Machine Learning will be used to reduce prediction error for future return, which is now 2x larger than for future risk, very large error before & after the bottom of 2008 or 2020 market crash.

Join our effort if you want to crack this hard problem. In the meantime, you will sleep better if you compare, research, then predict, allocate, to plan ahead for both up and down markets.

Exhaustive search for all selections and allocations, to suggest what to buy or sell, so your portfolio will move to the efficient frontier. But will you trust AI robots to handle your hard earned savings? We hope not. This is why we collect all the values in a colored spreadsheet, to let you review and make the final decision.

Great if you can help us develop a better front-end than a colored spreadsheet, while we concentrate on improving accuracy of gain prediction, on producing spreadsheets for other stock markets. Help us make how-to videos, answer questions in our forum, or just spread the word to your friends & relatives if you find this spreadsheet useful.

Each of us will have a personal investment adviser that tirelessly computes risk and predicts return for all the tickers, to see the current and future efficient and swing frontiers, then prepare ahead a ranked list of suggestions, what to buy, hold, sell, with ready explanations to our why... questions.

On days when the market is bullish, your adviser will recommend you to buy undervalues passed their bottoms, then raise cash by selling overvalues passed their tops, so your portfolio is more diversified and converges to next year's efficient frontier.

On days when the market is bearish, your adviser will tell you to stay calm because you already have limit orders to buy more efficient tickers that are now on sale. You must wait for a market bounce to sell at these retracement levels…

Even in a big crash, when fear dominates, the best thing is to buy on sale and not sell, and this is why…

But you can always override, because you know better… The software is good only at searching all scenarios, number crunching like robust statistics, estimations, predictions, but very poor at common sense reasoning like humans.