| Forward Premium phenomenon is broadly studied by various references and reports. However, there are rarely studies on the systematic features of the forward premium structure with different terms to maturity. The study on this problem involves various foreign exchange rate theories to explain the movement pattern of forward premiums with different terms to maturity.This article adopts 11 FX spot rate relative to USD, and forward rate with terms to maturity of 1 months, 3 months, 6 months, 9months, 12 months, 18months, 2 years, 3 years and 5 years as dada samples, to analyze the term structure features of forward premium, as well as the impact from different exchange rate regimes. Descriptive statistics results suggest that the standard deviation of most counties'forward premium increases with the term. Compared with FX rate under the free floating regime, the variation of FX rate under managed floating regime is relative larger. In most cases, the standard error of f m ,t- st + m is bigger than that of st + m - st. That is to say, estimation of the FX rate in the future by spot rate has a better effect than that by forward rate.The Risk Premium Model based on the optimal least square (OLS) estimation suggests that the coefficientβ- 2 of most FX rate is obviously different from 1, demonstrating the existence of the"forward premium biasness". Regard with the FX rate under floating FX regime, coefficientβ- 2 under the terms from 1 month to 18 month is less than 0 (accordingly, coefficientβ-1 >1 (β1 = 1-β2)). The risk premium and expected change of foreign exchange rate has a minus covariance. With the increase of the term,β- 2 decreases at first and then increases like a U shape. The coefficientβ- 2 of managed RX rate regime countries is closer to 1, suggesting a relative smaller risk premium coefficientβ-1 (β-1 = 1 -β-2).This article adopts Seemingly Unrelated Regression (SUR) to improve the model, to consider the cross-correlation of the FX rate of different currencies under the same term. The forward premiums with the term to maturity of 1 month, 3 month and 1 year are conformed to the estimated results according to UIP hypothesis best.Since the FX regime of RMB experiences various reform stage, we can use the sample data of RMB to analyze the forward premiums'term structure feature under different FX regimes. The result shows the risk premium coefficientβ?1 decreases with the RMB FX regime reform form managed floating to free floating. |