Mission for Better Portfolio Risk Management
Unisolver Ltd develops unique decision support solutions for recognising and eliminating risk factors in investments and for planning transactions. Our belief is that investing decision process has to start out from vision of possible macroeconomic scenarios. We use as our key tools fundamental analysis accompanied with future scenario forecasts of operating environment circumstances. The scenario options affect into timing and choice of instruments. We use mean-variance optimisation to find most efficient balancing of portfolios taking into account the various scenarios and the selected risk levels.
Basically we believe in the usability of fundamental analysis and classical company valuation models in investment decision making. How interest rate or energy cost increases would affect to the profitability of a company and to its fair price? How to rate IFRS-based asset valuation induced changes in company results? To these kinds of questions we are interested in finding answers.
Today's market phenomena caused by excess liquidity due to lax interest rates have made us cautious in just relying on fundamentals. Recent market behaviour reflects as much the financial health of the market participants, including leveraged hedge funds and derivatives market makers, as the underlying real-world fundamentals of companies and macroeconomics. There is plenty of room for systemic hazards in today's financial markets. Efficient market hypothesis continues to hold on the average and in the long run. However, market bubbles and turbulences may dislocate prices much temporarily. We are helping to make rational decisions also in those special market situations.
Our financial engineering toolbox includes:
investing strategy planning using machine learning, e.g. rule synthesis, inferential statistics and simulation
qualitative company screening using soft computing techniques
company pricing with respect to scenarios of fundamentals and market situations
(re)balancing of portfolios using mean-variance optimisation
Our background is in artificial intelligence, language engineering and software engineering starting from the late 1970's. Knowledge in automata theory and machine learning of grammars helps acquire and test strategies based on diverse time-series inputs. Experience also includes systems for planning and executing complex transactions and decision support systems for multiple-goal optimisation in highly combinatorial cases, optimising using interval arithmetics etc.
Updated September 20, 2006