Global
Of Disasters, Recessions, Crunches and Busts
August 14, 2008

By Joachim Fels | London

Recession risks in the G7 economies are clearly on the rise.  In Japan, GDP declined by a not-annualised 0.6% between the first and the second quarter.  In the euro area, 2Q GDP data released tomorrow are also likely to show a contraction of output.  And our US economists continue to forecast two consecutive quarters of GDP declines in the US starting in the final quarter of this year.  To be sure, full-blown recessions in any of the major economies over the next year or two are NOT our central case – the most likely outcome still appears to be a broad economic stagnation lasting well into 2009.  However, a worse outcome is a distinct possibility that warrants further investigation.  To this end, we take a closer look today at what history tells us about the frequency, depth and duration of recessions, and whether recessions differ if they coincide with house price falls, equity busts or credit contractions.  Moreover, we strongly recommend our readers to peruse a recent report by our global economics and strategy team exploring what investors should do in the various asset classes if the cycle turns really nasty (see Hedging Global Recession, Morgan Stanley Cross-Asset Strategies, August 7, 2008).

In this article, we survey the historical evidence on recessions.  We draw heavily on two excellent recent papers by Harvard economists Robert Barro and Jose Ursua (Macroeconomic Crises Since 1870, NBER Working Paper 13940, April 2008) and by IMF economists Stijn Claessens, Ayhan Kose and Marco Terrones (What Happens During Recessions, Crunches and Busts? August 5, 2008). 

Disaster probability 3.6% per year.   The study by Barro and Ursia (BU) focuses not on garden-variety recessions but on really large macroeconomic disasters, which they define as a cumulative decline in real per capita consumption or real per capita GDP of at least 10%.  Such disasters have been rare in recent decades, especially in the advanced economies, but using GDP data for 35 OECD and non-OECD countries extending back to 1870, in many cases they find a surprisingly large number of economic disasters.  To be precise, they count 148 such crises for GDP, implying disaster probabilities of around 3.6% per year.  Disasters last 3.5 years on average and GDP typically declines by a cumulative 21% over that time.  These are very big numbers by today’s standards, and it is important to point out that the bulk of these disasters occurred before 1950, and many were associated with the two world wars and the great depression of the 1930s.  The first half of the last century showed much bigger macroeconomic volatility than both the preceding 100 years and the years since then.

Few disasters in recent decades.   In the last 50 years, the BU data show only one GDP disaster in an OECD country. It occurred in Finland between 1989 and 1993, when per capita GDP fell by a cumulative 12.4% in response to the breakdown of trade with the former Soviet Union when the Soviet empire disintegrated.  There were several disasters in non-OECD countries, though, some of which occurred during the Latin American debt crisis in the 1980s and other during the Asian crisis in the late 1990s.

No less than 112 recessions in OECD countries since 1960.   Barring a war or another cataclysmic event, a macroeconomic disaster of the dimensions discussed above is unlikely to occur in the advanced economies.  However, normal recessions are far more frequent, as a recent thorough and detailed study by three IMF economists (Claessens, Kose, and Terrones, see above, henceforth CKT) reminds us.  On their count, there have been 112 recessions in the 21 OECD countries since 1960.  In addition, they study 112 cases of credit contractions, 114 episodes of house price declines and 234 equity price declines and their various overlaps with recessions.  The following results stand out from their analysis:

•           The typical recession lasts a little less than a year (3.6 quarters to be precise) and involves a GDP drop (from peak to trough) of a little less than 2%.  The more severe recessions, defined as the worst quartile of all recessions, last about 4.7 quarters and have an amplitude of close to 5% of GDP.

•           Contractions in credit to the private sector, house price declines and equity price declines typically last longer than economic recessions and involve larger peak-to-trough declines.  For example, the severe credit contractions (‘crunches’) last on average more than 10 quarters and involve a peak-to-trough decline in private sector credit of 17%.  Severe house price declines (‘busts’) last 18 quarters and show a median decline in house prices of close to 30%.

•           Recessions that coincide with either a credit contraction, a house price decline or an equity price plunge last longer and carry larger output losses than recessions without such events.  For example, a recession accompanied by a severe house price bust lasts 4.6 quarters and displays a 2.6% amplitude, while recessions without a house price decline only last 3.2 quarters and have an amplitude of 1.5%.  Recessions accompanied by an oil price surge or an inflation surge are deeper, but not necessarily longer, than recessions without such events.

Finally, it is worth comparing the current US downturn (whether it’s a recession or not is still not clear) to the average US and non-US OECD recession.  A few things stand out:

•           First, US GDP growth, though a bit weaker than ahead of the typical recession, has followed a roughly similar path.  The jury is still out on whether the US will experience a recession.

•           Second, US residential investment has declined much more sharply than in the typical US or OECD recession.  This is clearly related to the much stronger-than-usual decline in US house prices.

•           US credit growth was stronger than typical before a recession, but has recently declined sharply and is now more in line with past pre-US recession experiences.

•           Real short rates have fallen by much more than in or ahead of previous recessions.  This confirms a point we have been making several times before: The Fed has responded much more aggressively to the decline in house prices and the related mortgage crisis than ahead of or during the typical recession.

•           Similarly, fiscal policy, measured here by the growth rate of real government consumption, has been more expansionary than usual ahead of or during a recession.

The jury is still out.  Where does this all leave us?  If the historical evidence is anything to go by, a recession – if it happens – could be expected to be longer and deeper than normal because it is accompanied by both a credit crunch and a severe house price decline.  On top of this, the oil price and related inflation surge also points to a deeper-than-usual recession, judged by the historical evidence.  However, there is one important difference to past recessions: US monetary and fiscal policy has been much more proactive and aggressive in response to early signs of the slowdown.  Only time will tell how these divergent developments play out.  One thing is clear, though: With a barrel of oil trading at around US$113 today rather than US$145 a month ago, the growth outlook for the big oil importers is a little less bleak than it was then.

UK
UK Inflation Report Broadly Consistent with Rates on Hold View
August 14, 2008

By Melanie Baker, CFA | London

Broadly, the forecasts in the Inflation Report look consistent with our view that the single-most likely outcome for interest rates is that rates are on hold this year and next, with the balance of risks to that view more to the downside than upside heading into 2009.

The MPC presented its latest inflation and growth forecasts in its quarterly Inflation Report.   The committee significantly lowered its GDP growth forecasts for the next few quarters and now expects broadly flat GDP over the next year and for year-on-year GDP growth to decline to close to zero.  Its latest forecast for inflation consistent with that GDP growth profile (and largely unchanged interest rates) shows inflation moving up to around 5% in the near term, then falling back relatively rapidly to just below the target at the two-year horizon (and declining somewhat beyond that).  However, while the MPC thinks that the balance of risks to its GDP outlook is to the downside, it thinks that the balance of risks to its inflation profile is to the upside.

We think that the bank will likely keep rates on hold rather than cut rates:

1.         Despite seeing a downside balance of risks to its GDP growth profile, the MPC sees an upside balance of risks to its inflation profile in the medium term.

2.         Our own and the Bank of England’s central forecasts show inflation far above target in the near term and above target for most of 2009.  The MPC remains concerned about the risk that continued above-target inflation raises the longer-term inflation expectations of price and wage setters, making the rise in inflation more persistent.  The committee thinks that this risk has risen.

3.         GDP growth recovers on the Bank of England’s central profile despite no assumption of rate cuts.

4.         A sharp slowing in UK GDP growth is seen as necessary to ensure that inflation returns to target.

5.         Beyond the two-year horizon, the market interest rate used in the bank’s forecast rises to 5.2%.  If rates stayed on hold at 5.0%, it is not clear that the profile would still show inflation continuing to slow below the target beyond the two-year horizon.

6.         We think that the Bank of England’s central forecast for GDP growth is sensible – a major recession is not the most likely outcome.  Our own forecasts for GDP growth aren’t too different to the Bank of England’s, though its central profile shows a lower ‘trough’ than our own forecasts.  Our own forecasts show the economy just skirting a technical recession, but we continue to think that an early 1990-type recession is not the most likely outcome.  This seems to be the Bank of England’s view too.

Focusing solely on its central forecast for inflation at the two-year horizon might make it look like the bank itself thinks that a rate cut is likely in the near future, but that mechanical reading would be mistaken, we think, for two reasons: 1) it implicitly assumes that the bank’s role is to focus on inflation two years ahead.  In fact, the bank’s mandate is simply to keep inflation close to target – and it is very clear that inflation will be above target for much of the next two years; and 2) risks around that central inflation profile are clearly asymmetric.

A near-term rate cut should look like a significant probability into 2009: Given the likely sharp economic slowdown (and the risk of something more severe) and given that we expect inflation to return rapidly to the target (with inflation peaking in about September already), there is a significant chance of a rate cut as we head into next year.  We also think that the neutral rate is now somewhere between 4.50% and 5.00%, such that the Bank of England may feel more comfortable moving rates somewhat below 5%, as inflation returns towards target and against a weak economic backdrop. However, with many employees covered by pay deals agreed in the first four months of the year, the MPC might prefer to delay any such cut (and by 2Q09, the economic outlook is anyway likely to look somewhat brighter).

Our central case remains that the Bank of England is likely to keep rates on hold, assuming that the UK economy slows significantly but avoids a significant recession.