When Ludwig von Mises’ treatise, “Human Action,” was published in 1949, Austrian Business Cycle Theory (“ABCT”) was focused on the supply-side (credit expansion distorting inputs to what the corporate finance sector would later quantify and label “WACC vs. IRR”). This was because the supply side – primarily, commercial lending and investment – was the primary channel through which central bank credit expansion flowed into the economy. Subsequent innovation in consumer credit has made ABCT more, not less, relevant, as interest rate fluctuations now both reduce apparent WACC and boost baseline sales (and, with baseline sales, projected IRR). The 2000s housing bubble is a recent example. As the below timeline illustrates, central bank credit expansion distorted the signals relied upon on both the supply and demand side, by lenders, investors, home-buyers, homeowners, builders and workers alike. This enabled and induced a series of mal-investments that often compounded each other,[i] causing the biggest boom-bust cycle of all time, the effects of which still dominate the financial world, and are still felt by tens of millions of people.
In Q101, Alice and Alan buy a home. A&A receive the median household income of $42,228.[ii] Applying a 33% payment-to-income ratio, A&A can afford payments of $1,173. FRM and ARM rates are equivalent,[iii] and like 90% of borrowers, A&A opt for an FRM. At 7%, A&A qualify for a loan of $176K. With 80% financing, their price point is $221K.
In Q203, Brenda and Ben buy a home. B&B receive the median household income of $43,318 – 2½% more than A&A’s income in 2001.[iv] Applying a 33% payment-to-income ratio, B&B can afford payments of $1,203 – only a $30 increase. But ARM rates have dropped 300-350 bps, despite a falling savings rate.[v] This has happened because the Federal Reserve has, through Open Market Operations, purchased, with fiat credit, securities on the open market, in order to implement its target overnight rate of 1%, a 425 bps drop.[vi] At an ARM rate of 3¾%,[vii] B&B qualify for a loan of $260K, translating into a price point, with 80% financing, of $324K. This is 46% higher than A&A’s price point just 9 quarters earlier. Borrowers take out ARMs in much greater numbers,[viii] and home prices, predictably, are bid up at the fastest rate in a generation.[ix]
With no prepayment penalty in the US, existing homeowners can refinance at the new rates and values. 15 million households refinance in 2003, versus 2.5 million in 2000.[x] Refinancing enables existing homeowners to reduce monthly payments or take cash out, at up to 80% loan-to-value, with “value” determined based on the price that B&B paid for a “comparable” home within the preceding twelve months.[xi] There is no limit to the number of times that one purchase can serve as the “comp” to support the valuation for a refinance, and there is a refinance “boom-let” whenever rates plunge. In 2002-2003, hundreds of millions of dollars of paper equity is created, then monetized, then used to pay down credit cards and fund home improvements and everyday consumption.[xii] As a result, total US home mortgage debt more than doubles, with over 1/3 of new originations structured as ARMs at artificially, and then-historically, low interest rates.[xiii] By Q2 2004, prior to, and without the aid of, any large-scale shift in underwriting or lending metrics, there is already a massive bubble that must end in a future day-of-reckoning, and it is entirely the result of central bank credit expansion.[xiv]
Despite the leveraging-up, the mortgage delinquency rate plummets – having never in decades fallen below 2% for more than two consecutive quarters, it falls below 2% in Q103 and remains below 2% through Q306.[xv] This is primarily the temporary result of cash-out refis (debt proceeds servicing debt) and temporarily lower payments on ARMs. Delinquency rates, collateral value trends and debt service ratios are the fundamentals used by banks to approve credit and determine lending standards, and in 2004-5 these ratios all appear to be moving in the right direction. Banks take these ratios at face value, failing to understand how the ratios have been temporarily distorted by credit expansion. Banks shift lending thresholds according to that misinterpretation. Credit spreads decline and ARMs now feature rates only 100-150 bps higher than comparable-tenor Treasury rates. They make more “exotic” loans[xvi] and more loans to borrowers with low credit scores.[xvii] RMBS investors read the same ratios and make the same misinterpretation and mal-investment as the banks.
Also responding to low rates and rising prices are homebuilders. Most private housing developments are financed at floating, Prime-based rates with interest-carry. Builders’ cost of capital is thus directly a function of Fed policy, and it has been reduced dramatically. Builders, like banks (and many economists), misinterpret rising home prices and home improvement spending as signals of increased real demand. Home construction and employment increase dramatically[xviii] as homebuilders’ “WACC vs IRR” inputs are distorted on both sides of the equation. Over a million workers who might otherwise have continued their education and/or pursued other jobs not related to the artificial boom (e.g., in technology) take construction jobs that will disappear in 2008 – a personal mal-investment of time and training.[xix] Countless others take jobs in other real estate-related fields. They make further personal decisions and obligations (marriages, children, mortgages, car loans) informed by their boom-time incomes. For builders and construction workers, the dynamic is a textbook example of ABCT, without any new or “exotic” features to cloud the matter.
After a few years of the asset price boom, the accompanying inflation in consumer and producer prices cannot be ignored, and from mid-2004 to 2007 the Fed gradually raises rates.[xx] The rest of the story is well-known: the spiral reverses itself. Home prices, no longer bolstered by credit expansion, stop rising. The increased pace of building produces excess supply, and now prices start to fall. ARMs re-set, generating payment shock. Troubled borrowers cannot refinance at higher rates and are too leveraged to sell at the new, lower prices. Delinquencies spike. “Underwater” borrowers become “strategic defaulters.” Delinquencies soar. RMBSs become worthless. Securities “derivative” of these underlying mortgages become worthless. Trading units leveraged up by the same math as the homeowners (flat income stream covers more debt) fail as rates rise and defaults increase – (financial firms’ leverage is high, but the crash has to happen regardless of leverage, as a result of the crash in the underlying mortgage market). Credit markets freeze. Major banks fail. Builders and other real-estate-related firms fail. Millions of workers, primarily in boom-related sectors, are laid off, and, having spent 4-5 years working construction, lack the skills needed to fill openings in sectors such as technology. In turn, many of them default on their loans, and spend less than at the peak of the boom. The same economists who saw increased real demand in 2003-4 now misread the situation as being a matter of “insufficient aggregate demand.” The financial crisis and Great Recession are the necessary outcomes of the unnecessary – and unsustainable – boom that originated in the loose central bank policies of the early 2000s.
[i] My analysis ties the boom and bust to rate policy, not government home-ownership promotion or “Wall Street Greed,” both of which are problems worth studying in their own right, but both of which had been omnipresent for decades prior to the 2000s.
[ii] Source: US Census Bureau.
[iii] Source: Bankrate.com.
[iv] Source: US Census Bureau.
[v] Source: NPR: http://www.npr.org/news/graphics/2009/mar/saving-rate/.
[vi] Source: US Federal Reserve web site http://www.federalreserve.gov/monetarypolicy/openmarket.htm. Of note, the ECB follows a similar rate path, generating bubbles in real estate markets featuring floating or adjustable rate mortgages with long amortization periods (e.g., Spain, Ireland).
[vii] Source: Bankrate.com.
[viii] ARM market share rises dramatically, exceeding 33% for purchase mortgages, and 38% for all mortgages, by 2005. Source: NY Fed. http://www.newyorkfed.org/research/current_issues/ci16-8.pdf
[ix] The median new home price rises from $175M in 2001 to $221M in 2004. The average new home price rises from $213M to $274M. Both median and average prices peak in 2007 at levels fully explained by the above-calculated increase in purchasing power. Source: US Census Bureau web site -http://www.census.gov/const/uspriceann.pdf
[x] Source: US Department of Housing and Urban Development, Office of Policy and Research, “An Analysis of Mortgage Refinancing, 2001-2003,” November 2004.
[xii] Ibid. It is worth noting that, according to circular flow theory, in which one person’s spending, even of borrowed funds, is another’s income, this should have marked the dawn of a long period of widely shared prosperity.
[xiii] US Federal Reserve web site: http://www.federalreserve.gov/econresdata/releases/mortoutstand/current.htm.
[xiv] This is not to suggest that this day of reckoning, now seen in the rear view mirror as being obvious, was recognized at the time by those who had the data. See: http://www.newyorkfed.org/research/current_issues/ci9-12.pdf
[xvi] Note that, for the first several years of any 30 year loan, there is minimal amortization, and that the crisis occurred only a few years after the ramp-up of “exotic” lending; thus the oft-cited interest-only and modest negative amortization provisions in many late-bubble loans, while “exotic” in form, are more of a red herring than a smoking gun, as such terms typically have a 10-15% effect on payment level, vs. a 30-60% effect from the rates. The rates were set at the central bank-manipulated index plus a standard margin. “Predatory lending,” then, primarily meant exposing borrowers to the Fed’s volatile rate policy.
[xvii] Not only were such loans made in response to a declining overall delinquency rate, they were made possible by the expansion of credit. There had never been regulatory prohibitions of loans to borrowers with low scores – the prohibitive factor had, prior to the 2000s, always been pricing, reflecting the limits to the supply of credit. One effect of increasing the supply of something is to reduce its price, and loans to people with poor credit scores were more likely to be made, and to cash-flow, in the high single digits in the 2000s than in the double digits at which such loans would have been priced in earlier periods.
[xix] The US Dept. of Labor estimates that 1.2-1.7 million construction jobs in existence in 2005 were solely the result of bubble-related demand Source: US Bureau of Labor Statistics web site: http://www.bls.gov/opub/mlr/2010/12/art1full.pdf. Of note, including the further positive feedback of inflated credit fueling the temporary increase in jobs and incomes would further strengthen the argument for the housing bubble as an example of an ABCT-described credit bubble. Note that we are referring to a growth in employment as a negative – the combination of increased household spending and boom-sector job creation are to Keynesians very real – “increased investment and spending, thus increased job creation,” thus a positive, in their circular flow diagram – to Austrians, these represent a the mirage of prosperity – an unsustainable boom or bubble, as the spending cannot be propped up forever, and the investment is made in production of something for which consumers are not in fact saving up to purchase. This is why, after the crash, Keynesians call for a reinflation effort, to try to support demand growth to the pre-crash level that they still believe was real, while Austrians oppose such measures on the ground that the previous consumption level was not sustainable at it was achieved with borrowed funds – indeed, the housing bubble was created partially in response to the collapse of the dot.com bubble.