Asset Class Allocation

The winning formula for accurate forecasts

Drawing upon the work of Philip Tetlock's Good Judgment Project (Superforecasting) and analysis of various prediction markets, Signal Sigma employs an optimal timeline in order to create accurate forecasts. In order for any system to make a prediction, it needs to draw upon the recent past.

We have determined that looking back over the past two years gives our system the best shot at forecasting the next 10 months. We have kept this ratio (2 years of history / 10 months of projection) across most of our automated data analysis. The last 2 years of financial statements also carry the most weight when a bank decides to issue a loan. This historical analysis window is widely used in the financial industry when determining how the past informs our view of the future.

Investors have different time horizons and different risk tolerances. It’s easy to say that over a 100-year time horizon, risk assets always appreciate in value. While that is certainly true, near term drawdowns can cause immense emotional hardship and nobody likes to see their hard earned capital evaporate during a crisis event. In recognition of these facts, we believe our method delivers a balanced risk-reward profile.

1.1 Analysis and Investment Horizon