Weekly competition

We make predictions on selected asset prices by asking a pool of traders and investors what they think about what might happen, and how others around them think about what might happen. We then eliminate individual biases through our proprietary network analysis which is the cornerstone of the BASON®.

Our advantage over similar approaches that use surveys of institutional or individual investors, or surveys of forecasters to predict macro trends is our proven accuracy (<1% error in predicting Brexit and Trump in 2016, and Biden in 2020, along with consistent weekly accuracy of predicting S&P500 and the Dow Jones Industrial at <2% error for over 110 weeks during 2021, 2022, 2023, and 2024) alongside our proprietary innovation that is able to clearly discern signals from the noise. We successfully solve the representative sampling and non-response bias issues by using a combination of the wisdom of crowds and network analysis to eliminate groupthink bias. Our approach significantly reduces the margin of error compared to standard surveys and is commercially proven to accurately predict elections, consumer behavior, product demand, and even the post-COVID recovery and vaccine rollout.

We are perfectly aware that no single individual can predict markets. However, the logic applied to predicting markets is similar to that of predicting elections – no single person in any of our surveys ever gets it right within 1%, but collectively, once corrected for their biases, the crowd delivered stunning accuracy.

What are we predicting?

We make regular weekly predictions for the values of three market indices (S&P500, Dow Jones, the NASDAQ 100 ETF, QQQ), the volatility index (VIX), the 10-year Treasury bond yield, and the AAPL stock.

The BASON® Survey is based on two pillars:

  • Wisdom of crowds:

    what you think will happen (e.g. S&P target for end of week), and what others think will happen => Bayesian self-correction

  • Network analysis:

    homogenous vs heterogenous groups – looking for best observers, eliminate
    echo chamber bias (e.g. bearish or bullish)