Idiosyncratic Risk and Creative Destruction in Japan

1
Idiosyncratic Risk and Creative Destruction in
Japan

• Yasushi Hamao
• Jianping Mei
• Yexiao Xu

2

• It was the best of times, it was the worst of
  times, it was the age of wisdom, it was the age
  of foolishness, it was the spring of hope, it was
  the winter of despair, we had everything before
  us, we had nothing before us, we were all going
  direct to Heaven, we were all going direct the
  other way.
• Charles Dickens, A Tale of Two Cities

3
Motivation

• Capital market perspective on Japanese stagnation
  in the 1990s (The Lost Decade)
• Japan show different patterns from the U.S.
• A sharp reduction in firm-level volatility
  relative to market volatility immediately
  following the Japanese market crash.
• While market-wide volatility has increased, there
  is a significant drop in the variation of
  systematic risk across firms.

4

• Lack of Restructuring
• We discover a positive relation between changes
  in aggregate firm-level volatility and corporate
  bankruptcies.
• It appears that increasing bankruptcies after
  1997 have led to higher firm-level volatility.
• These results suggest that the sharp fall in
  firm-level volatility during the 1990-1996 period
  could be due to a lack of corporate
  restructuring.

5

• The Role of keiretsu
• A decrease in industrial production growth tends
  to increase the difference of firm-level
  volatility between non-keiretsu and keiretsu
  firms in 1990-96.
• This means that during the recession period in
  Japan, there was a greater disparity of stock
  performance among non-keiretsu firms, indicating
  the presence of protection among group-affiliated
  firms.

6

• We find similar results when comparing the
  firm-level volatility between firms with main
  banks and firms without main banks.
• Thus, idiosyncratic volatility for firms with
  business group and main bank affiliations is much
  less responsive to economic conditions than that
  of firms without such affiliations

7
Implications

• Macroeconomic models of cleansing recessions,
  (Caballero and Hammour (1994), Eden and Jovanovic
  (1994)) emphasize the impact of firm-level
  volatility on resource allocation during
  recession.
• Durnev, Morck, and Yeung (2001) find that firms
  in industries with greater firm-specific return
  variation exhibit a higher quality of capital
  budgeting (their profitability indices, marginal
  Q ratios, are closer to one).
• Wurgler (2000) demonstrates that equity markets
  do improve the allocation of capital.

8

• A reduction in firm-level volatility and in the
  variation of systematic risk among firms leads to
  a more homogeneous firm valuation, implying that
  Japanese stocks were treated much less
  discriminatingly after the crash. Thus, both
  low-quality and high-quality firms had a similar
  cost of capital.
• The reduction in firm-level volatility may also
  explain the increasing trend in market-wide
  volatility. (Good ones also become like
  lemons.)

9
Data and Background

• Data
• First and Second Sections of the Tokyo Stock
  Exchange (TSE). Monthly individual stock returns
  and volume. Annual financial statements data for
  book value of equity.
• From 1975 to 1999. Subdivided to five 5-year
  periods. The numbers of stocks 1174, 1275,
  1368, 1521, and 1607.

10
GDP Growth Rate 1956-2000
11
(No Transcript)
12
(No Transcript)
13

• Bubble in stock market popped at the beginning of
  1990.
• Slow responses from government and business.
• Banking system still has bad loan problem (as
  much as 100 trillion, 18 of GDP).
• Major bank failures in 1997 and 1998
• End of convoy system.

14
Measuring Idiosyncratic Volatility

• Direct Decomposition Using Asset Pricing Models

15

• Indirect Decomposition (Xu and Malkiel, 2001).
• Total aggregate volatility is calculated
  by value weighting an individual stocks total
  volatility .
• The aggregate idiosyncratic volatility is
  computed as the difference between the total
  aggregate volatility and the market volatility.

16

• Lo and Wang (2000), Cremers and Mei (2002)
  duo-factor model.
• Rit Et(Rit ) f1t bi1 ... fKt biK eit
  
• i
  1,...,N t 1,...,T ,
• Et( Rit ) rft l1t bi1 ... lKt biK,
  
• tit ti di1g1t ... diKgKt xit
  

17
Time-Varying Market Volatility and Idiosyncratic
Volatility

• Overall market Table 1.
• The dynamic behavior of market and idiosyncratic
  volatility
• Visual impression from Figure 2.
• Augmented Dickey-Fuller test for nonstationarity.

18

• Market volatility is stationary for the entire
  sample period.
• The linear trend is positive and statistically
  significant at 1 level.
• Aggregate idiosyncratic volatility ln(?IV) for
  the entire sample period appears to be
  non-stationary, with Augmented Dickey-Fuller test
  statistics of -18.75 compared with the critical
  value of -24.5.
• The trend coefficient for idiosyncratic
  volatility is not significantly different from
  zero.

19

• For idiosyncratic volatility in 1985-1999, the
  ADF test strongly rejects the unit root
  hypothesis.
• There is a statistically significant downward
  trend in the idiosyncratic volatility.
• This is also robust to other measures of
  idiosyncratic volatilities.
• Therefore, the aggregate stock market volatility
  in Japan has trended upwards over the years, but
  firm-level volatility decreased during the 1990s.

20
R2 and Correlation Coefficient

• Average R2 as a measure of market information
  inefficiency (Morck, Yeung, and Yu, 2000 Durnev,
  Morck, Yeung, and Zarowin, 2002 and Wurgler,
  2000).
• Higher R2 implies that the market is less
  informationally efficient.
• Figure 3a shows R2 from market model (24-month
  rolling regression). A sharp fall in information
  efficiency during the post-bubble period and a
  slight improvement in more recent years.

21

• Similar pattern is shown in Figure 3b for average
  pair-wise correlations using past 24 months data.
• Stocks in the Japanese market lost their
  individuality and started to move together after
  1990.

22
Idiosyncratic Volatility and R2 in a Multi-Factor
Model

• Table 4 3 or 4 factors in returns, 3 to 6
  factors in turnover.
• Average R2s from multifactor models sharply rise
  in 1990-94 and 1995-99 periods.

23
Why Did Idiosyncratic Volatility Fall after 1990?

• Main Bank system.
• Banks provide firms with liquidity at times of
  financial distress (Hoshi, Kashap, and
  Scharfstein, 1990).
• Examples of Main Bank rescue Yamaichi Sec.
  (1965), Mazda, Daishowa Paper, etc.
• Keiretsu.
• Sumitomo, Mitsubishi, etc. Presidents council.
• Expensive and tough bankruptcy procedures.

24

• All the above resulted in slow response to
  problems caused by burst bubble.
• Very small number of bankruptcies of listed
  firms. (Figure 4b.) Too large to fail.

25

• Bankruptcies and Idiosyncratic Risk (Table 5).
• Regress annual change in idiosyncratic risk on
  number of bankruptcies and annual growth rate of
  GDP.
• Weak positive relation between ?idio. risk and
  bankruptcies.
• Regress changes in information efficiency
  variables on bankruptcy and DGP growth.
• Negative relation Increase in bankruptcy tends
  to improve efficiency.
• Rise in R2 may be due to lack of restructuring.

26

• Recent increase in bankruptcies may be related to
  decrease in R2 and correlations.
• Some anecdotal evidence Mycal, Daiei, etc.

27

• Keiretsu, Main Bank relation and Idio. risk.
• Regress difference (nonkeiretsu keiretsu), or
  (nonMB MB) of idio. risk on growth rate of
  industrial production.
• Positive relation in pre-bubble and bubble
  period. An increase in IP tends to increase
  difference in firm level volatility.
• Negative relation in post-crash (stagnation)
  period. Lower IP growth is related to higher
  disparity in firm level volatility of
  non-keiretsu (non-MB) vs. keiretsu (MB) firms.
• Keiretsu (MB) firms are less sensitive to bad
  economic conditions protected by keiretsu and
  Main Banks.

28
Vicious Cycle

• Lack of restructuring (bankruptcies) may have
  contributed to a reduction of efficiency of the
  Japanese market.
• This makes it difficult to distinguish the good
  firms from bad ones, causing good ones to suffer,
  too.
• Then further bailout may become necessary.

29
(No Transcript)
30
Conclusions

• More bankruptcies!
• Future topics
• Did reduction of efficiency affect capital
  budgeting?
• Any effect on asset pricing?