Enter apple or AAPL to search

Use ; to split multiple search items

Leave blank and press "Run" to go Home


6m ahead

Predictive Factors

The engine identifies what we call the causally predictive factors: a small set of assets judged to be the most important driving forces behind your selected item's forecast.


2024-06-23T05:43:55.050929 image/svg+xml Matplotlib v3.7.3,

Nvidia Corporation

2024-06-23T05:43:55.103589 image/svg+xml Matplotlib v3.7.3,

China / U.S. Foreign Exchange Rate USDCNY

2024-06-23T05:43:55.152750 image/svg+xml Matplotlib v3.7.3,

SPDR S&P Midcap 400 ETF Trust

2024-06-23T05:43:55.199711 image/svg+xml Matplotlib v3.7.3,


2024-06-23T05:43:55.249867 image/svg+xml Matplotlib v3.7.3,

as of

Our Probabilistic Prediction

Based on the predictive features above, we run many models repeatedly, each predicting a set of possible outcomes for the date in 6m and aggregate them.

2024-06-23T05:43:55.753250 image/svg+xml Matplotlib v3.7.3,
Goodness of Probabilistic Prediction

Test of hypothesis that a sample of the realised probability level is uniformly distributed between 0 and 1 (whitepaper), using significance level 0.001 for the p-value. The user is free to use another significance level e.g. 0.01, 0.05 for the test.


🟢 p-value: 0.002 > 0.001

2024-06-23T05:43:55.892800 image/svg+xml Matplotlib v3.7.3,

Each model casts a single vote on the likely direction of change. If one actually makes trades, a trade will be opened every day with the size in number of shares/lots proportional to the majority vote and kept open for 6m. The currently open trades result in the position for the next end-of-day time point (when the market is open),

% Up % Down % Un­de­cided
SPDR S&P 500 ETF 77% 6% 17%

with respect to the following time points,

Time Value
SPDR S&P 500 ETF Thu 2023-12-14 UTC 472.01

Performance of Past Trades

The overall profit and loss of the trades opened in the above manner, in contrast to the buy-and-hold strategy of one constant share, would have been as plotted below. The Sharpe ratio is calculated on returns measured as the difference between values rather than in percentage, assuming zero trading cost.

Active Buy-and-Hold
Sharpe ratio
(the higher the better)
1.5 1.3
Solvency II Value-at-Risk 99.5%
(the lower the better)
- -
2024-06-23T05:43:56.483729 image/svg+xml Matplotlib v3.7.3,