Time Series Terminal
p/time-series-terminal
Driving factors and predictions for any time series
J Lee

Time Series Terminal — Driving factors and predictions for any time series

Featured
1
Time Series Terminal continually discovers driving factors for any time series you search for and makes predictions based on them.
Replies
Best
J Lee
Hello, After 50 months of work, this is the first release of Time Series Terminal. At the moment it has two main functionalities: 1. Find out driving factors for a given prediction horizon For example, for 2 day ahead prediction, GameStop GME used to be often the top predictive factor for BTCUSD, as of late May however, BTCUSD is being mostly driven by peer cryptocurrencies as ETHUSD and industrials as EMR and NXPI 2. Make predictions based on identified predictive factors For example, as of Fri 27 May, for the next business day, Nasdaq 100: net 70% models predict down S&P 500: net 80% models predict up DJIA: net 60% models predict up The prediction for Nasdaq 100 is down maybe given that it suffered heaviest retreat year-to-date. The models are not being strongly optimistic over the three indices. Some backtests on the predictions were included in the white paper on the site. At the moment it covers about 200 US stocks, US sector ETFs, major country ETFs, major currencies, US treasury interest rates, major cryptocurrencies and a few commodity ETFs. The update takes place at 4.30am ET the day following the close. In the future we may make it more real-time, expand to more instruments, and allow interactive use to explain the prediction, and much more. At the same time, we’d be keen to hear any feedback. Stay tuned.