Thank you so much for your replies.I refer to Sami ATTAOUI?s paper ?Hedging Performance of the Libor Market Model: The Cap Market case?. The author indeed did as what bearish said. He constructed the hedging portfolio for a cap as below:?Given this 3-factor model, a delta-based hedge portfolio is composed of three hedge instruments that are common to all caplets. The natural hedge instruments used here are the zero-coupon bonds. The first hedge instrument is a short-maturity bond (3 months). The second instrument is a spread of bonds consisting of short position of short-maturity bond and a long position of long-maturity bond (10 years). Finally, a butterfly composed of long positions of short- and long-maturity bonds and a short position of intermediate-maturity bond (5.5 years) is used to hedge against the curvature risk. We applied the same hedging strategy to both models standard LMM and CEV LMM. The Delta ratio is computed in both cases through Monte-Carlo simulation.?Then I have some further questions. The author only uses a 3-factor model rather than a higher one. From the practical point of view, is it because more factor models need more hedging instruments and the poor liquidity of long-maturity bond makes them fail to be the hedging instruments? So in practice, a 3-factor model balances the availability of hedging instruments and the hedging performance?