, to enable all establishments to raise the standard of capital intensiveness still further, to the point where no further
reduction in costs of production or improvement in the quality of products could be achieved by any greater availability of capital…. 
Long before such a point could ever be reached, time preference would put an end to further increases in the degree of capital intensiveness.
It is doubly absurd to believe that the source of a saving glut would be precisely countries possessing very little capital compared to the United States and other industrialized countries. But that is what Bernanke claims. He claims that countries such as Thailand, China, Russsia, Nigeria, and Venezuela are the source of the alleged saving glut.
There are further theoretical considerations that argue specifically against any form of “saving glut” being responsible for the housing bubble.
First, if saving had been responsible, and not credit expansion and the increase in the quantity of money, then the additional saving taking place in the countries providing it, would have been accompanied by a reduction in consumer spending in those countries. People would have had to spend less for consumption in those countries, in part, in order to make available funds for additional spending on capital goods that were exported to the United States. Such export of capital goods to the US would not have fueled a boom here. To the contrary, it would have resulted in lower prices of capital goods in the US. Only the portion of funds saved that was used to finance purchases within the US could have contributed to any higher prices of capital goods and land in the US. And, of course, whatever rise in the prices of capital goods and land that might have taken place in the US would have tended to be matched by a fall in the prices of consumers’ goods in the countries that had stepped up their saving. The only way that the demand for capital goods and land could rise without the demand for consumers’ goods falling would be on the strength of an increase in the quantity of money and the total, overall volume of spending in the economic system.
Indeed, the fact that in the absence of an increase in the quantity of money and volume of spending in the economic system, shifts in spending serve to reduce prices as much as increase them has a parallel in the further fact that increases in the relative size of some of the countries in the world’s economy imply equivalent decreases in the relative size of other countries in the world’s economy. In the absence of an increase in the quantity of money and volume of spending, growth in the relative size of the economies of many Asian countries would not by itself be sufficient for greater saving in those countries serving to increase global spending for capital goods. For that greater spending would be accompanied by reduced spending for capital goods in other countries, i.e., countries that were already in the category of developed economies and now had to yield some portion of their previous relative size.
In the present instance, what this means is that greater spending for capital goods and land in the US, financed by saving in parts of Asia, would be accompanied by less spending for capital goods in the US (and possibly elsewhere) financed by saving in the US or financed by saving elsewhere in the world. If spending for capital goods financed by saving in Asia is not accompanied by reduced spending for capital goods financed by saving elsewhere, the only ultimate explanation is an increase in the quantity of money and volume of spending in the world’s economy. Of course the source of such an increase in today’s conditions is none other than the Federal Reserve System.
Second, contrary to popular understanding, when saving is divorced from the increase in the quantity of money and volume of spending, and takes place without such increase, it does not tend to grow larger from year to year. Nor does consumer spending tend to decrease from year to year. And thus more saving would not serve to raise the prices of capital goods or land from one year to the next. Its effect would essentially be limited to a discrete, one-time only increase. Yet for the prices of capital goods and land to rise from one year to the next on the strength of an increase in the demand for capital goods and land based on an increase in saving, the increase in saving would have to become progressively larger from year to year. And this would mean that the demand for consumers’ goods would have to become progressively smaller from year to year.
For example, imagine that at the expense of an equal fall in the demand for consumers’ goods, the demand for capital goods rose by some given amount, say, 100. This 100 can represent however many billions or hundreds of billions of dollars as may be required to make it realistic in terms of present spending levels. In such circumstances, there would be nothing present that would make the prices of capital goods or land any higher in the second and later years of 100 of additional such spending than in the first year.
Indeed, as the years wore on, the increases in production achieved by a greater supply of capital goods would start reducing prices, including the prices of capital goods themselves, as the supply of capital goods itself was increased on the foundation of a general increase in production. Even land prices would fall to the extent that improvements in the supply of capital goods permitted the adoption of methods of production that allowed the economical use of previously submarginal land or so increased the output per unit of land as to make part of its supply redundant.
In circumstances of an unchanged supply of money and demand for money for holding, each act of greater saving and accompanying greater expenditure on capital goods operates in a manner analogous to the relationship between force and acceleration in the physical world. In the physical world, in the conditions of a friction-free environment, a single application of force to an object imparts continuous motion at a constant velocity. Similarly, in the economic world, in the conditions of an unchanged quantity of money and volume of spending, each act of reduced expenditure for consumers’ goods and increased expenditure for capital goods, causes the economic system to adopt a greater relative concentration on the production of capital goods and a reduced relative concentration on the production of consumers’ goods. This produces an inertial effect on capital accumulation.
The first result of the greater relative concentration on the production of capital goods is a greater production of capital goods, alongside a smaller production of consumers’ goods. These additional capital goods, however, obtained on the foundation of additional saving, are the basis of an increase in the ability to produce both consumers’ goods and further capital goods. That is to say, the additional capital goods make possible a general increase in production, an increase in the production of consumers’ goods and a further increase in the production and supply of capital goods as well. The process of an increasing supply both of consumers’ goods and capital goods, based on the foundation of a single fall in consumption and increase in saving, can go on indefinitely if it is accompanied by further scientific and technological progress. In these circumstances, a further fall in the demand for consumers’ goods and rise in the demand for capital goods would be analogous to a further application of force to an object and would result in an acceleration of the increase in production.
A further point must be mentioned here. And that pertains to the durability of capital goods and its implications for capital accumulation, saving, and spending. Thus, if the average life of the capital goods in our example of 100 of additional spending for capital goods were, say, 10 years, then a diminishing process of saving would go on for 10 years with no further fall in the demand for consumers’ goods nor rise in the demand for capital goods. Net saving and equivalent net investment in the economic system would take place in a pattern 100, 90, 80, …10, as the 100 of additional spending for such capital goods was accompanied by successive increases in annual depreciation charges. The additional depreciation charges would be 10 in the year following the first year’s expenditure of an additional 100 for such capital goods. In the next year, when there were two such batches of capital goods, depreciation would be 20. At the end of the tenth year, the depreciation charges on ten such batches of capital goods would be 100, and net saving and net investment would disappear, unless, of course, there were a further decline in consumption expenditure and increase in demand for capital goods.
What is particularly important to realize here is that the net saving of years 2 through 10 would not serve at all to raise the demand for capital goods and land nor their prices, but would contribute to the supply of capital goods being larger, production in general consequently being greater, and prices in general, including the prices of capital goods, being lower as a result. Such results, and those of the process of saving and capital accumulation in general that were described a moment ago, cannot be reconciled with the conditions of a bubble. They should not be cited as the basis of explaining a bubble.
Third, if somehow saving were responsible for the housing bubble, why did it suddenly collapse? Why did people suddenly stop saving and stop making funds available for the purchase of homes? Obviously, the explanation was that the bubble did not depend on saving but on credit creation and its acceleration and that when the ability to create sufficiently more credit came to an end, the props supporting the bubble were removed and it collapsed.
Fourth, if saving were responsible for the bubble, why have banks and countless other firms found themselves confronting an acute lack of capital? Saving provides new and additional capital. How can it be that an alleged process of saving has resulted in widespread major capital deficiencies? This situation of insufficient capital is the result of malinvestment and overconsumption, which are the consequences of credit expansion, not saving.
Fifth, if saving had been responsible for the increase in spending on capital goods and land, the rate of profit would have modestly fallen from the very beginning, and continued its fall until net saving came to an end. It would not have risen, let alone risen dramatically, as it did during the bubble.
This is the implication of the discussion, above in this section, of the second reason why saving was not responsible for the bubble. In particular it is the implication of the example of 100 more of spending for capital goods financed by 100 of saving derived from 100 less of spending for consumers’ goods. In that example, in which there is no increase in the quantity of money or total volume of spending, the global economic system would have had the same total aggregate business sales revenues, with the sales revenues coming from the sale of consumers’ goods diminished by the amount of saving, and those coming from the sale of capital goods equivalently increased. At the same time, however, it would have had a tendency toward a rise in the aggregate costs of production deducted from those sales revenues.
The rise in costs would have been the result of such things as additional depreciation charges on the new and additional capital goods purchased, or additional cost of goods sold following additional purchases of materials and labor on account of inventory. In the example of 100 more being spent for capital goods each year with an average life of 10 years and accompanying depreciation charges in the respective amounts of 10, 20, …, 100 in the 10 years following the rise in demand for capital goods, aggregate profit in the economic system would have been falling year by year by an amount equal to the increase in depreciation.
A falling aggregate amount of profit together with the increasing amount of capital invested in the economic system, would have progressively reduced the economy-wide average rate of profit. It would have been a case of a falling amount-of-profit numerator divided by a rising-amount-of-capital denominator.
Totally contrary to what one would expect from these effects of a rise in saving, the reality, of course, was a sharply higher average rate of profit in the economic system so long as the bubble lasted. This can be explained only on the foundation of credit expansion and an expanding quantity of money and volume of spending, not on the basis of saving.
If none of these five reasons are sufficient to dispel the notion that a saving glut was responsible for the bubble, then hopefully it will be sufficient to point out that there simply was no saving glut, but rather only a very modest rate of saving, a mere trickle of saving. For it turns out that over the 13 year period 1994-2006, the rate of saving in the US, together with all foreign saving entering the country in connection with deficits in the current account, never exceeded 7 percent, and in 8 of those 13 years was 3 percent or less. In 5 of those years it was a mere, 1 or 2 percent. And what is of special significance is that in the years of the housing bubble, 2002-2006, it was especially low: 2 percent in 2002, 1 percent in both 2003 and 2004, 3 percent in 2005, and 4 percent in 2006.
To see this result, it is necessary to begin by removing all fictional elements in the reported amounts of domestic net saving and GDP. These fictional amounts consist of various “imputations.” The leading imputations that are relevant here are those that arbitrarily convert what is in fact consumption expenditure into investment expenditure. These have the effect of reducing reported consumption and equivalently increasing reported saving., 
The two most important such imputations are these: 1) the treatment of the purchase of single family homes that the buyer intends to occupy and that thus will not be a source of any money revenue of income to him, as though they were nonetheless income producing assets and therefore represented an investment; 2) the treatment of government expenditure for fixed assets such as buildings, as though it were an investment expenditure rather than a consumption expenditure.
When such imputations are removed from the calculation of net saving and from GDP, the very modest extent of saving that has been going on over the last decade or more is clearly shown. Indeed, since 2002, domestic net saving has been negative to the extent of several hundred billion dollars each year.
The following table describes the situation:
The table has 6 columns. Column 1 lists the years 1994 through 2006, the period encompassing both the stock market and the real estate bubbles. Column 2 shows the current account deficit in those years. This deficit is taken as representing the foreign savings coming into the United States. (For this reason it is shown as a positive number.) Column 3 shows net saving in the United States in those years when such savings are calculated free of imputations. Column 4 is the sum of Columns 2 and 3. It shows total saving in the United States as the sum of foreign saving entering the country together with domestic saving. Column 5 is GDP year by year, with all imputations removed. Column 6 is the sum of imputation-free foreign and domestic saving divided by such GDP, presented in decimal format.
The notion that there was a saving glut behind the housing bubble is simply a fiction. Its proponents could manufacture as much of a glut as they like simply by reclassifying such things as expenditure for automobiles, major appliances, furniture, and clothing as investment expenditures, on the grounds that these goods too are durable, like houses. That would equivalently reduce consumption expenditure and increase reported saving in the economic system.
Bernanke and Greenspan et al. focus on deficits in the current account as representing the counterpart of foreign saving and investment, which they believe must be present to finance the deficits. There is certainly a very close relationship between foreign saving and investment on the one side and the financing of deficits in the current account on the other. The following example may help to highlight this relationship.
Thus imagine Saudi Arabia back in the days when geologists had determined that the country possessed vast oil reserves but before it had any oil wells, pipelines, refineries, or facilities for the handling of supertankers. Those things had yet to be built.
Now how could those facilities be built? The only way was by means of the arrival of shiploads of equipment and construction materials from Europe and the United States. In addition, large quantities of various consumers’ goods were required for the foreign engineers and other workers who were required to carry out the construction. All these goods coming into Saudi Arabia were imports of foreign goods. But Saudi Arabia had hardly anything to export before its ability to produce oil was developed. Thus, in the interval, there was a massive excess of imports over exports. That excess represented foreign investment in Saudi Arabia. Its physical form was all of the facilities under construction and then, ultimately, the completed facilities for producing oil.
Foreign investment very often, perhaps most of the time, has this kind of close connection to the existence of an excess of imports over exports and, more broadly, an excess of outlays of all kinds on current account over receipts of all kinds on current account. (As previously explained, the balance on current account includes not only the difference between the imports and exports of goods, but also of services. In addition, it includes the difference between incomes paid to abroad and incomes paid from abroad, and finally, the difference between remittances to and from abroad.)
Nevertheless, it should be realized that the essential, core concept of the current account, namely, the so-called balance of trade, which is the difference simply between the import and export of goods, was developed long before the emergence of any significant international investment. It was developed and employed by a school of writers known as the mercantilists, who were current from the 16th to the third quarter of the 18th Century, when the school was laid to rest by Adam Smith.
The main concern of the mercantilists was the accumulation of gold and silver within the borders of their country and the prevention of any loss of gold or silver by their country. Gold and silver were the money of the day everywhere and, it was believed, needed to be accumulated within the country in order to be available if and when the government might need them, in order to finance military operations outside the country or any other activities in which circumstances might operate to draw precious metals away from the country.
Inasmuch as already by that time, most of the European countries had no gold or silver mines within their territory, the only way they could gain gold or silver was by means of the export of goods. The import of goods was seen as constituting a loss of gold or silver by the country. Accordingly, the goal of mercantilist policy was to maximize exports while minimizing imports. That would allegedly ensure the greatest possible accumulation of the precious metals within the country.
Centuries later, in the chapter “On Foreign Trade” in his Principles of Political Economy and Taxation, Ricardo developed the principle that the supply of the precious metals tends to be distributed among the different countries essentially in proportion to the relative size of their respective economies. He wrote: "Gold and silver having been chosen for the general medium of circulation, they are, by the competition of commerce, distributed in such proportions amongst the different countries of the world as to accommodate themselves to the natural traffic which would take place if no such metals existed, and the trade between countries were purely a trade of barter."
The operation of this principle can, of course, be modified by the operation of other principles working alongside it. Thus a country with a relatively small economy, but with an exceptional reputation for the security of property and the enforcement of contracts, might well have a quantity of money within its borders far in excess of what corresponded to the relative size of its economy. By the same token, countries with larger economies but in which property rights and the enforcement of contracts were in retreat, could possess a proportion of the world’s money supply substantially less than what corresponded to the relative size of its economy.
It follows from Ricardo’s principle that countries with gold and silver mines will experience a chronic excess of imports over exports. The gold and silver that they mine cannot all be retained within their borders. If they were retained, the country would have a disproportionately large supply of the precious metals. This would serve to raise prices in that country relative to prices abroad. The effect would be an outflow of the precious metals until their buying power at home did not fall short of their buying power abroad by more than the costs of shipping them abroad.
Today, the US dollar is in a position similar to that of gold under an international gold standard. The dollar is a virtual world money—not completely, but substantially. The United States is the country with the “dollar mines.” When dollars are created in the US, a substantial portion of them will flow abroad. And this applies not just to currency, but also to checking deposits and all other short-term financial instruments easily convertible to currency.
Most of the dollars that “flow abroad” need not actually circulate abroad but to a large extent serve as mere precautionary holdings of money, and, to an important extent, as reserves for financial institutions that create various moneys other than dollars. These other moneys that are created on the foundation of additional dollars circulate abroad.
Now the fact that the United States compared to almost all other countries in the world still has the most reliable protection of property rights and enforcement of contracts, is responsible for the fact that much or most of the money that “flows abroad” does not in fact leave the country. Rather it passes into the ownership of foreign individuals, firms, and governments who continue to hold it within the United States.
The increase in such foreign owned assets within the United States has the appearance of foreign investment. Actually, it is nothing more than the by-product of credit expansion and the increase in the quantity of money within the United States.
There is no genuine surge in foreign saving. There is domestic credit expansion and money supply increase that serves to increase imports and shift ownership of a substantial portion of the additional money supply, and short-term claims to money, to foreigners.
Ironically, Bernanke himself helps to confirm this interpretation of the increase in the current account deficit. He says: “First, the financial crises that hit many Asian economies in the 1990s led to significant declines in investment in those countries in part because of reduced confidence in domestic financial institutions and to changes in policies—including a resistance to currency appreciation, the determined accumulation of foreign exchange reserves, and fiscal consolidation—that had the effect of promoting current account surpluses.” (Bundesbank Lecture, Berlin, Germany, September 11, 2007.)
What Bernanke describes here is not any sudden increase in foreign saving but rather decisions to change the way in which a portion of previously accumulated savings are held, i.e., to hold them to a greater extent in the form of US dollars and short-term claims to dollars.
In the same passage, Bernanke presents a second reason for the alleged growth in foreign savings, namely the sharp increase in the price of oil that had taken place. He says, “sharp increases in crude oil prices boosted oil exporters' incomes by more than those countries were able or willing to increase spending, thereby leading to higher saving and current account surpluses.”
Here, Bernanke overlooks the role of credit expansion and the increase in the quantity of money in bringing about the higher price of oil. He also overlooks the effect of the higher price of oil on the real incomes and ability to save of everyone who had to pay that higher price.
The role of credit expansion and the increase in the quantity of money in causing the rise in oil prices was confirmed by the subsequent plunge in oil prices once credit expansion was brought to an end and appeared to be about to turn into massive credit contraction. It has since been further confirmed by the recent rise in oil prices following the growing belief that the government’s program of renewed credit expansion will be sufficient to eliminate the danger of a financial collapse and will serve to maintain and increase the demand for oil.
Net Saving as a By-Product of the Increase in the Quantity of Money
My discussion of the fallacy of a saving glut as being responsible for the housing bubble and its aftermath would not be complete if I did not point out that the continued existence of net saving is itself a by-product of the increase in the quantity of money and volume of spending in the economic system. In the absence of increases in the quantity of money and volume of spending, economy-wide, aggregate net saving would tend to disappear. It would cease when total accumulated savings came to stand in a ratio to current incomes and consumption that people judged to be sufficiently high that they had no further need to make still greater relative provision for the future.
What keeps net saving in existence is that the increase in the quantity of money and volume of spending tends continually to raise incomes and consumption in terms of money. In order to maintain any given ratio of accumulated savings to a rising level of income and consumption, it is necessary to increase the magnitude of accumulated savings. At the same time, the increasing quantity of money provides the financial means of spending more and more each year for capital goods as well as consumers’ goods and for thus maintaining the desired balance in the face of growing magnitudes of spending.
Thus it is the increase in the quantity of money and the volume of spending that it supports that is responsible for net saving continuing in being. In the absence of the continuing increase in the quantity of money, net saving would disappear, and capital accumulation would take place simply by means of a continually increasing purchasing power of the same capital funds. That growing purchasing power would be created by the increase in the production and supply of capital goods and the fall in prices of capital goods that would result.
Summary and Conclusion
The real estate bubble, like the stock market bubble before it, was caused by credit expansion. The credit expansion was instigated and sustained by the Federal Reserve System, which could have aborted it at any time but chose not to. As a result, the Federal Reserve System and those in charge of it at during the real estate bubble bear responsibility for major harm to tens of millions of Americans.
In order to avoid having to accept this responsibility, a specious doctrine has been advanced by Alan Greenspan and Ben Bernanke, the former and present Chairman of the system, and others. That is the doctrine of a “global saving glut.” Not credit expansion but the saving glut was responsible, they claim.
The truth is that time preference puts an end to further saving long before it could outrun the uses for additional saving. This makes a saving glut impossible. In addition, there are five major reasons why saving could not have been responsible for the real estate bubble in particular. First, if saving had been responsible, rather than credit expansion and the increase in the quantity of money, there would have been a corresponding decline in consumer spending in the countries allegedly doing the saving. The fact is that there was no such decline.
Second, saving implies a growing supply of capital goods, more production, and lower prices, including lower prices of capital goods and even of land. These are results that are incompatible with the widespread increases in prices typically found in a bubble.
Third, if somehow saving had been responsible for the housing bubble, the spending it financed would not suddenly have stopped. Such stoppage is a consequence of the end of credit expansion and the revelation of a lack of capital.
Fourth, if large-scale saving rather than credit expansion had been present, banks and other firms would have possessed more capital, not less. They would not be in their present predicament of having inadequate capital to carry on their normal operations. This situation of insufficient capital is the result of malinvestment and overconsumption, which are the consequences of credit expansion, not saving.
Fifth, in the absence of increases in the quantity of money and overall volume of spending in the economic system, saving also implies an immediate tendency toward a fall in the economy wide average rate of profit. This is another result that is incompatible with what is observed in a bubble or boom of any kind, which is surging profits so long as “the good times” last.
Especially noteworthy is the fact that in the real estate bubble, there simply was no saving glut. In the 13 year period 1994-2006, the rate of saving in the US, together with all foreign saving allegedly entering the country in connection with deficits in the current account, never exceeded 7 percent, and in 8 of those 13 years was 3 percent or less.
What has served to conceal how low the actual rate of saving has been is the fact that major fictional items have been counted in saving, which add hundreds of billions of dollars every year to its reported amount. The most notable instance is that purchases of single family homes that the buyers intend to occupy and that will thus not be a source of any money revenue or income to them, are treated as though they were nonetheless purchases of income producing assets and therefore represented an investment. Similarly, government spending on account of buildings and structures is treated as investment. Such overstatement of investment correspondingly understates consumption expenditure in the economic system. And when the artificially reduced amount of consumption is subtracted from any given amount of national income or GDP, saving appears to be equivalently larger.
The alleged saving entering the American economy via deficits in its current account is in fact largely not saving at all, but the by-product of US credit expansion and money supply increase. Dollars today are a virtual global money. And in conformity with Ricardo’s principle concerning the distribution of the precious metals throughout the world based on the relative size of the economies of the various countries, most of the additions to the supply of dollars and short-term claims to dollars cannot remain in the possession of Americans but must gravitate into the ownership of foreigners. This creates a deficit in the balance of trade and in the whole of the so-called current account. While it may appear that increased foreign holdings of dollars and short-term dollar-denominated securities represent foreign investment, the truth is that much or possibly even all of the alleged foreign saving entering the United States is nothing other than a consequence of US credit expansion and money supply increase.
Finally, net saving itself, as a continuing phenomenon is nothing more than a by-product of the increase in the quantity of money, in that it would come to an end if the money supply were to stop increasing.
The conclusion to be drawn is that the housing bubble was indeed the product of credit expansion, not a “saving glut.”
 David R. Henderson and Jeffrey Rogers Hummel, Greenspan’s Monetary Policy in Retrospect, Cato Institute Briefing Paper 109, Cato Institute, Washington, D.C., November 3, 2008, pp. 4f.
 George Reisman, Capitalism: A Treatise on Economics (Ottawa, Illinois: Jameson Books, 1996) p. 512.
 This figure is arrived at by taking the sum of M1, sweep accounts, money market mutual fund accounts both retail and institutional, and one half of savings deposits as the measure of money market deposit accounts, the data for which are apparently otherwise unavailable. The same procedure is used as the basis of all other statements of the money supply or changes in the money supply.
 Italics in original.
 Reisman, Capitalism, p. 57.
 “Homer Smith Lecture,” St. Louis, MO, April 14, 2005.
 For a comprehensive explanation of the role of the quantity of money in determining the volume of spending in the economic system, see Reisman, Capitalism, chaps. 12 and 19.
 For an explanation of the role of saving in capital accumulation, see ibid., pp. 621-642.
 For a thoroughgoing discussion of the determinants of the rate of profit and its relationship to saving and capital accumulation, see ibid., chaps. 16 and 17.
 For a comprehensive explanation of the distinction between capital goods and consumers’ goods and investment or, better, productive expenditure and consumption expenditure, see ibid., pp. 445-456.
 For a detailed critique of the imputed income doctrine, see ibid., pp. 456-459.
Copyright © 2009 by George Reisman. George Reisman, Ph.D. is the author of Capitalism: A Treatise on Economics (Ottawa, Illinois: Jameson Books, 1996) and is Pepperdine University Professor Emeritus of Economics. He is also a Senior Fellow at the Goldwater Institute. His web site is www.capitalism.net and his blog is www.georgereisman.com/blog/. A pdf replica of his book can be downloaded to the reader’s hard drive simply by clicking on the book’s title, above, and then saving the file when it appears on the screen.