Tradelegs AI Logic and Search Optimization seeks to select the optimal basket of options at each roll date, maximizing performance objectives, subject to explicit risk controls, given current and expected market conditions. The possible combinations of Puts and Calls, long and short, strike price, maturity and quantity, all relative to possible changes in market variables that can affect the value of an options position over time, result in a search of > 3*10^47 combinations. This is not amenable to conventional techniques. The system can also identify when no solution exists that satisfy the risk constraints.
AI Deep Learning and proprietary options analytics significantly enhance the accuracy of modeling options behavior compared with conventional approaches.
Adaptive Strategies
The Optimize Advisors strategies are defined by performance objectives and risk constraints, rather than by static options selection rules.
At each roll date, the Tradelegs system accounts for current market parameters such as underlier price, implied volatility, interest rates, and the liquidity of different options. It also performs a forward analysis of probable and stressed changes to these.
The Optimize Advisors strategies adapt to changing market conditions. They are the first systematic strategies that are at the same time active and non-discretionary.
Any approaches that do not (a) take a forward look at potential and stressed market conditions, (b) accurately model options behavior, and (c) perform a complete search that results in optimal basket selections, expose clients to more risk and/or leave money on the table.
Reduced backfitting bias
Using AI Logic and Search Optimization, and an analysis of multiple potential future paths, allows the Tradelegs platform to avoid the over-reliance on market history associated with using Deep Learning alone.
At each roll date, the Tradelegs system accounts for current market parameters such as underlier price, implied volatility, interest rates, and the liquidity of different options. It also performs a forward analysis of probable and stressed changes to these.
Our backtests simulate what the system would have selected in the past if it were limited to the data available to it at the time. This is in marked contrast to conventional backtests, which rely on historic data analysis to identify options selection rules that would have worked in the past, but may significantly degrade performance when market conditions change. This approach aims to achieve robust backtests.
