Transcript Advanced Valuation Analysis Tools and Simulation - cgu-emp

Advanced Valuation Analysis
Tools and Simulation
Brian Stonerock
CGU EMP Independent Study
December Update
Overview
Objective: Evaluate advanced investing and valuation concepts for
investments through the development of robust cutting edge
platform using the latest technologies
 December Update
Project Plan and Progress
Technical Analysis
Technology / Data Sources
Demo
Next Steps
Project Plan
Research and Plan
Develop Framework – Vaadin / Java
Implement Simple Tools
Implement Stock and Technical Analysis
Connect to Historical Servers
Implement Analysis Tools
 Data Mining (IP)
 Back Casting (IP)
 Bubble Bursting
 Documentation and Deployment
The Potential Rewards
How can market timing can benefit returns?
The only problem is that you have to be
very good at it….
Alternative Market Strategies (1964 to 1984)
Strategy
Buy and Hold
Avoid Bear Markets
Long and Short Major Swings
Long and Short Every 5% Swing
Avg. Annual Gain
11.46%
21.48%
27.99%
93.18%
Based on work from Norman Fosbeck 1984
$10,000 Grows To
$
87,500
$
489,700
$
1,391,200
$ 5,240,000,000
The Potential Rewards (Cont)
The benefit of being smart enough to miss the worst
5 days of the year between Feb ‘66 and Oct ‘01
Source: “The Truth About Timing,” by Jacqueline Doherty, Barron’s (November 5, 2001)
Technical Analysis
 Technical analysis: The attempt to forecast stock
prices on the basis of market-derived data
 Technicians (also known as quantitative analysts
or chartists) usually look at price, volume and
psychological indicators over time
 Basic Tools
Breakout
Trend Lines
Moving Averages
Price Patterns
Indicators
Cycles
Support
Resistance
Technical Indicators
 There are, literally, hundreds of technical indicators used
to generate buy and sell signals
 We will look at just a few that I use:
 SMA – Simple Moving Average
 EMA – Exponential Moving Average
 RSI - Relative Strength Index (by Welles Wilder)
 0 to 100 measurement the speed and change of price
movements, >70 overbought and <30 oversold
 MFI - Money Flow Index
 Similar to RSI but volume weighted
 CCI - Commodity Channel Index
 Identifies cyclical turns in commodities seeking overbought
and oversold conditions
Technology Overview
Vaadin
Java / Tomcat
JFreeChart
Data Sources
JStock
Interactive Brokers
Trader Work Station
JBookTrader
http://code.google.com/p/cgu-emp
Technology Overview
Vaadin Architecture
http://vaadin.com
Technology Overview
Development Process
Technology Overview: Eclipse
Dynamic Web Project
Data Sources
Real Time & Historical Data Servers
Interactive Brokers
Yahoo EOD, ID for various all countries
Google EOD
Tickers, Quotes, and more
Demo
Next Steps: Emotionless Trading
Back Casting
 JStockTrader Demo
Bollinger Bands Example
Source: Stock Market Prediction Using Online Data:Fundamental and Technical Approaches By Nikhil Bakshi (2008)
Next Steps (Cont): Predicting Bubbles
"the basic intuition is straightforward: if the reason that the price is high today is only because investors believe that the
selling price will be high tomorrow-when "fundamental" factors do not seem to justify such a price-then a bubble exists."
(Stiglitz 1990, p 13)
 Ideal Type 1: Pure Speculative Bubble
 Asset price today is too high and the price eventually will fall…. Speculators believe
that the price will continue to rise for some time, with potential to sell with a profit
before the price falls
 Ideal Type 2:Irrational Expectations Bubble
 Speculators become overoptimistic and think the price will continue to grow rapidly.
The growth is expected to outperform history or fundamentals…. Therefore it seems
rational to pay a high price
 Ideal Type 3: Irrational institutions Bubble
 Principal-agent problem, where
Speculators have incentives to pay higher
prices than what is supported by historical
patterns or strong evidence
Source: Price Bubbles on the Housing Market: Concept, theory and indicators Hans Lind (2008)
Next Steps (Cont)
Bubble Equation
9 Parameter equation that
requires iterative “fitting”
algorithm to predict falls
http://frog-numerics.com/blog/2009-12_blog.html
Source: D. Sornette and A. Johansen ('Large Financial Crashes', Physica A 245,pp. 411-422, 1997)