Behavioral Forecasting

1
Behavioral Forecasting

• MSE 444 Final Presentation
• Rachit Prasad, Sudeep Tandon, Puneet Chhabra,
  Harshit Singh
• Stanford University

2
Motivation

• Division of Investor Classes
• Fundamentalists Trade on belief in intrinsic
  value of asset
• Chartists Trade on current market trend, and use
  knowledge of previous movement of prices
• Assumptions
• Bounded Rationality Agents cannot assimilate all
  the information in a market, so perfect foresight
  may not hold
• Prediction Based on heuristic techniques
• Fundamentalist Mean
  reversion to intrinsic value
• Chartist
  Extrapolation of historical prices

3
Agent Prediction Model

• Fundamentalists
• Ef(?t,t1S) -? (St St)
• St Asset price at time t
• ? Mean-reversion coefficient
• St Fundamental price at time t
• Chartists
• Ec(?t,t1S) a0 b0t S2i1aisin(bit ci)
• ai, bi, ci constants found by fitting across a
  window of past asset prices

4
Fundamentalist Prediction
5
Chartist Prediction
6
Agents Predictions
7
Market Prediction Model

• wf fundamentalists / investors
• wc chartists / investors
• wf exp(?Pf)/ exp(?Pf) exp(?Pc)
• Pf Risk-adjusted profitability (over
  training period)
• ? Learning rate parameter
• Pf ?Pf - µsf
• µ Risk aversion parameter
• sf Volatility of profits
• E(?t,t1S) wf Ef(?t,t1S) wcEc(?t,t1S)

8
Model Prediction
Fitting Window
9
Dynamic Weight Adjustment
Fundamentalists Dominate
Chartists Dominate
10
Dependence on Learning Rate
11
Estimation of Model Parameters

• Model parameters (?, ? , µ, S) change with
  feedback (profits)
• The optimal parameters found by grid search and
  nonlinear optimization

12
USDJPY Exchange Rate

• Window Length 15
• Transaction Cost 0 01/02/1975 09/26/1979

13
Daily Returns USDJPY

• 01/02/1975 11/15/1985

14
Cumulative Profit USDJPY

• 01/02/1975 09/26/1979

15
Microsoft Stock

• 04/28/1986 09/28/1989

16
Binary Model USDJPY

• 09/05/2000 06/20/2002

17
Constant Parameters USDJPY
18
Conclusions

• Hit-Rate of about 53 is observed across asset
  classes.
• Profits generated are sufficient to overcome
  transaction costs.
• In addition to the base model, various strategies
  were attempted. The binary model showed good
  promise.

19

• Thank You !