Peru Balance Of Trade

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

Model: Agentic Investor TSFM (foundation model)

Accuracy vs naive

-25%Trails the no-change baseline over 7 tests (7 tests, 2 won)

Direction

58%No clear edge over a coin flip (49 of 84 calls right in 7 tests)

Turning points

34 of 46Anticipated most turns (46 turns in 7 tests, 0 without a turn)

Error vs baseline

25% higherError ratio 1.25; 1.00 is a tie.

Tests won

2 / 7Complete rolling backtests; ties count as no win.

Test period

2019-10-01 to 2026-09-0112 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 2026-09-01.
• This is an evaluation of the serving model; it does not establish a winner across four model classes.