This article is a qualitative overview of the issues involved, which I will pursue in later articles in more detail. Once again, I am making a rough draft of ideas that will end up as a chapter in my business cycle book. (Sorry, no news on the breakeven inflation book.)
IntroductionMy argument is straightforward: most recessions in the post-1990 era in the developed countries are associated with financial or banking system crises, with a few classes of exceptions noted below. If we look at the pre-1990 period, the linkage is somewhat more strained. I would note that the length of business cycles in this era means that we have few data points to work with.
I will immediately note exceptions.
- Sufficiently tight fiscal (and possibly monetary) policy can induce a recession. The euro periphery after the Financial Crisis demonstrates this.
- It might be possible that some other sufficiently large shock -- like an oil price spike -- could induce a recession.
- Countries following an export-led strategy (for the developed economies, Japan and Germany are the main examples), weakness in export markets can cause a recession even if the domestic economy is otherwise sound.
- When I write "developed economies," I am dividing the world as a fixed income investor would; I would not include the Asian economies hit in the 1997 crisis as being part of that basket, even though they are relatively rich countries.
One may also note that I am not necessarily claiming causality; it is possible that recessions cause financial crises. I am going to put that objection aside until a later article.
A Blind Spot for Mainstream Macro?The possibility of a financial crisis was certainly a blind spot for mainstream macro going into 2007; whether the situation has improved since then could be debated. I have not looked at the recent literature, but I would characterise the post-2007 response that I looked into as being underwhelming.
The fundamental issue is that workhorse mainstream models are built around optimising decisions of representative households. This is not a fertile starting point for a financial crisis, which is built around the notion of a panic.
A standard post-2007 fix is to embed credit spreads into the model, and so widening credit spreads act as a "random shock" that causes the economy to move from steady state growth into contraction. The problem with this fix is straightforward: credit spreads widen in response to a financial crisis. In other words, the crisis is already underway, and so the alleged "causal factor" is actually a lagged variable. From the perspective of a fixed income investor, such a model is useless: we want to price credit products on perceived recession odds.
There may be better post-2007 models out there; part of my book's research will be tracking down the existence of such models. Therefore, I want to emphasise to readers that my summary here is known to be incomplete.
A Very High Level History of Modern RecessionsI will now outline the history of recessions in the era that I am interested in. I will start with the United States, as it most familiar, and then outline the experiences in selected other countries.
The United States has had three recessions in the period of interest.
- The first is the recession of the early 1990s. The exact dating of the recession is associated with the oil spike that coincided with the invasion of Kuwait and the UN response. The United States was stumbling through a collapsed regional real estate bubble associated with the failed regulation of the Savings and Loan (S&L) industry. There was a policy angle: the very brief "peace dividend" after the end of the Cold War led to a sharp reduction in defence spending. (I graduated as an electrical engineer in the early 1990s, and the North American job market for such engineers imploded as a result of defense cutbacks. It only recovered once the telecommunications boom built up steam.) If we take an extremely narrow definition of "financial crisis," it may be that the early 1990 recession does not qualify, but it certainly met looser definitions.
- The early 2000 recession was associated with the demise of the technology bubble, as well as general over-investment by firms more generally. The banking system was largely untouched by this bubble, and I believe that this was partly due to the pioneering use of derivatives in credit risk management. (This partly explains why their use did not raise eyebrows in later years.) However, the corporate bond market was hammered by defaults and downgrades; if you owned previously AAA-rated European incumbent telecom bonds, the crisis was very real. The shut down of the corporate bond market helped scupper the investment boom by corporations.
- The Asian Crisis of 1997 caused a short panic in the United States (just when I started working in finance), but the associated collapse of oil prices helped bail out the real economy, and so there was no associated recession.
- The 2007-2008 Financial Crisis do not need much of an explanation at this point. The key point to note is that this cycle was fully global; the pattern of earlier recessions was somewhat regional in nature. Once again, housing was an important part of the crisis; the fact that modern home owners are highly indebted turns housing market hiccups into events of concern for finance.
I will go through various countries and regions in no particular order.
- Canada largely followed the cyclical pattern in the United States, although the reasons behind them diverged. There size of the government sector was starting to be rationalised, and so there was a policy element behind slow growth. The early 1990s recession was less associated with a financial crisis, although there were regional condo booms that went bust.
- The United Kingdom had a fairly ugly housing bust courtesy of the end of the Lawson Boom. (Coincidentally, I was studying there at the time.) Interest rates were being set in a fashion to defend the sterling parity in the ERM. Whether or not this was a "financial crisis" could be debated, but there was certainly a financial/housing aspect to it. The U.K. economy followed a somewhat distinct path until its cycle recoupled with the rest of the developed economies.
- Japan had an epic financial bubble that burst by the early 1990s. The banking system was left in a largely zombie state thereafter for an extended period. Given the collapse of dynamism, it is not surprising that cyclical variations in the 1990s were muted. The Japanese economy was more sensitive to overseas developments given its strength as an exporter, and limited domestic sources of volatility.
- Australia has largely managed to avoid recessions during this era. The economy has been riding a wave of resource investment and housing boom.
- I would need to refresh my knowledge of the Nordic countries, but there was a housing bubble/banking bust in Sweden in the early 1990s (also associated with the ERM currency parity defense). I will need to find better references, but here is the Wikipedia article.
- The German economy had to deal with the impact of the integration of East Germany, and then various reforms. German domestic financial practices are famously stodgy (with her international banks less so). Germany falls into the camp of export-led growth, and so the cycle is somewhat more dependent upon trends in importer nations.
- Continental Europe generally rode the wave of euro convergence, which turned into divergence after 2007.
The above comments hardly constitute a rigorous statistical survey of the data. However, it indicates the pattern that I am discussing, and the countries that I am discussing.
The analysis of the economist Hyman Minsky gives the cleanest explanation for the linkage between financial crises and recessions. I expect that one of my next articles will be a lengthier description of his "financial instability hypothesis."
For academically-inclined readers, I will emphasise that I wrote "cleanest explanation," I am not claiming that he invented the entire concept single-handedly. Minsky himself largely deferred to Keynes in terms of his thinking. For readers who want mathematical economic models, Minsky is perhaps not the best source.
For readers who are not familiar with Minsky, I will summarise his financial instability hypothesis as follows. In order for a capitalist economy to achieve its "normal" state of growing nominal incomes, we would expect that some sectors are running increasing financial deficits. (It is theoretically possible to achieve increased nominal income growth by increasing the velocity of circulation of financial instruments, but that is not typically seen, other than for short periods.) If the sector is the government, it runs a fiscal deficit. For non-governmental sectors, we need some entities within the aggregate to be issuing financial instruments, which are normally debt instruments in practice.
The financial sector (including banks) makes loans based on various conventions: debt-to-income ratios, etc. Historical default rates are used to calibrate the riskiness of loans. In a growing economy, nominal incomes grow, and so borrowers can avoid default. In other words, increasing loan amounts ratify the previous decision to make loans. Lenders that grow their market share are more profitable than stodgy lenders, and so there is a Darwinian pressure to eliminate conservative financing practices.
This works until it doesn't: eventually lending standards become some hair-raising insane that eventually even investors allocating other peoples' money get scared. The lending cycle goes in reverse rapidly, until the central bank steps in.
When we look at the importance of housing in modern economies, it is no surprise its fluctuations dwarf other cyclical factors, like the inventory cycle.
I expect to write articles discussing the following topics.
- Minsky's Financial Instability Hypothesis.
- How do we measure financial crisis risk?
- Steven Keen's work.
- Causality: do financial crises cause recessions, or vice-versa?
- Can we predict financial crises?
- Does post-2007 mainstream macro have anything useful to add on this topic?
- Historical analysis of how financial practices change.
- More historical surveys of crises (and surveys of the empirical literature).
(c) Brian Romanchuk 2018