Malawi Temperature

The serving forecast made larger errors than the no-change baseline in these tests.

Model: Agentic Investor TSFM (foundation model)

Accuracy vs naive

-14%Trails the no-change baseline over 12 tests (12 tests, 4 won)

Direction

58%No clear edge over a coin flip (21 of 36 calls right in 12 tests)

Turning points

3 of 17Anticipated some turns (17 turns in 12 tests, 1 without a turn)

Error vs baseline

14% higherError ratio 1.14; 1.00 is a tie.

Tests won

4 / 12Complete rolling backtests; ties count as no win.

Test period

1989-12-01 to 2024-12-013 observations ahead per test.

80% band coverage

83%Observed outcomes inside the band across all stored test windows.

• Band coverage uses all stored test windows, including model-selection windows when present; it does not determine the confidence label.
• Evaluated 2026-11-15 against the cached history through 2024-12-01.
• This is an evaluation of the serving model; it does not establish a winner across four model classes.