Record
Two records, never mixed: the live record scores the forecasts this site published, as their target days close; the backtest record replays the past with the same models and rules. Figures are per tier: the mean over its assets, each weighted equally.
Live record
1 daily forecast since 2026-11-15, scored as of 2026-11-15. No day missed.
Does TimesFM 3.0 beat naive?
| Tier | Horizon | Scored days | Ratio | 90% interval | Verdict |
|---|---|---|---|---|---|
| control | 1 | 0 of 180 | – | – | too early to tell |
| control | 7 | 0 of 180 | – | – | too early to tell |
| control | 30 | 0 of 180 | – | – | too early to tell |
| large | 1 | 0 of 180 | – | – | too early to tell |
| large | 7 | 0 of 180 | – | – | too early to tell |
| large | 30 | 0 of 180 | – | – | too early to tell |
| mid | 1 | 0 of 180 | – | – | too early to tell |
| mid | 7 | 0 of 180 | – | – | too early to tell |
| mid | 30 | 0 of 180 | – | – | too early to tell |
| micro | 1 | 0 of 180 | – | – | too early to tell |
| micro | 7 | 0 of 180 | – | – | too early to tell |
| micro | 30 | 0 of 180 | – | – | too early to tell |
Backtest record
A replay, not live forecasts: every day of the past, each model forecast from the data before
that day only. Run backtest_2026-10-21_20261026T160033Z, data to 2026-10-21, enigma
6c03bbb, timesfm 3.0.2,
torch 2.14.1+cpu, statsforecast 2.1.1.
Verdicts
| Tier | Horizon | Models | Ratio | 90% interval | Verdict | Forecasts |
|---|---|---|---|---|---|---|
| control | 1 | timesfm-2.5 / naive | 0.999 | 0.986 – 1.010 | no detectable difference | 5308 |
| control | 1 | timesfm-3.0 / naive | 0.982 | 0.972 – 0.990 | better | 5308 |
| control | 1 | timesfm-3.0 / timesfm-2.5 | 0.983 | 0.977 – 0.989 | better | 5308 |
| control | 7 | timesfm-2.5 / naive | 0.982 | 0.961 – 1.008 | no detectable difference | 5302 |
| control | 7 | timesfm-3.0 / naive | 0.980 | 0.957 – 1.007 | no detectable difference | 5302 |
| control | 7 | timesfm-3.0 / timesfm-2.5 | 0.998 | 0.986 – 1.012 | no detectable difference | 5302 |
| control | 30 | timesfm-2.5 / naive | 0.976 | 0.929 – 1.037 | no detectable difference | 5279 |
| control | 30 | timesfm-3.0 / naive | 1.007 | 0.958 – 1.064 | no detectable difference | 5279 |
| control | 30 | timesfm-3.0 / timesfm-2.5 | 1.031 | 0.995 – 1.066 | no detectable difference | 5279 |
| large | 1 | timesfm-2.5 / naive | 1.035 | 1.024 – 1.046 | worse | 15947 |
| large | 1 | timesfm-3.0 / naive | 1.016 | 1.012 – 1.019 | worse | 15947 |
| large | 1 | timesfm-3.0 / timesfm-2.5 | 0.982 | 0.972 – 0.991 | better | 15947 |
| large | 7 | timesfm-2.5 / naive | 1.060 | 1.028 – 1.085 | worse | 15947 |
| large | 7 | timesfm-3.0 / naive | 1.031 | 1.019 – 1.042 | worse | 15947 |
| large | 7 | timesfm-3.0 / timesfm-2.5 | 0.973 | 0.952 – 0.994 | better | 15947 |
| large | 30 | timesfm-2.5 / naive | 1.044 | 1.006 – 1.077 | worse | 15947 |
| large | 30 | timesfm-3.0 / naive | 1.034 | 1.013 – 1.052 | worse | 15947 |
| large | 30 | timesfm-3.0 / timesfm-2.5 | 0.991 | 0.972 – 1.013 | no detectable difference | 15947 |
| mid | 1 | timesfm-2.5 / naive | 1.024 | 1.013 – 1.034 | worse | 4742 |
| mid | 1 | timesfm-3.0 / naive | 1.012 | 1.004 – 1.020 | worse | 4742 |
| mid | 1 | timesfm-3.0 / timesfm-2.5 | 0.989 | 0.979 – 0.997 | better | 4742 |
| mid | 7 | timesfm-2.5 / naive | 1.060 | 1.036 – 1.088 | worse | 4742 |
| mid | 7 | timesfm-3.0 / naive | 1.029 | 1.006 – 1.049 | worse | 4742 |
| mid | 7 | timesfm-3.0 / timesfm-2.5 | 0.971 | 0.936 – 1.000 | better | 4742 |
| mid | 30 | timesfm-2.5 / naive | 1.109 | 1.047 – 1.178 | worse | 4742 |
| mid | 30 | timesfm-3.0 / naive | 1.055 | 0.996 – 1.119 | no detectable difference | 4742 |
| mid | 30 | timesfm-3.0 / timesfm-2.5 | 0.952 | 0.869 – 1.019 | no detectable difference | 4742 |
| micro | 1 | timesfm-2.5 / naive | 1.016 | 1.008 – 1.026 | worse | 6453 |
| micro | 1 | timesfm-3.0 / naive | 1.007 | 0.998 – 1.016 | no detectable difference | 6453 |
| micro | 1 | timesfm-3.0 / timesfm-2.5 | 0.991 | 0.982 – 0.999 | better | 6453 |
| micro | 7 | timesfm-2.5 / naive | 1.023 | 1.006 – 1.044 | worse | 6453 |
| micro | 7 | timesfm-3.0 / naive | 1.007 | 0.993 – 1.020 | no detectable difference | 6453 |
| micro | 7 | timesfm-3.0 / timesfm-2.5 | 0.985 | 0.971 – 0.997 | better | 6453 |
| micro | 30 | timesfm-2.5 / naive | 1.033 | 0.992 – 1.080 | no detectable difference | 6453 |
| micro | 30 | timesfm-3.0 / naive | 1.005 | 0.967 – 1.046 | no detectable difference | 6453 |
| micro | 30 | timesfm-3.0 / timesfm-2.5 | 0.973 | 0.930 – 1.011 | no detectable difference | 6453 |
Every model, backtest
MASE (lower is better; the lowest per row in bold)
| Tier | Horizon | timesfm-3.0 | timesfm-2.5 | auto-arima | auto-ets | naive | seasonal-naive |
|---|---|---|---|---|---|---|---|
| control | 1 | 0.703 | 0.715 | 0.754 | 0.715 | 0.716 | 1.577 |
| control | 7 | 1.545 | 1.547 | 1.841 | 1.577 | 1.576 | 1.576 |
| control | 30 | 2.748 | 2.664 | 4.062 | 2.727 | 2.730 | 2.916 |
| large | 1 | 1.443 | 1.469 | 1.468 | 1.423 | 1.420 | 3.871 |
| large | 7 | 4.089 | 4.202 | 4.114 | 3.998 | 3.965 | 3.965 |
| large | 30 | 9.853 | 9.947 | 10.369 | 9.842 | 9.527 | 10.409 |
| mid | 1 | 0.764 | 0.772 | 0.767 | 0.757 | 0.755 | 2.058 |
| mid | 7 | 2.130 | 2.195 | 2.152 | 2.071 | 2.070 | 2.070 |
| mid | 30 | 4.787 | 5.029 | 4.842 | 4.543 | 4.537 | 5.040 |
| micro | 1 | 0.790 | 0.798 | 0.822 | 0.795 | 0.785 | 2.234 |
| micro | 7 | 2.252 | 2.287 | 2.393 | 2.252 | 2.237 | 2.237 |
| micro | 30 | 4.705 | 4.835 | 5.379 | 4.740 | 4.680 | 5.160 |
Relative error (absolute error as a share of the last close; lowest in bold)
| Tier | Horizon | timesfm-3.0 | timesfm-2.5 | auto-arima | auto-ets | naive | seasonal-naive |
|---|---|---|---|---|---|---|---|
| control | 1 | 0% | 0% | 0% | 0% | 0% | 0% |
| control | 7 | 0% | 0% | 0% | 0% | 0% | 0% |
| control | 30 | 1% | 1% | 1% | 1% | 1% | 1% |
| large | 1 | 3% | 3% | 3% | 3% | 3% | 9% |
| large | 7 | 9% | 9% | 9% | 9% | 9% | 9% |
| large | 30 | 23% | 23% | 23% | 23% | 23% | 25% |
| mid | 1 | 4% | 4% | 4% | 3% | 3% | 10% |
| mid | 7 | 10% | 10% | 10% | 10% | 10% | 10% |
| mid | 30 | 23% | 23% | 24% | 21% | 21% | 23% |
| micro | 1 | 4% | 4% | 4% | 4% | 4% | 11% |
| micro | 7 | 11% | 12% | 12% | 11% | 11% | 11% |
| micro | 30 | 25% | 26% | 29% | 25% | 25% | 27% |
Direction (share of up or down calls that were right; naive makes none)
| Tier | Horizon | timesfm-3.0 | timesfm-2.5 | auto-arima | auto-ets | naive | seasonal-naive |
|---|---|---|---|---|---|---|---|
| control | 1 | 57% | 56% | 54% | 55% | – | 56% |
| control | 7 | 57% | 57% | 53% | 52% | – | – |
| control | 30 | 61% | 62% | 56% | 53% | – | 57% |
| large | 1 | 49% | 49% | 50% | 52% | – | 51% |
| large | 7 | 49% | 50% | 49% | 51% | – | – |
| large | 30 | 50% | 52% | 50% | 50% | – | 48% |
| mid | 1 | 51% | 51% | 46% | 51% | – | 50% |
| mid | 7 | 50% | 48% | 46% | 51% | – | – |
| mid | 30 | 50% | 45% | 45% | 50% | – | 49% |
| micro | 1 | 51% | 54% | 52% | 51% | – | 50% |
| micro | 7 | 51% | 50% | 52% | 50% | – | – |
| micro | 30 | 50% | 48% | 51% | 51% | – | 49% |
80% interval coverage (calibrated at 80%)
| Tier | Horizon | timesfm-3.0 | timesfm-2.5 | auto-arima | auto-ets | naive | seasonal-naive |
|---|---|---|---|---|---|---|---|
| control | 1 | 80% | 87% | 92% | 93% | 93% | 91% |
| control | 7 | 80% | 81% | 90% | 94% | 96% | 91% |
| control | 30 | 82% | 83% | 87% | 96% | 97% | 94% |
| large | 1 | 81% | 87% | 83% | 87% | 84% | 82% |
| large | 7 | 81% | 80% | 81% | 84% | 83% | 82% |
| large | 30 | 78% | 78% | 78% | 80% | 80% | 79% |
| mid | 1 | 84% | 90% | 91% | 86% | 91% | 90% |
| mid | 7 | 80% | 82% | 90% | 85% | 91% | 90% |
| mid | 30 | 75% | 77% | 89% | 84% | 89% | 88% |
| micro | 1 | 81% | 87% | 93% | 89% | 93% | 93% |
| micro | 7 | 79% | 79% | 92% | 86% | 92% | 93% |
| micro | 30 | 78% | 78% | 92% | 86% | 92% | 92% |