Similar to Katin, Igor , Mockus, Jonas , Katina, Joana „Apie eksperimentinį investavimo strategijų tyrimą realiose ir virtualiose finansų rinkose“ (VU MII)
Similar to Katin, Igor , Mockus, Jonas , Katina, Joana „Apie eksperimentinį investavimo strategijų tyrimą realiose ir virtualiose finansų rinkose“ (VU MII) (20)
Polkadot JAM Slides - Token2049 - By Dr. Gavin Wood
Katin, Igor , Mockus, Jonas , Katina, Joana „Apie eksperimentinį investavimo strategijų tyrimą realiose ir virtualiose finansų rinkose“ (VU MII)
1. On the Experimental
Investigation of Investment
Strategies in the Real and
Virtual Financial Markets
Jonas Mockus, Igor Katin, Joana Katina
Šiauliai, 2013 m.
2. Introduction
The optimal financial investment (Portfolio) problem was
investigated by leading financial organizations and
scientists. Nobel prices were awarded for the Modern
Portfolio Theory (MPT) and further developments. The aim
of these works was to define the optimal diversification of
the assets depending on the acceptable risk level.
In contrast, the objective of this work is to provide a flexible,
easily adaptable model of virtual financial markets
designed for the needs of individual users in the context of
utility theory and to present this model on the web. The aim
is to optimize investment strategies. This aim is the new
element of the proposed model and simulation system
since optimization is performed in the space of investment
strategies; both daily and long-term.
3. The PORTFOLIO model of a virtual
financial market
• Predictions:
o AR(p) ( Used for experiment: AR(1), AR(3), AR(6), AR(9) )
o AR-ABS(p) ( Used for experiment: AR-ABS(1), AR-ABS(3), AR-ABS(6), ARABS(9) )
• Buying-selling strategies:
o Short terms
• Strategy No. 1, risk-aware stockholders: buying the best - selling the
losers by three profitability levels
• Strategy No. 2, risk-aware stockholders: buying the best - selling all the
losers
• Strategy No. 3, risk-neutral stockholders: buying the best - selling all the
rest.
• Strategy No. 4, risk-averse stockholders: selling and buying in
proportion to profitability.
4. The PORTFOLIO model of a virtual
financial market
o Buying-selling strategies:
• Long term strategies:
o Maximizing Sharpe Ratio - Modern Portfolio Theory (MPT)
o Defining risk by survival probabilities and individual utility function
o Applying short term strategies in the long term investment
21. Software of the PORTFOLIO model
The PORTFOLIO model is a part of the general on-line
system for graduate studies and scientific
collaboration.
The most recent PORTFOLIO versions are on the websites:
• http://www.getweb.lt/igor/
• http://www.getweb.lt/joana/
the later site is meant for the prediction models.
22. Average profits of eight prediction models by the
first strategy using virtual data
23. Average profits of eight prediction models by the
first strategy using historical data
25. The most profitable portfolio defined using the
first strategy and AR(6) prediction model
26. Conclusions
•
•
•
1. The growing power of internet presents new problems and opens new possibilities for
distant scientific collaboration and graduate studies. Therefore some nontraditional ways
for presentation of scientific results should be defined.
2. The optimization models show the possibilities of some non-traditional ways of graduate
studies and scientific collaboration by creating and using a specific environment for Eeducation
3. After experiment with real data results shows that profit doesn’t relate on prediction
accuracy. Profit does not depend on prediction error.