20.8 The boom causes the bust

Credit booms tend to produce asset bubbles. Investors borrow from banks to buy land, shares, or other assets, not for the income those assets can produce but because they expect to sell them later at a higher price. They are trying to make money without producing anything: essentially, they are gambling on prices with borrowed money. Their purchases push prices higher, the apparent gains attract more borrowers, and the rising assets can be pledged as collateral for still larger loans.

Banks have no difficulty supplying the money for this process. They do not need to find an existing saver for each loan; a commercial bank normally creates a new deposit when it extends credit, increasing the money supply. The asset bubble and the money used to inflate it can therefore grow together.

Under VoTa, which has sound money, lending transfers existing money from a saver to a borrower. Speculators must compete for a limited supply of genuine savings, every loan has an opportunity cost, and purchasing power cannot multiply merely because asset prices have risen. The extreme, self-financing asset bubbles produced by bank credit creation would be impossible.

The Austrian explanation of the business cycle is adopted here. Credit creation suppresses interest rates and allows investments to proceed without corresponding real savings. It does not create the labour, materials, or capital that all of those investments require, so many projects that appear profitable during the boom cannot be completed or sustained. These are malinvestments. Once they have been made, a bust is inevitable because the claims created during the boom cannot all be satisfied from the available real resources.

The economic failure therefore occurs during the boom, not the bust. The bust merely reveals and corrects the malinvestment that has already occurred (see Hayek's Prices and Production).

The Mississippi Bubble

John Law's financial system in early eighteenth-century France joined money creation, government debt, and asset speculation. His bank issued paper money while his Mississippi Company acquired trading monopolies and assumed much of the government's debt. New company shares could be bought for only 75 livres down on an issue price of 500 livres, with the balance paid in instalments, while the bank created notes and supported the share price. The apparent success of the company and the money used to buy its shares were therefore produced by the same system (see the Federal Reserve Bank of New York's The Mississippi Bubble of 1720 and the European Debt Crisis).

Mississippi Company shares rose from about 500 livres at the beginning of 1719 to approximately 10,000 livres. Over roughly the same period, the bank's note issue expanded from about 40 million livres to almost one billion, an increase of approximately twenty-five times, while the share price increased twentyfold. The note issue is not a modern measure of a bank loan book, but it measures the new bank money available to finance and support the speculation. The company's paper value reached an estimated twice the annual output of France (see The Man Who Invented Money). The twentyfold share-price increase did not represent a twentyfold increase in France's productive resources; it demonstrated what newly created purchasing power could do to the price of a favoured asset.

Japan's asset bubble

Japan's asset bubble of the late 1980s illustrates how credit creation can make a boom self-reinforcing. Bank lending raised land prices; more valuable land could then be pledged as collateral for larger loans, and the additional credit financed further purchases. Bank of Japan research found that the initial rise in asset prices was caused by a sharp increase in lending to the real-estate sector and that the interaction between land values and collateral-constrained credit helped link the land and stock-market booms (see Explaining Asset Bubbles in Japan).

The expansion can be seen directly in bank lending. BIS figures show that Japanese bank lending for real-estate purposes grew at an average annual rate of 19 percent between 1983 and 1990, implying that it more than tripled over seven years. The composition of bank balance sheets shifted with it. From 1982 to 1989, direct real-estate loans rose from approximately 5.5 percent to more than 12 percent of all city-bank loans and from 6 percent to 11 percent of regional-bank loans. Over the same period, the share of city-bank loans secured by land increased from 17 percent to more than 24 percent. These measures come from bank loan portfolios rather than changes in land prices, so they independently record the flow of credit into the bubble (see BIS Economic Paper No. 34 and Did Investors Regard Real Estate as Safe During the Japanese Bubble?).

The resulting paper valuations were extraordinary. In 1988, Japan's theoretical land value was estimated at four times that of all land in the United States. A contemporary estimate valued Tokyo real estate at about US$8 trillion—enough, on paper, to buy all real estate in the United States together with every company listed on its stock exchanges (see Power from the Ground Up and Why Japan Could Take a Fall). These were extrapolated market valuations, not sums that could actually have been realised by selling the land, but that limitation is itself part of the point: prices had become detached from plausible economic value.

When land prices stopped rising, the collateral-dependent debts and investments created during the boom could no longer be sustained. The bust followed from the structure of the boom; it did not create the underlying failure.

Australia's 1880s land boom

Australia experienced the same process during the 1880s. Rapidly expanding bank lending, supplemented by capital from Britain, flowed into pastoral and urban property. New building societies and property-finance companies competed for business, lending standards deteriorated, and some of the so-called land banks both financed land and speculated in it themselves. The apparent prosperity included extraordinary construction and private investment, but these activities were being directed by prices inflated through credit (see the Reserve Bank of Australia's comparison of the 1890s and 1930s depressions).

The expansion was enormous relative to the economy. The RBA's historical series shows bank credit rising from approximately 40 percent of GDP around 1880 to about 70 percent around 1890, while the assets of all financial institutions rose from 55 percent of GDP in 1881 to 115 percent in 1891. The banks could not fund all of this lending from Australian deposits. From 1886 to 1891, their liabilities to British residents averaged 45 percent of their liabilities within Australia (see A Comparison of Indicators of Financial System Stability and the discussion of Three Australian Asset-price Bubbles).

Melbourne was the centre of the speculation. One estimate puts the average nominal return on Melbourne land at about 35 percent a year from 1880 to 1892. In Prahran, the average property price exceeded £1,000 in 1888; in Brighton it reached about £950. Collins Street sites reportedly rejected offers of £2,000 per foot of frontage during the boom, yet later failed to find buyers at £600 (see Three Australian Asset-price Bubbles). The subsequent banking collapse and depression revealed the failure, but the misallocation occurred while rapid credit growth, construction, and rising land prices were still being counted as signs of success.

The Great Depression

The stock-market boom preceding the Great Depression produced striking examples of prices becoming detached from plausible returns. The Dow Jones Industrial Average rose sixfold, from 63 in August 1921 to 381 in September 1929. Radio Corporation of America, the era's celebrated new-technology company, rose from about US$1.50 in 1921 to US$549 in 1929. At its peak it traded at approximately 73 times earnings despite paying no dividends. The Goldman Sachs Trading Corporation, an investment trust issued at US$104 in December 1928, more than doubled from that price within three days during 1929; adjusted for splits, its shares were worth US$1.75 by mid-1932 (see Stock Market Crash of 1929, the BIS paper The Great Depression as a Credit Boom Gone Wrong, and Goldman Sachs Trading Corporation Bears Full Brunt of 1929 Crash).

These prices, the use of borrowed money to buy shares on margin, and the rapid creation of leveraged investment trusts provide direct evidence of speculative excess. Federal Reserve data show that total borrowing by New York stockbrokers rose from US$4.4 billion on 1 January 1928 to US$8.5 billion on 1 October 1929, nearly doubling in twenty-one months. The total is broader than bank lending alone: on 4 October 1929, US$1.885 billion came directly from banks and US$6.640 billion came from other sources, including corporations, individuals, private banks, and foreign banking agencies. All of it was credit secured by securities and used by brokers to carry inventories or provide margin loans to their customers. The figures were compiled by the Federal Reserve from reports by New York banks and the New York Stock Exchange (see Was There a Bubble in the 1929 Stock Market? and The Great Margin Call).

Credit creation financed the boom and created debts whose viability depended on inflated asset prices continuing to rise. Under a sound monetary system in which lending transfers existing savings rather than creating new money, credit could not have expanded in this way and the bubble and ensuing bust could not have reached the same scale. The failure occurred while the 1920s were still being celebrated as prosperous; the collapse exposed it.

The global financial crisis

The global financial crisis followed the same broad pattern in a different institutional form. Expanding mortgage credit enabled more spending on houses, rising house prices made borrowers appear wealthier and mortgages appear safer, and lenders could sell securitised loans into global markets and lend again. The official US inquiry into the crisis concluded that the growth of risky mortgages both fuelled the housing bubble and spread toxic loans through the financial system (see the Financial Crisis Inquiry Report).

Federal Reserve balance-sheet data show the scale of the expansion relative to the productive economy. US household mortgage debt rose from 61 percent of GDP in 1998 to 97 percent in 2006. This was an increase of 36 percentage points, or approximately 59 percent relative to its starting share, in eight years. The measure is outstanding household mortgage debt rather than the annual flow of newly issued mortgages, and expressing it as a share of GDP prevents ordinary economic growth from being mistaken for a credit boom (see The Great Recession and Its Aftermath).

When house prices stopped rising, the mortgages and securities premised on continued appreciation failed. The crisis revealed losses embedded throughout the financial system during the preceding credit boom.