A perspective on 2000's illiquidity and capital crisis: past banking crises and their relevance to today's credit crisis.
The current financial crisis is marked by the simultaneous occurrence of several severe economic and financial conditions. These circumstances reinforce the need to provide expectations for investors and consumers in an environment marked by risk and uncertainty.
This author strongly believes the current financial crisis shares a key characteristic of prior banking crises. Sharp credit contractions were preceded by inflationary asset bubbles and credit expansion. It was therefore felt useful to obtain a perspective on today's crisis by viewing it in the context of 15 prior systemic banking crises. This setting enables an assessment of the possible course of today's financial crisis.
If today's crisis turns out to be similar to the benchmark, a recovery period of 10.6 years is expected for real equity prices to regain their peak. Then, equity prices could return to their inflation-adjusted peaks by mid-2018, from the prior peak of October 9, 2007. Real total equity returns of about 10 percent annually for nine years would be implied, if March 9, 2009 were adopted as the date of the true trough in equity prices.
The unemployment rate is likely to peak at about 11.5 percent if today's crisis turns out to be similar to past credit crunches. And real Gross Domestic Product is expected to decline by four percent between 2008 and 2009.
In the second half of 2009, concern is being voiced of floundering in uncharted waters, with no clear indication of the depth and length of the economic recession and its accompanying state of illiquidity, especially for small business. Moreover, housing prices continue to decline, with foreclosures increasing markedly (Norris, 2009). The unemployment rate continues to rise, and energy prices are once again increasing (Wyss, 2009). Banking institutions, which had been besieged by actual and near-insolvencies, also faced capital shortages. Losses from household loans and credit cards are mounting (de la Merced, 2009).
Investors are awaiting clearer evidence that the recent months' rally in the stock markets was not unwarranted, perhaps to be "retested" with a typical 10-20 percent market contraction (Lim, 2009). And wary consumers are saving at the highest level since 1998, as expressed by the personal savings rate as a percent of disposable personal income (Healy, 2009). During the second half of 2009, the Consumer Confidence Index[R] reflected weak economic conditions, uncertainty about future job prospects and a lowered outlook on spending (Conference Board, 2009). While a welcome relief from extremely volatile equity markets and illiquid credit markets, the overall mood is one of risk and uncertainty.
NYU Professor Nuriel Roubini presciently stated in March 2007, if the economy slips into a recession, "then you have a systemic banking crisis like we haven't had since the 1930s" (Roubini, 2008).
I believe the current financial crisis can properly be viewed as a systemic banking crisis due to its sharp credit contraction, preceded by a housing market price bubble and credit expansion. I maintain we are not merely experiencing another cyclical economic crisis characterized by a deep recessionary period or just a bear market phase of the stock market. It was therefore felt useful to view the current liquidity and capital crisis in the context of prior systemic banking episodes.
The objective of this paper is to provide an assessment of the current position and possible course of the 2000's financial crisis for investors and consumers in an environment marked by risk and uncertainty. An analysis of previous banking crises is undertaken to provide a comparison with the current crisis. By determining the most appropriate benchmarks based on 15 prior systemic banking contractions, we can discern the risk of further consequences on the ongoing economic/financial crisis. For example, the duration of the equity price downturn and the peak unemployment rate may not yet have been realized.
Five specific issues of interest and general concern will be addressed in the Findings section, based on past protracted economic crises of the type we are currently experiencing:
1. The duration and magnitude of the downturn in stock market prices.
2. The number of years for equity prices to regain their inflation-adjusted peaks.
3. The real annual total returns required for regaining those peaks.
4. The peak unemployment rate.
5. The real decline in the nation's output from peak to trough. ("Real" refers to inflation-adjusted dollars, rather than current dollars).
Uncertainty may surely be with us, but current conditions resemble past credit episodes. Those shared common characteristics will serve as bearings to our current circumstance, and guideposts to investors, consumers, economists and regulators.
As Abraham Lincoln said in 1858, "If we could first know where we are and whither we are tending, we could better judge what to do and how to do it."
Basic Approach and Data Description
Today's credit, illiquidity and capital crisis bears strong resemblances to prior systemic banking crises. This article adds to the literature sharing this perspective, specifically the work of economists Carmen M. Reinhart of the University of Maryland and Kenneth S. Rogoff of Harvard (Reinhart, 2009) and (Reinhart and Rogoff, 2008, 2009, 2009a), and chartered financial analysts Jon Ruff and Vincent L. Childers (Ruff and Childers, 2009). This author has extended the work of Ruff and Childers to derive new, more robust benchmarks to serve as more appropriate bases of comparison to the current credit crisis.
In characterizing banking crises, Reinhart and Rogoff state: "We find that systemic banking crises are typically preceded by asset price bubbles, large capital inflows and credit booms, in rich and poor countries alike." (Reinhart and Rogoff, 2008). "Financial crises are protracted affairs. Asset market collapses are deep and prolonged. Real housing price declines average 35% stretched out over six years." (Reinhart, 2009).
Ruff and Childers view a large credit contraction as the defining element, noting: "It is this characteristic--a severe shock to a country's financial system--that differentiates the current crisis from more typical business cycle-related recessions and bear markets." (Ruff and Childers, 2009, p. 1).
Reinhart and Rogoff present average changes in real housing prices, equity prices, the unemployment rate, national output, and government debt for 14 banking episodes over the past three decades (Reinhart and Rogoff, 2009a). Ruff and Childers exhibit eight crisis characteristics, as listed and defined in Exhibit 1, for each of 15 systemic banking crises. Changes in equity prices, real national output and the unemployment rate are displayed by both studies. The decline in real earnings per share (EPS) and changes in real loan growth and contraction, as defined in Exhibit 1, are included only by Ruff and Childers for each of their episodes. Nine credit events are common to both samples, with a full listing available in Appendix II.
I eschewed studying the ten official, post-1947 recessions as inappropriate bases of comparison from which to view the severe crisis that occurred in the U.S. beginning in December of 2007 (National Bureau of Economic Research, 2009). Rather, further analysis utilizing the Ruff-Childers sample set of 15 banking crises has been undertaken, because this author fully supports their choice of severe credit contractions--rather than typical business cycle-related recessions--as apt comparisons to the current crisis.
Ruff and Childers compared characteristics of the current crisis against their benchmark of prior credit contractions and found that "If today's crisis turns out to be "average,"... prices could return to their inflation-adjusted peaks by the year 2016." And "a recovery by then would imply eight years of real double-digit stock returns." (Ruff and Childers, 2009, p. 3).
Two shortcomings of their work were identified. Their benchmarks were determined by taking simple averages of the banking episodes contained in several universes of the 15 banking crises. This is cause for concern, because some of the "average" results presented in the Ruff-Childers paper might be skewed by assigning the same weight, or influence factor, to each crisis. Also, by publishing in January 2009, before the S&P 500 nadir of March 9, 2009, was realized, Ruff and Childers identified the trough in equity prices to have occurred on November 20, 2008.
My contributions are generating new, more sharply-honed and robust benchmarks by:
1. Determining the weighted average, rather than the simple average, of the banking crises, accompanied by 95 percent confidence intervals to summarize the risk associated with the weighted mean.
2. Selecting new universes or subsets of the fifteen prior episodes to serve as more appropriate bases of comparison to the current credit crisis.
This author suggests in the Methodology section:
* Based on observations of current conditions compared to those of the 1930s, the unemployment rate and the decline in real Gross Domestic Product (GDP) associated with the Great Depression are extremely unlikely to recur
* New, adjusted benchmarks for just these two indicators were created, omitting that banking crisis
* In addition, estimation of the year in which the prior peak of real equity prices could be attained was updated by tentatively adopting March 9, 2009, as the possible trough in stock prices
Application of the weighted mean to data consisting of characteristics associated with systemic banking crises is highly appropriate to represent the relative importance or magnitude of the credit crunches in the sample. This application is original, as neither set of co-authors, Ruff-Childers or Reinhart-Rogoff, utilized this statistical measure of centrality.
The weights selected to represent the relative influences of the crises were Relative Wealth of the crisis-impacted countries--roughly, the affected country's GDP relative to the United States' GDP. [Relative Wealth is more precisely defined in the ensuing Methodology section.] Ruff and Childers, by calculating simple averages of their constituent banking episodes, assigned the same weight, or influence factor, to, say, the systemic banking crises of Colombia in the late 1990s and the U.S. in the late 1980s. The weight I assigned to the latter episode was nearly five times Colombia's.
The differences between the more robust and representative weighted average and the simple average benchmarks are delineated in the Comparison of Results section. For example, based on differently-calculated benchmarks based on the same 15 banking contractions, I expect real equity prices to reach their peak in 2018, two years later than the Ruff-Childers result cited above.
Results are also presented in terms of annual real total returns to recovery from the assumed equity price trough of March 9, 2009 to the (prior) inflation-adjusted peak in real equity. Total return is defined as price-plus-gross-cash-dividend return (Standard & Poors, 2009). These findings provide a practical measure of attainability of the peak in real equity to which investors and consumers can relate.
The three groups of five banking crises each and the universe of all 15 crises which Ruff and Childers specified, and which I adopted, are listed with further details in Appendix I (Ruff and Childers, 2009, p. 2). The eight characteristics of a crisis, also based on the Ruff-Childers paper, as previously noted, are listed and defined in Exhibit 1.
Exhibit 1: Definitions of the Crisis Characteristics Characteristic Definition Real Economy EPS--Peak to Trough Percentage change in one-year operating earnings per share from peak to trough, at end of month, adjusted for inflation. Like all other crisis characteristics, EPS is derived for "host" country affected by banking crisis. EPS, sometimes called earnings (E), is derived using country price index P and country price/earnings (P/E) ratio history. For U.S. episodes, Price P is defined as S&P 500 Stock Price Index, and EPS as as earnings accruing to this Index. GDP (per Capita, Percentage change in real Gross Domestic Product Annualized) (GDP) (annualized) from peak to trough calendar year, adjusted for inflation and seasonal fluctuations; also known as real economic growth or national output. Unemployment: % Pt. Percentage point increase in the unemployment Incr.--Trough to rate from its trough to its peak; Peak seasonally-adjusted unemployment rate of civilians does not include those who have stopped looking for work. Real Equity Price Peak-to-Trough Drop Percentage change in equity price (stock market (%) index), adjusted for inflation, from peak to trough end-of-month; price or S&P 500 Index excludes cash dividends. Peak-to-Trough Years Number of years of real equity price decline from prior pre-crisis peak to trough of crisis. Years to Regain Peak Number of years for real equity prices to regain prior inflation-adjusted peak from prior peak. Real Loan Growth Three-Years-Prior to Percentage change in domestic credit market, Peak adjusted for inflation, from three-years-prior to peak (un-annualized). Peak to Trough Percentage change in domestic credit market, adjusted for inflation, from peak to trough (un-annualized),
To derive the weighted-average benchmark, the readily-available Relative Wealth data was utilized, with values for each banking crisis contained in Appendix I of the Ruff-Childers (2009, p. 8) paper. Relative Wealth was defined as "real GDP per capita of [the affected] country divided by the greater of U.S. or UK real GDP per capita." (Ruff and Childers, 2009, p. 8) Normalizing the Relative Wealth factors for each crisis to form an index of weights that sum to 100 percent for each universe, each factor was then divided by the sum of Relative Wealth figures for each banking crisis for a given universe.
The largest weights of 9.63 percent were associated with U.S. banking crises #1, 11, 12 and 13, as identified in Appendix |; the smallest weight of 2.02 percent was assigned to crisis # 10 (Colombia). The more robust and representative results obtained by the weighted-average benchmark are mainly attributable to not applying the implied same fixed weight of 6.67 percent to each of the 15 crises, as associated with the "average" figures derived in the Ruff-Childers paper.
The results associated with the new benchmark based on all fifteen prior crises are presented in Exhibit 2, accompanied by their 95 percent confidence intervals. For each of the benchmark's characteristics, say, the percentage change in real equity price from peak to trough end-of-month, the expected (or mean) figure of -50 percent, is shown, along with a lower bound of -61 percent and an upper bound of -38 percent. (Exceptions for displaying the expected figure based on all fifteen crises for two characteristics are noted below.)
Exhibit 2: Benchmarks for Banking Crisis Characteristic Benchmark: 15 95% Confidence Crises Interval: 15 Crises Lower Bound Upper Bound Real Economy EPS--Peak to Trough -102% -151% -53% GDP (per Capita, -6.7% *# -9.8% * -3.6% * Annualized) Unemployment: % Pt. Incr. 6.6% * 3.9% * 9.3% * Real Equity Price Peak-to-Trough Drop (%) -50% -61% -38% Peak-to-Trough Years 3.0 1.2 4.8 Years to Regain Peak: Prior Peak to Peak 10.6 6.0 15.2 Real Loan Growth Three-Years-Prior to Peak 21% 14% 28% Peak to Trough -21% -30% -12% Characteristic All 15 Group I: Great Crises Developed Depression Countries U.S.-1930's Post-WWII Real Economy EPS--Peak to Trough -102% -130% -67% GDP (per Capita, -6.7% *# -4.0% # -30.8% Annualized) Unemployment: % Pt. Incr. 6.6% * 5.2% 21.0% Real Equity Price Peak-to-Trough Drop (%) -50% -51% -81% Peak-to-Trough Years 3.0 4.1 2.8 Years to Regain Peak: Prior Peak to Peak 10.6 6.9 26.5 Real Loan Growth Three-Years-Prior to Peak 21% 18% 27% Peak to Trough -21% -20% -45% * Figures shown are for the adjusted benchmark, a more representative benchmark for these two characteristics; the Great Depression crisis is excluded from the adjusted benchmark. # The most representative benchmark for GDP change is -4.0% associated with Group I developed countries. Note: The benchmark for GDP, utilizing all 14 banking episodes for which data exists, is -9.1%; benchmark for change in unemployment rate, utilizing all 15 banking episodes, is 8.0% points.
To generate a wider set of results, supplementing those associated with the universe of all 15 banking crises, two other event sets are introduced: (1) a second benchmark generated by taking a weighted average of banking crises for Group I, consisting of five major credit contractions associated with developed countries after World War II; and (2) the Great Depression of the 1930s. These two event sets were both included in the total universe of 15 banking crises, but with diminished influence. The decreases in real equity price from peak to trough are-51% and -81% for the Group I benchmark and the Great Depression, respectively. (One additional universe is presented for two characteristics--the unemployment rate increase and the change in real GDP - as noted below.)
Thus, five figures for each of eight crisis characteristics are presented as benchmarks for the current credit crisis of the 2000s, with the Great Depression set of numbers emanating directly from Ruff and Childers (2009, p. 8).
The likelihood of the Great Depression with all of its manifestations making a return appearance is probably very small, for the following reasons:
1. Bank depositors are now insured for $250,000 per account by the Federal Deposit Insurance Corporation, which was established in 1933, after the Great Crash and after the beginning of the Great Depression. (The FDIC raised the standard maximum deposit insurance amount per account for member banks from $100 thousand to $250 thousand during the current crisis in 2008.)
2. The magnitude of the 25 percent decline in the consumer price index, or deflation, in the Great Depression was attributable to continued adherence to the gold standard, since discontinued. Under the gold standard, governments had little influence on the money supply. (Ruff and Childers, 2009, p. 5)
3. "Today's fiscal and monetary policies are certainly a lot better than what the world saw in the Great Depression of the 1930s." (Reinhart and Rogoff, 2009)
Thus, the magnitude of the real per capita Gross Domestic Product decline of 30.8%, associated with the banking crisis of the 1930s, is extremely unlikely to recur.
The Great Depression figures for the decline in real GDP and the increase in the unemployment rate [for similar reasons, as explained in the Findings section] were treated as influential outliers in the universe of all fifteen banking crises, which, in my judgment, are highly unlikely to be experienced again, due to changed external conditions. Therefore, I created, just for those two characteristics of the eight considered, an "adjusted benchmark" consisting of all banking crises but excluding the U.S. 1930s episode.
However, other characteristics associated with the Great Depression might very well merit attention, perhaps even as attainable observations, as noted below.
* The "housing price decline experienced by the United States to date during the current episode (almost 28% according to the Case-Shiller index) is already more than twice that registered in the U.S. during the Great Depression." (Reinhart, 2009)
* Today's financial crisis is also far more global than its 1930's counterpart, with downturns and recoveries more "interconnected" to those of other countries.
Moreover, Simon Johnson, former chief economist of the IMF, has recently written: "What we face now could, in fact, be worse than the Great Depression--because the world is now so much more interconnected and because the banking sector is now so big. We face a synchronized downturn in almost all countries, a weakening of confidence among individuals and firms and major problems for government finances." (Johnson, 2009).
The S&P 500 Index and its total return, (TR), were selected as representing equity prices and total equity return in the U.S. stock market for the current crisis (Standard & Poors, 2009). A decline in actual real total equity return of 56.0 percent was determined by the following steps. First, the nominal total return percentage, TR(%), for a period was derived by calculating the percentage change in the broad-gauged S&P 500 Index of total returns, from its peak value on October 9, 2007, to its "assumed" trough on March 9, 2009, using the formula:
TR(%) = [(Index of new date-Index of old date) / (Index of old date)]. (1)
Then the real total return was calculated by the formula:
Real total return = [(1+Nominal TR(%))/(1+Inflation rate)]-1. (2)
The current year's lowest total return, with twelve months of actual 2009 data available, was attained by the S&P 500 Index on March 9, 2009, with the S&P 500 at 676.53 and its TR of 1095.04. The decline in real total returns was 56.0 percent from the recent peak S&P 500 TR of 2447.03 on October 9, 2007. Similarly calculated, the decrease in real equity prices was 57.5 percent from the S&P 500 peak of 1565.15. The 17-month inflation rate was 1.81 percent (InflationData, 2009).
The nominal declines of 55.3 percent for total return and 56.8 percent for equity prices were calculated in a similar manner with the inflation rate set to 0.0 percent in Equation 2.
The four annual returns required for equity market recovery--total returns and price returns, with both returns stated on a real and nominal basis--are determined as explained below, and introduced and discussed in the Findings section. The following three equations were utilized:
Real index peak = [(Real decline to trough)/(Inflation rate +1)].
Percent to real recovery = [(Real index peak)/(Index trough)]-1. (4)
Annual return to recovery = [(Real index peak/Index trough) ^(1/Time to recovery)]-1 (5)
where the notation "b ^ c" represents base b being taken to the power c.
Observations associated with the benchmark for 15 banking crises and other benchmarks, as summarized in Exhibit 2, include the following findings for the five specific issues identified in the Introduction:
1. Duration and Magnitude of the Downturn in Stock Market Prices
* Real equity prices are expected to decline for 3.0 years from peak to trough years. This observation suggests, since the S&P 500 Index peaked in October 2007, stock prices may continue to decline until October 2010, if the 2000's credit crisis turns out to be similar to its fifteen predecessors. Through early March 2009-before the stock market rally of the last several months--the current bear market had been declining for 1.4 years.
* In an acknowledgment that real equity prices could further decline beyond the 2009 low value attained on March 9, 2009, and could extend the bear market beyond its 1.4 years to the three-year downturn associated with the benchmark, Reinhart and Rogoff (2009) warn, "The bad news is that these down price cycles typically last for several years. So, even if the big hit on stocks and house prices has come already, the bottom might not be reached until the end of 2010."
* In their recent study of 14 global episodes of bear markets during banking crises, Reinhart and Rogoff found "The average historical decline in equity prices is 55.9 percent, with the downturn phase of the cycle lasting 3.4 years." (Reinhart, 2009). The latter figure represents duration 13 percent longer than the 3.0 years associated with the benchmark.
* The decline in real equity prices for the benchmark is 50%; the upper and lower bound decreases are 38% and 61%, respectively. The expected decreases in stock prices are 51% for the benchmark for Group I and 81 percent for the Great Depression.
* From the S&P 500 Index peak on October 9, 2007, through its current year's low on March 9, 2009, the real total equity return exhibited a decline of 56.0 percent. The real equity price decrease over that period was 57.5 percent. Total return is defined as price-plus-gross-cash-dividend return, as previously noted. [Equity returns and price changes were derived in the Methodology section.]
* The nominal (i.e., non-inflation-adjusted) declines were 55.3 percent for total return--including cash dividends--and 56.8 percent for the equity prices. As noted above, Reinhart (2009) suggests "The average historical decline in equity prices is 55.9 percent"--virtually the same figure as the actual decline of 56.8% already experienced by the S&P 500 Index
* The expected decline in real equity prices of 50 percent associated with the benchmark based on the fifteen banking crises has been exceeded by current events, as represented by the real decline of 57.5 % in the S&P 500 Index attained on March 9, 2009. The 95 % confidence bound of a 61% decline in real equity prices could be regarded as an upper limit to the decline, based on past banking episodes.
* However, I am unwilling to suggest the equity price decline has necessarily reached its trough in early March 2009, marking the end of the bear market, because there are indications of longer downturn duration. As previously noted, the benchmark expects real equity prices to decline for 3.0 years. Also, we may be experiencing a rally with diminished volatility which is likely to be followed by a "retesting" of those lows. Some observers expect increased volatility and a possible severe contraction of perhaps 10-20% (Lim, 2009).
2. Number of Years for Equity Prices to regain their Inflation-adjusted Peaks
* If today's crisis turns out to be similar to the benchmark, a peak-to-peak recovery period of 10.6 years in real prices would be expected. Then, prices could return to inflation-adjusted peaks by mid-2018, from the prior peak of October 9, 2007.
* The recovery period based on the benchmark 15 banking crises could be as low as 6.0 years or as high as 15.2 years. The benchmark for Group I suggests a 6.9-year recovery period. It is worth recalling, following the Great Depression, stock market prices took 26.5 years to regain their pre-1930s inflation-adjusted peaks.
3. Real annual Total Returns required for regaining prior Peaks
* The expected real annual total returns associated with equity prices regaining their inflation-adjusted peaks in 10.6 years-- with a time from trough to recovery of 9 years, assuming the trough in the S&P 500 Index of March 9, 2009--is 9.6%, or about 10%. A real equity price return of 10.0% over 9 years is implied. For comparative purposes to other studies, nominal annual total returns--unadjusted for inflation--and nominal annual equity price returns--unadjusted for inflation and excluding dividends--of 9.3% and 9.8%, respectively, would be expected for the nine-year recovery period. The close proximity of these four returns - lying in a fairly tight band of 9.3-10.0% - is attributable to the future inflation rate, assumed equal to the actual rate of 1.81% experienced over the 17-month peak-to-trough period and adoption of the actual dividend return over this period of about 2.9%, with the latter two figures close to 2%. In the future, the real annual total returns will depend to some extent upon the relative offset of inflation by dividends, and the close association between these four figures of return will likely be loosened.
* However, if the recovery period of 6.9 years associated with the benchmark for Group I were adopted--instead of the 10.6 years of the benchmark--then prices could return to inflation-adjusted peaks by late-2014, from their prior peak of October 9, 2007. Annual real total returns of 16% would be implied for the 5.5 years of recovery, if March 9, 2009 were adopted as the true trough date of the S&P 500 Index.
* The real annual total equity returns of approximately 10% may be compared to the findings of Wharton Professor Jeremy Siegel--the stock market has generated an annual average, inflation-adjusted return of 6.8 % from 1802 through year-end 2006. For post-World War II periods, the compound annual total real equity returns were 6.9% for 1946-2000, 10.0% for 1946-1965, (0.4)% for 1966-81, 13.6% for 1982-99, and 8.4% for 1985-2006. (The 1966-81 period, which exhibited negative compound annual returns, did not follow a severe banking crisis and thus was not a "recovery period.") "This remarkable stability is called the mean reversion of equity returns, which means that returns can be very unstable in the short run but very stable in the long run. ... Periods during which the market fell below the trend line, such as during the early 1980s, pointed to promising future returns." (Siegel, 2008, p. 13).
4. Peak Unemployment Rate
* The October 2009 unemployment rate of 10.2% is the highest since December 1982 (10.8%) and has increased 5.3 percentage points over its trough of 4.9% in December 2007 (Bureau of Labor Statistics, 2009). The latter date is the official beginning of the recession (National Bureau of Economic Research, 2009). Nevertheless, observers might consider it highly unlikely, given our current social net, for the rate to increase an additional 15.7 percentage points to match the 21.0 percentage point increase of the Great Depression. The Great Depression figure cited was treated as an influential outlier, which, in my judgment, is highly unlikely to be experienced again, due to changed external conditions. Therefore, I created an "adjusted benchmark" consisting of all banking crises but excluding the U.S. 1930s episode.
* Peak unemployment of about 11.5% is expected based on the trough-to-peak 6.6 percentage-point increase in the unemployment rate associated with the adjusted benchmark for 14 banking crises. The increase in the unemployment rate for the adjusted benchmark of 6.6 percentage points represents a marked difference from the benchmark's 8.0 percentage points. The former figure of the adjusted benchmark is displayed in Exhibit 2 [with an asterisk], along with its lower and upper confidence bounds of 3.9 and 9.3 percentage points, respectively. Based on these two bounds, unemployment is very likely to peak between 8.8% and 14.2%, with the lower bound already surpassed by October's rate of 10.2%.
* In close agreement with the 6.6 percentage-point increase associated with the adjusted benchmark, Reinhart states, "The unemployment rate rises an average of 7 percentage points over the down phase of the cycle, which lasts on average over four years." (Reinhart, 2009). Reinhart and Rogoff (2009) also state that "If the United States follows the norm of recent crises, as it has until now ... Unemployment will continue to rise for three more years, reaching 11-12 % in 2011." Roubini expects the unemployment rate "to peak at around 11%" and to remain at a very high level for two years or more (Roubini, 2009).
* The unemployment rate was 10.2 % in October, with many economists now expecting a peak of over 10% in 2010 (Herbert, 2009). The government's stress tests for banks used 10.3% as the worst-case unemployment rate for 2010, which may prove to be an insufficiently-low figure. About 8 million jobs have been lost since the recession's inception in December 2007, with nearly 16 million people now unemployed. President Obama's top economic advisor, Lawrence Summers, admitted in July 2009 that job losses and unemployment at the end of the second quarter of 2009 were higher than the government expected (Calmes, 2009). The underemployment rate--which also counts civilians who have given up searching for a job--was 17.5% in October, almost certainly its highest level since the Great Depression (Leonhardt, 2009).
5. Real Decline in the Nation's Output from Peak to Trough
* Real per capita Gross Domestic Product (GDP) is expected to decrease on an annualized basis by 4.0% from peak to trough calendar year, based on the benchmark for Group I. I believe this benchmark, consisting of five major credit contractions associated with developed countries after World War II, may be the most representative benchmark for the current U.S. crisis. This assertion follows from an astute Ruff-Childers observation, "This pattern [of GDP contraction being minimal relative to loan contraction for developed-market crises] indicates that modern developed economies have tended to be less severely affected by crisis-induced credit crunches, in part reflecting the powerful role now played by government fiscal and monetary authorities in responding to financial and economic emergencies." (Ruff and Childers, 2009, p. 6).
* Based on the adjusted benchmark for 13 banking crises--which excludes the U.S. 1930's banking episode deliberately and the U.S. 1870's financial crisis (for which GDP data was unavailable)--real per capita GDP is expected to decrease by 6.7% on an annualized basis. The real GDP decline of 30.8% associated with the Great Depression is unlikely to recur, as previously discussed in the Methodology section. Both the 6.7% and 4.0% GDP declines associated with the adjusted benchmark and Group I, respectively, are depicted in Exhibit 2 [with an asterisk and pound sign].
* Real GDP shrank at an annual rate of 2.6% from 2008 to 2009, based on quarterly GDP data in chained 2005 dollars (Bureau of Economic Analysis, 2009). This actual decrease in year-over-year GDP is well within the confidence bound associated with the expected decline of 4.0% associated with the benchmark for Group I. Actual real GDP declined by 4,0% from the first quarter of 2008 to the first quarter of 2009, according to revised figures from the BEA.
* An unlikely expected total decline of 9.1% in the real GDP is suggested by the upward-biased benchmark for all 15 banking episodes, including the Great Depression. This figure is much higher than the declines of 4.0% and 6.7% associated with the benchmark for Group I and the adjusted benchmark, respectively. This observation may tend to cast some doubt on the applicability of the following valid statement of Reinhart: "As to real per capita GDP around banking crises, the average magnitude of the decline is 9.3%." (Reinhart, 2009). Since the latter average of 9.3% includes the U.S. financial crisis of 1929, it too is unrepresentative as a likely outcome due to its upward-bias.
Comparison of Results
A comparison of the results for each crisis-characteristic associated with the more robust weighted-average benchmark adopted in this paper with the "average" figures depicted in the Ruff-Childers paper reveals:
* a significant difference of 22% in the decline of real EPS;
* a negligible difference (of less than 4%) in the percentage decline for real equity prices; and
* moderate differences of 11 - 15% in the other six characteristics.
For example, there is a difference of 15% in the number of years of peak-to-peak recovery for real equity prices--10.6 years associated with the weighted-average benchmark constructed in this paper vs. 9.2 "average" years in the Ruff-Childers paper. There is an 11% difference in the increase in the unemployment rate from trough to peak--8.0 percentage points associated with the weighted-average benchmark vs. an "average" 7.2 percentage points by Ruff and Childers.
Comparisons of benchmarks in this paper with corresponding average rates generated by Reinhart and Rogoff [R&R] were discussed for 5 characteristics in the Findings section. There is very close agreement in the typical unemployment rate increase (adjusted benchmark with 14 episodes of 6.6 percentage points vs. 7 percentage points of R&R) and the expected peak unemployment rate (adjusted benchmark of 11.5% vs. 11--12% of R&R), while the benchmarks for the decline in stock prices and their associated downturn period are about 13% below the average figures of R&R.
I am uncomfortable with the applicability of the average decline in real per capita GDP of 9.3% displayed by R&R, since the average includes the unlikely-to-recur 30.8% decrease associated with the Great Depression (Reinhart, 2009). Also, the robustness of their results is uncertain, for they did not utilize the weighted-average technique strongly advocated in this paper. Reinhart and Rogoff did not generate annual real total returns needed to regain the inflation-adjusted equity price peaks (Reinhart and Rogoff, 2009a).
This recession is already the longest since the Great Depression of the 1930s. It is unlike previous recessions, as noted by former Federal Reserve Chairman Paul Volker, because it was not brought about by tight credit and high interest rates but by an excess of capital (Volker, 2009).
The current illiquidity, credit and capital crisis was meaningfully viewed from the perspective of fifteen prior banking crises selected by Ruff and Childers
1. That shared the characteristic of "a severe shock to a country's financial system" in terms of impairing the effectiveness of credit intermediaries in maintaining liquidity.
2. Were preceded by inflationary asset bubbles and multiyear credit expansion, followed by credit contractions of roughly the same magnitude.
In addition, this author was successful in addressing five specific issues of interest and general concern which underlie the current mood of uncertainty.
(1) Real equity prices are expected to decline for 3 years from the last peak of the S&P 500 Index in October 2007 to their trough, based on the benchmark generated from the 15 prior banking events. Therefore, stock prices may continue to decline until October 2010, if the 2000's credit crisis turns out to be similar to its predecessors.
(2) A recovery period of 10.6 years in real prices from prior peak to peak would be expected if today's crisis turns out to be similar to the benchmark. Prices could then return to inflation-adjusted peaks by mid-2018, if March 9, 2009, were adopted as the date of the true trough in equity prices.
(3) Real total returns of about 10% annually for nine years would be implied for equity prices to regain their inflation-adjusted peaks, if today's crisis turns out to be similar to the benchmark and assuming March 9, 2009, as the date of the true trough in equity prices.
(4) Peak unemployment of about 11.5% is expected, based on the increase in the unemployment rate associated with the adjusted benchmark for 14 banking crises. Many economists now expect a peak of over 10% in 2010.
(5) Real per capita GDP is anticipated to decrease by 4.0% on an annualized basis from 2008 to 2009, based on the benchmark for Group I, consisting of five major credit contractions associated with developed countries after World War II. Actual GDP decline was 2.6%.
The unprecedented magnitude of government intervention, including possible "jobs creation" programs in 2010, may tend to blunt the severe impacts anticipated in the peak unemployment rate and the duration of the equity market downturn.
August 28, 2009
Revised: January 31, 2010
The author particularly wishes to acknowledge the assistance provided by Mr. Vincent L. Childers in fully explaining the methodology he utilized, as well as for furnishing several unpublished details underlying the Ruff-Childers (2009) paper. In reaction to my suggestion of generating a benchmark crisis by taking a weighted average of the component credit contraction events, he agreed that "some kind of weighted average like you've laid out would be more accurate in trying to really look at an "average" crisis."
Appendix I: Fifteen Systemic Banking Crises * Group I: Developed Countries, Post World War II 1. U.S.--Late 1980s Deregulation contributed to unsound real estate lending, ending in failure of 2700 banks/S&Ls 2. Japan--1990s Cheap credit led to bad loans and massive asset bubble, engulfing banking system as it deflated 3. Norway--Late 1980s Deregulation led to lending and house price boom, ending in recession and eventual banking crisis 4. Sweden--Early 1990s Credit and real estate boom followed financial deregulation, ending in banking/currency crisis 5. Finland--Late 1980s Deregulation led to lending and asset boom that ended in banking/ currency crisis Group II: Open Developing/Emerging Countries, 1990+ 6. South Korea--Late Excessive lending and over investment ended in 1990s banking/currency crisis 7. Argentina--Early Recession and large public debt led to loss of 2000s confidence in currency and run on banks 8. Thailand--Late Economic and real estate boom supported in 1990s part by foreign debt ended with banking/currency crisis 9. Turkey--Early 2000s Bank unwinding of large government bond positions snowballed and resulted in banking/currency crisis 10. Colombia--Late Credit and real estate boom followed financial 1990s liberalization that ended in banking/currency crisis Group III: U.S., 1870--World War II 11. U.S.--1930s Credit and stock market boom ended in crash and eventual banking crises in the Great Depression 12. U.S.--Mid-1910s Outbreak of WWI caused foreigners to sell U.S. assets for gold, resulting in closing NYSE and capital controls 13. U.S.--Early 1900s Failed stock corner resulted in runs on the banks that ended in a rescue by J. P. Morgan 14. U.S.--1890s Lending boom related to railroads overbuilding ended with 600+ bank failures and run on U.S. gold supply 15. U.S.--1870s Railroad- and reconstruction-related lending boom ended with 400+ bank failures and the "Long Depression" Source: International Monetary Fund, National Bureau of Economic Research, AllianceBernstein * Appendix I is a copy of Display 1 (Ruff and Childers, 2009, p. 2). Appendix II: Systemic Banking Crises Crises of Ruff and Childers Crises of Reinhart and Rogoff (2009) (2009a) 1. U.S.--Late 1980s 2. Japan--1990s Japan--1992 3. Norway--Late 1980s Norway--1987 4. Sweden--Early 1990s Sweden--1991 5. Finland--Late 1980s Finland--1991 6. South Korea--Late 1990s Korea--1997 7. Argentina--Early 2000s Argentina--2001 8. Thailand--Late 1990s Thailand--1997 9. Turkey--Early 2000s 10. Colombia--Late 1990s Colombia--1998 11. U.S.--1930s U.S.--1929 12. U.S.--Mid-1910s 13. U.S.--Early 1900s 14. U.S.--1890s 15. U.S.--1870s Spain--1977 Philippines--1997 Malaysia--997 Indonesia--1997 Hong Kong--1997
Bureau of Economic Analysis. 2009. "Current-Dollar and Real Gross Domestic Product" and "Personal Income and its Disposition" tables, National Economic Accounts, U.S. Department of Commerce. http://www.bea.gov/national/index. Accessed in July 2009 and January 2010.
Bureau of Labor Statistics. 2009. "Civilian Unemployment Rate" and "Personal Savings Rate as a Percent of Disposable Personal Income" (NIPA Table 2.1), Bureau of Labor Statistics, U.S. Department of Labor, http://www.research.stlouisfed.org. Accessed in July 2009 and January 2010.
Calmes, Jackie. 2009. "Summers Says Stimulus Plan on Track Despite Job Losses," The New York Times, July 18, 2009.
Conference Board. 2009. Consumer Confidence Survey Press Releases, June 30, July 31 and December 31, 2009.
De la Merced, Michael. 2009. "Profits at Two Banks Mask Signs of More Trouble Ahead," The New York Times, July 18, 2009.
Healy, Jack. 2009. "U.S. Savings Rate at Highest Point in 15 Years." The New York Times, June 26, 2009.
Herbert, Bob. 2009. "No Recovery in Sight," The New York Times, June 26, 2009.
Inflation Data. 2009, "Current Inflation Rate" and "Inflation Calculator" tables based on the Consumer Price Index (CPI-U) as compiled by the U.S. Department of Labor: Bureau of Labor Statistics, as viewed at InflationData.com and accessed July 2009-January 2010.
Johnson, Simon. 2009. "The Quiet Coup," Atlantic Monthly, May 2009.
Leonhardt, David. 2009. "Broader Measure of U.S. Unemployment Stands at 17.5%," The New York Times, November 6, 2009.
Lim, Paul. 2009. "What About the Valley After the Rally?" The New York Times, May 31, 2009.
Lincoln, Abraham. 1858. "House Divided" speech.
National Bureau of Economic Research (2009). "Official Recession Dates," NBER, http://www.nber.org/cycles/, as accessed in May 2009.
Norris, Floyd. 2009. "A Recession Measured by New-House Sales," The New York Times, June 26, 2009.
Rampell, Catherine and Jack Healy. 2009. "U.S. Economic Contraction Slowed in Quarter," The New York Times, July 31, 2009.
Reinhart, Carmen M. 2009. "The economic and fiscal consequences of financial crises," in VOX, Research-based policy analysis and commentary from leading economists, www.VoxEU.org, January 26, 2009.
Reinhart, Carmen M. and Kenneth S. Rogoff. 2008. "Banking Crises: An Equal Opportunity Menace," National Bureau of Economic Research, December 17, 2008.
Reinhart, Carmen M. and Kenneth S. Rogoff. 2009. "Kenneth Rogoff and Carmen Reinhart: The history of banking crises indicates this one may be far from over," Newsweek, March 22, 2009.
Reinhart, Carmen M. and Kenneth S. Rogoff. 2009a. "The Aftermath of Financial Crises," American Economic Review 99(2): 466-72, May 2009.
Roubini, Nouriel. 2008. "Worse than LTCM: Not Just a Liquidity Crisis; Rather a Credit Crisis and Crunch," Global EconoMonitor, July 9, 2008.
Roubini, Nouriel. 2009. "The Worst is yet to Come: Unemployed Americans Should Hunker Down for More Job Losses," rgemonitor.com/roubini-monitor/257978/, November 15, 2009.
Ruff, Jon and Vincent L. Childers. 2009. "Is This Time Different? Systemic Financial Crises and Their Effect on Economies and the Markets," Alliance Bernstein Research report, January 2009.
Siegel, Jeremy J. 2008. Stocks for the Long Run: The Definitive Guide to Financial Market Returns and Long-Term Investment Strategies, Fourth Edition. New York: McGraw-Hill.
Standard & Poors. 2009. "S&P 500 Index and S&P 500 Total Returns Table," Standardandpoors.com, accessed in April--December 2009.
Volker, Paul (2009). "Crisis May Be Worse than Depression: Volcker," Reuters, February 20, 2009.
Wyss, David. 2009. "S&P Economic Outlook," Standardandpoors.com, June 19, 2009.
Serge L. Wind, New York University-SCPS and Keller Graduate School of Business at DeVry Institute of New York email@example.com.
|Printer friendly Cite/link Email Feedback|
|Author:||Wind, Serge L.|
|Publication:||Review of Business|
|Date:||Sep 22, 2010|
|Previous Article:||Testing the trade-off theory of capital structure.|
|Next Article:||IRS offers relief for investors of fraudulent investment schemes.|