1\begin{table}[tp]2\centering3\footnotesize4\begin{threeparttable}5\caption{Out-of-sample $R^2$ of the daily factor changes under alternative targets and regressors, by sub-period}6\label{tab:fidelityalt}7\setlength{\tabcolsep}{3pt}8\begin{tabular}{p{3.9cm}rrrrrrrrrrrrrrr}9\toprule10Target / regressors & \multicolumn{3}{c}{2011--2026} & \multicolumn{3}{c}{2011--2015} & \multicolumn{3}{c}{2016--2019} & \multicolumn{3}{c}{2020--2021} & \multicolumn{3}{c}{2022--2026} \\11\cmidrule(lr){2-4}\cmidrule(lr){5-7}\cmidrule(lr){8-10}\cmidrule(lr){11-13}\cmidrule(lr){14-16}12 & L & S & C & L & S & C & L & S & C & L & S & C & L & S & C \\13\midrule14GSW zero curve, 1--30 years & 0.95 & 0.83 & 0.86 & 0.94 & 0.85 & 0.81 & 0.96 & 0.80 & 0.89 & 0.96 & 0.81 & 0.77 & 0.96 & 0.82 & 0.91 \\15CMT, maturities $\geq 1$ year & 0.95 & 0.74 & 0.79 & 0.95 & 0.87 & 0.81 & 0.94 & 0.67 & 0.75 & 0.94 & 0.71 & 0.61 & 0.96 & 0.66 & 0.83 \\16CMT, maturities $\geq 2$ years & 0.95 & 0.70 & 0.76 & 0.95 & 0.75 & 0.75 & 0.95 & 0.50 & 0.66 & 0.93 & 0.67 & 0.71 & 0.96 & 0.71 & 0.81 \\17Ceiling: cash 2, 5, 10, 30-year CMT yields & 0.99 & 0.85 & 0.84 & 0.99 & 0.96 & 0.91 & 0.98 & 0.82 & 0.75 & 0.99 & 0.82 & 0.82 & 0.97 & 0.76 & 0.85 \\18Six roots $+$ SR3 (estimates from 2021Q1) & 0.95 & 0.76 & 0.78 & -- & -- & -- & -- & -- & -- & 0.95 & 0.93 & 0.75 & 0.95 & 0.74 & 0.78 \\19\bottomrule20\end{tabular}21\begin{tablenotes}[flushleft]\footnotesize22\item \textit{Notes:} Out-of-sample $R^2$ of daily factor changes (level L, slope S, curvature C) on the mimicking-portfolio prediction, same rolling protocol as Table~\ref{tab:fidelity} (weights estimated on the previous three years, re-estimated quarterly, applied to the following quarter). Rows change the daily target (Nelson--Siegel factors from the G\"urkaynak--Sack--Wright zero curve; from CMT yields of at least 1 or 2 years) or the regressors (the four cash constant-maturity yields 2, 5, 10 and 30 years instead of futures, which removes the futures basis and gives an upper bound; the six futures plus the front 3-month SOFR future SR3). In the SR3 row the columns headed 2011--2026 and 2020--2021 cover 2021Q1--2026Q4 and 2021 only, respectively, and are not comparable with the same columns of the other rows or of Table~\ref{tab:fidelity}; compare the 2022--2026 column. L, S and C denote level, slope and curvature. Entries are point estimates without confidence intervals. Pre-registered target: Table~\ref{tab:fidelity}.23\end{tablenotes}24\end{threeparttable}25\end{table}26