Prediction markets are often used to forecast when an event will occur. Our newest question type combines our standard option markets with an automated rolling function to streamline administration, and improve accuracy of these questions.
Before continuous date markets, you’d have to create stocks for each possible timeframe, with the name describing when that time frame began and ended. As time passed and a time frame ended, you would manually have to cash out an expired stocks. This limits the amount of information you’re able to get from a market, since each close out leaves the market with fewer stocks.
With continuous date markets, all the mechanics of running a market, that predicts a date, are taken care of for you. By specifying the time frame (daily, weekly, monthly) and number of time frames between date buckets, our backend takes care of stock name formatting, as well as keeping track of the actual date and time that the stocks begin and end.
Additionally, when a stock date bucket is about to expire, we “roll” the market in to the future. This means we resolve the closing stock to 0 (since it hasn’t happened yet) and we split the last stock of the market into two. This maintains the same granularity in forecasting until the answer is known. When we do this, we make sure to keep trader’s net worth the same, by giving them an equal position in the newly created stock as they did in the previous last stock.