Wednesday, December 3, 2008

Credit Crunch Subprime Recession: How long?

Readers may recall my prediction a few weeks ago that US recession probably started in the Winter 2007. I explained that it takes time for GDP data to be corrected and revised (major revisions possible for up to 2 years, mainly due to lateness in getting all profit/loss, unearned income and spending data in). I expect UK recession will be back-dated to the Summer of 2008 and EU recession (currently official) data to be revised and pushed forward, not back, to end of 2009. The US NBER has now confirmed the first part of that prediction stating that revised data suggests recession started in December 2007. Now that US economy is officially in recession, the big question on everyone’s mind is, how long will it last? Given that US recession (which may come to be called the Credit Crunch or Subprime Recession) is officially already 12 months old, the current recession is the fourth longest in the past 80 years and some others predict therefore it is on track to be as long as recessions in the ‘70s and ‘80s i.e. four quarters, 16 months. We are a long way from the Great Depression, which lasted 43 months or 3.5 years. On Monday, Ben Bernanke made a speech saying it is not useful to compare the current recession with the Great Depression because that was much worse. In looking at the above chart from WSJ we have a cluster of recession duration data, and an outlier (the Depression). The use of an overall average makes sense i.e. typically 9 months to crash and 72 months to recover to pre crash values. Or, we have an average duration for downturns unrelated to the long-term credit cycle, and another longer duration for small-sample downturns representing unwinding of te credit (crunch) cycle. The following is the U.S. long-term credit cycle chart (from SocGen). There are many clues for pondering the timing of the credit /business /economic cycle. Calling the turning point is not easy. The debate is often couched in terms of whether the recession will be V, U or L shaped. Gauging the magnitude of the financial sector's ills on top of the economic is the problem. I estimate them both to be of equal strength. Fiscal & monetary policy actions on top of guarantees, sureties, asset swaps and direct investment in banks by Government I calculate will cut the problem in two halves. Monetary and Fiscal measures are essentially aimed at mitigating normal recession while financial sector measured aim to restore banks to a more normal condition whereby they can play a part in recovery by maintaining a reasonable (and economically realistic) level of lending. There is a problem if the banks and the so-called 'real economy' do not respond intelligently and recession is therefore longer than usual, say 2-3 years! The conventional assumptrion is that banks will recover 55% of their booked losses, which may be somewhat at the upper end of past experiences. Various attempts by profilers to characterize the cycle are throwing out confusing signals. Different economic and financial measures are making different forecasts that adjust as the data is revised about timing of the economic contraction. Has it only just begun, in the middle or the beginning of the end?Depending on the data one chooses to highlight, all three scenarios look plausible.
Consider the chart below (courtesy of, published a few weeks ago. Setting aside the NBER's announcement, a simple review of earlier GDP numbers suggested economic activity peaked earlier in 2008 (Q2 posted a 2.8% annualized real rise in US GDP). The casual observer might think that Q3's -0.5% fall in US GDP (when UK posted a 0.3% rise) was the start of the recession, with most economists saying Q4 will suffer a bigger decline and 2009's looking weak. If we are in a V-shaped recession it should look like this. Non-farm payrolls are often resorted to in the US as a main cycle indicator. Year-over-year comparisons of this series have a habit of diving sharply ahead of the trough. The labour market has certainly been weak, 2.5 million jobs lost in the past year or so, not counting those who are unemployed do not yet qualify to register as unemployed, suggesting that a peak may have passed? Year-over-year comparisons of non-farm payrolls tend to coincide with periods when the economy's at its worst. Alternatively, stock market prices are signalling that we are deeper into the cycle and close to the cyclical bottom than GDP or payrolls suggest. The stock market has a history of falling ahead of the trough. Then there is the spread between long and short rates, signalling that the economic trough is behind us and the recovery phase has begun. The yield curve tends to invert ahead of the trough. But, the curve has inverted on and off for several years? The naïve explanation suggests the economy is set to rebound soon. But soon could mean the middle of next year just as in the UK the Government predicts recovery starting at the back-end of next year. The above metrics should be in sync for identifying where we are in the business cycle. This time, there is more variety in the reading the tea-leaves. In a year when rules of thumb and other prudential measures such as standard risk gradings have appeared to fail or look perverse, our challenge is determing which metrics harbour faulty unreliable signals. One valid question on reflection is why has the market been so stupid as to over-sell stocks so far below their book value? One reason may be the extreme experience of the opposite, the dotcom bubble, when investors prices Tech stocks far above their net present value or book value, pricing some in $billions that were generating mere millions in revenues. Then too the bubble in residential and commercial real estate. Two years ago, the all-knowing market bid up commercial property rates to 3%-4% rental return and in residential often below 3%. Dividend yields of 3% on REITs were rationalized as appropriate because of the infinite growth in lease rates ahead. Today, dividend yields on REITs are around 15%, just as profit ratios to share prices on many stocks are double this. Commodities were pumped up sometimes tenfold compared to prices today. Oh, but the market got banks right in the summer and they tanked, or was the market hijacked by short-sellers? Less than two years ago, junk spreads were at all time lows. Emerging market sovereigns were trading as if they were on the verge of becoming developed world credits. The near-perfect information markets were telling us how low solvency risks are and that liquidity risk had been abolished in a free flowing globalised world and how we were in the middle of the strongest period of co-ordinated global growth in history. Today, the markets are throwing babies out with the bath water. Risky assets are deemed too complex, therefore too risky, in a back=to=basics cash-is-king world, putting them up at fire-sale prices for the vulture funds to buy as never before. From junk bonds to investment-grade bonds to stocks to commercial real estate, anyone with a time frame longer than my nose is able to buy distressed assets from insolvent owners at 70-80% discount from pre-crisis values.
If you are playing the long game as a trader, this is a horrible market, probably more downside to come. If you are playing the long gsme as an investor, the market is offering up what is a one in a generation (some say once in a lifetime) opportunity to buy.

No comments:

Post a Comment