Title: What Causes Persistence of Stock Return Volatility One Possible Explanation with an Artificial Stock
1What Causes Persistence of Stock Return
Volatility? One Possible Explanation with an
Artificial Stock MarketRyuichi
YAMAMOTOBrandeis University
20. Review What are about Persistence of
Return Volatility?
- A well-known feature
- Identified if ARCH process exists. (Engle (1982))
- Evidence (Pagan(1996)).
- Not investigated much about how the volatility
clustering is created.
30. Review What are about an Agent-based
Stock Market?
- Deals with stock market
- A set of interacting heterogeneous agents
- Learning/Adaptation/Evolution
41. Motivation
- Why would be an Agent-based Stock Market useful?
- the evolution of the agents strategies
- the dynamics of the return series with a relation
to the interaction of the heterogeneous
investors. - What is new?
- How volatility persistence is endogenously
created.
5 How is the volatility persistence explained with
an agent-based stock market?
- (Two Steps)
- Agent-based stock markets, which can and cannot
produce the volatility clustering, are examined.
(social and individual learning) - The relation between the persistence and agents
expectation is investigated.
6Review What are Individual Learning and
Social Learning?
Figure 1 Individual Learning
Figure 2 Social Learning
Imitative behavior
Privately distributed
7Results
- A social learning economy produces persistence of
return volatility while an individual learning
economy does not. - Expectations cluster in social learning economy
but not in an individual learning economy. -
8Outline
- 0) Preview
- Motivation
- Market structure
- Computer Simulations
- Conclusion
92. Market Structure (LeBaron et al. (1999))
- 2 tradable assets a stock and a bond.
- The risk-free bond 10.
- The stock pays a dividend
- of shares is 150 of agents in the market.
10 How do events in this artificial market
proceed? 1. Information set
- At time t, agents observe the past price and
dividend, and calculate technical indicators. -
- where k5 and 10.
- Form an information set, .
11- The information set, , includes
- 1)
- 2)
- 3)
- 4)
- 5)
- At time t, dividend, , is revealed and paid.
122. Prediction
13- Permitting to range with the allowable
bounds for and as in LeBaron
(1999), that is, and , - Agents are heterogeneous in terms of their
expectation.
143. Strategy making
4. Price determination
155. Updating wealth
- After revealing the price, wealth, , is
recorded. - Wealth, w, for individual i is evolved according
to
166. Updating Forecast Strategies
- Step 1-5 are repeated for 25 periods.
- Figure 4 Timing of the market
-
- The fitness criterion
- Individual and Social Learning
25 50 .. 500
17Figure 1 Individual Learning
Figure 2 Social Learning
183. Experiments
- Simulations are conducted in each market
separately and are repeated for 10 times. - The series of stock price, dividend, and point
estimates on future prices made by agents,
, are recorded for the last 5,000
periods. - Two steps
19Step 1 Volatility Clustering in Individual and
Social Learning Economies
- Following LeBaron et al. (1999), the estimated
residual series are analyzed.
Table 1 Summary Statistics
ARCH(1)
Note Means over 10 runs. Numbers in brackets are
the fraction of tests rejecting the no ARCH null
hypothesis at the 95 confidence level.
20Step 2 What Causes the Volatility Persistence?
- Why is it in the social learning economy but not
in the individual learning economy? - Difference?
- The important phenomena in social learning is the
following the herd. - Agreement on
- Figure.
-
- The changes in the expectations over time
- (18)
21(No Transcript)
22Figure 4 Dynamics of the Variation of
Expectations and Return Series.
23Relation between dynamics of expectations and
return volatility
- GMM
- An Explanatory variable
- A Dependent variable
24There IS a Relation between dynamics of
expectations and return volatility
- Economics
- When the agents point estimates on future prices
differ, some traders shift expectations up and
others down. - As a result, the trading volume additionally
increases so that the price changes. - A large swing of the expectations is related to
the large change in returns.
25What Causes the persistence of Stock Return
Volatility?
- Volatility clusters in a social learning economy
since expectations shows clustering. - Volatility clusters in an individual learning
economy since expectations doesnt show
clustering.
26Conclusion
- A social learning economy produces persistence of
return volatility while an individual learning
economy does not. - A social learning economy is more likely to show
the agreement of the point estimates on future
prices than in an individual learning economy. - The changes in the expectations explain the
variability of the stock returns in both
individual and social learning economies.
27Conclusion
- If the expectations cluster, the volatility of
the returns series would also show the serial
correlation. This phenomenon is found only in a
social learning economy. - Those of the results indicate that the
persistence of stock return volatility would be
caused endogenously with the behavior of the
heterogeneous investors.