Great Events
Section 4.5 concludes the rejection of the random walk null hypothesis for the four frequencies and provides the supporting evidence of mean reversion for low frequency data. 1n Section 4.6, this study investigates whether the great events contaminate the above conclusion in Section 4.5 and compares the differences of the influences among the five great events. The empirical results are reported in Table 7.
Panels C and D of Table 7 present that the majority of 1 +Mr(q) for the monthly and quarterly retums on commercial papers surrounding the five great events are significantly different from 1, indicating the rejection of random walk null hypothesis. The majority of 1+Mr(12) [and 1+Mr(16)] for the daily and weekly retums on commercial papers around the five great events, as shown in Panels A and B in Table 7, are also significantly 企om 1. The above results reveal that both before and after the five great events the four 企equencies of commercial paper rates do not follow random walk, which is consistent with the testing results of random walk hypothesis in Table 6. That is, the five great events do not contaminate the conclusion regarding random walk in Section 4.5.
1n contrast to the findings of mean reversion in Table 6, the majority of 1 +Mr(2) after the five great events in Table 7 are larger than 1 even for the low frequency data, indicating the positive first-order autocorrelation coe宜icients for the commercial paper rates across frequencies after the five great events.
Moreover, Table 7 presents that the differences of Lo and Mackinlay's (1988) variance ratios between before and after the five great events are numerically significant. When the five great events are viewed as a single event, the first three lines and the last three lines in Panels A-D of Table 7 reveal similar results with those of five individual great events. Therefore, as the findings in Tables 3 and 5, the great events have a big impact on the tendency of mean reversion.
Chiao Da Management Review Vol. 33 No.1, 2013 99
Table 7
Lo and Mackinlay's (1988) Variance Ratios of Commercial Paper Rates across Frequencies for the Periods before and after the Five Great Events
Number q of Base Observations Accumulated to Compute Variance Ratio
Before Presidential Election 0.960 1.019 0.985 0.859 0.706
(1982/01105 to 2000/03117) (-0.882) (0.302) ( -0.233) (-2.248)** (-4.866)**
After Presidential Election 1.181 1.233 1.206 1.295 1.361
AU
Chiao Da Management Review Vol. 33 No.1, 2013 101
Before Presidential Election 0.801 0.575 0.483 0.420 0.386 (1982/01105 to 2000/03/17) (-2.237)** (-2.887)** (-2.998)** (-3.298)** (-3.555)**
After Presidential Election 1.296 1.700 2.002 2.140 2.078 (2000/03/20 to 2012/02/22) (2.344)** (4.101)** (5.462)** (5.696)** (5.214)**
Before 911 Terrorist Attacks 0.803 0.576 0.482 0.421 0.393 (1982/01105 to 2001109/11) (-2.225)** (-2.898)申* (-3.024)** (-3.315)** (-3.542)**
After 911 Terrorist Attacks 1.307 1.827 2.142 2.086 1.923
(After Five Great Events) (2.365)** (5.142)** (6.524)** (5.606)** (4.632)**
(2001109/12 to 2012/02/22)
Note: Lo and Mackinlay's (1988) variance ratios, I+M,{q), are reported in the main rows. Figures
in parentheses are the standardized test statistic of M,{q) [ i.e., Z*(q)]. “*" and “**" refer to significant at the 10 and 5 percent level, respectively.
5. Conclusions
To get a more complete picture of the behavior of commercial paper rates, this study investigates the influence of data 企equencies and great events on commercial paper rate behavior. The empirical results show that first, the levels of commercial paper rates across the four different frequencies are all crucial to the determination of their volatility, and that the commercial paper rate process is stationary. Second, the double exponential or Laplace is a more suitable density function for the daily and weekly retums on the commercial papers. Third, the commercial paper rates across the four different 企equencies all reject the null hypothesis of random walk. Finally, the commercial paper rates across the four
different 企equencies all exhibit mean reversion over a
“
very" long period of time, and a contrarian s甘ategy will gain excess retums. This study believes that the reason for the mean reversion process of commercial paper rates comes 企omincorrect expectations of noise traders regarding the fair price of commercial papers when they encounter an information shock. Specifically, noise 甘aders
usually cannot make a correct valuation when faced with information shocks, or without the ability to predict the fair price of a commercial paper. Therefore, when a nOlse 仕ader does business in the commercial paper market 企equently， it may lead to a deviation between the market price of a commercial paper and its fair price, and result in a mean reversion.
Data Frequencies and Great Events
In terms of the effect of data frequencies on commercial paper rate behavior, the exhibition of mean reversion tendency and the rejection of the random walk nul1 hypothesis for low frequency data are s甘onger than those for high frequency data. However, the data frequencies have a limited impact on the sensitivity of commercial paper rate volatility to its level and the shape of the dis甘ibution of commercial paper rates. The findings across the four frequencies give us a more complete picture of commercial paper rate behavior.
F or the impact of great events on commercial paper rate behavior, the empirical evidence across frequencies indicates that there is an obvious difference of mean-reverting speed between before and after the five great events. In addition, the Taiwan Presidential election in 2000 had the weakest effect on the commercial paper rate behavior among the five great events. This study conc1udes t4at because the result of the presidential election in 2000 did not deviate from the expectation of investors in the money market, this political event had less impact on commercial paper rate behavior than the other four. This suggests that the variation of commercial paper rates depends on fundamental economic factors and investor sentiment changes, and the investors in the Taiwan' s commercial paper market should pay more attention to the events causing a substantial economic loss and over-pessimistic sentiments.
References
Adkins, L. C. and Krehbiel, T. (1999), “Mean Reversion and Volatility of Short-Term London Interbank Offer Rates: An Empirical Comparison of Competing Models," lnternational Review
01
Economics & Finance, 8(1), 45-54.Balvers, R. J., Wu, Y. and Gil1iland, E. (2000),“Mean Reversion across National Stock Markets and Parametric Contrarian Investment Strategies," Journal
01
Finαnce， 55(2), 745-772.
Brailsford, T. 1. and Maheswaran, K. (1998),“The Dynamics of the Australian Short-Term Interest Rate," Australian Journal
01
Manageme肘， 23(2),Chiao Da Management Review Vol. 33 No.1, 2013 103
213-234.
Brenner, R. J., Harjes, R. H. and Kroner, K. F. (1996),“Another Look at Models of the Short-Tenn Interest Rate," Journal of Financial and Quantitative
Analys缸， 31 (1)， 85-107.
Caporale, G. M., Cipolli凹， A. and Demetriades, P. O. (2005),“Monetary Policy and the Exchange Rate during the Asian Crisis: Identification through Heteroscedasticity," Journal of International Money and Finance, 24( 1), 39-53.
Casassus, 1. and Collin-Du企esne， P. (2005), “Stochastic Convenience Yield
Implied 企om Commodity Futures and Interest Rates," Journal 01 Finance, 60(5), 2283-2331.
Chan, K. C., Karolyi, G. A., Longsta缸~ F. A. and Sanders, A. B. (1992),“An Empirical Comparison of Altemative Models of the Short-Term Interest Rate," Journal 01 Fi胸前院的(3) ， 1209-1227.
Christiansen, C. (2010),“Mean Reversion in US and Intemational Short Rates,"
North American Journal 01 Economics and Finance, 21 (3), 286-296.
Christopher, S. J. (2003),“N onlinear Mean Reversion in the Short-Term Interest Rate," Review ofFinancial Studies, 16(3), 793 回843.
Cox, J., Ingersoll, 1. and Ross, S. (1980),“An Analysis of Variable Rate Loan
ContractsλJournal 01 Finance, 35(2), 389-403.
Dahlquist, M. (1996), “On Altemative Interest Rate Processes," Journal 01 Banking & Finance, 20(6), 1093-1119.
Engle, R. F. and Rangel, 1. G.. (2008), “The Spline-GARCH Model for Low-Frequency Volatility and Its Global Macroeconomic Causes," Review 01 Financial Studies, 21(3), 1187-1222.
Hays, P., Schreiber, M., Payne, 1. E., Ewing, B. T. and Piette, M. J. (2000), “Are Net Discount Ratios Stationary? Evidence of Mean-Reversion and Persistence," Journal 01 Risk and Insurance， 的(3)， 439-450.
Karol抖， G. A. (2002),“Did the Asian Financial Crisis Scare Foreign Investors out of Japan?" Pacific-Basin Finance Journal, 10(4),411-442.
Kim, J. H. and Ji, P. I. (2011),“Mean-Reversion in Intemational Real Interest Rates," Economic Modelling, 28(4), 1959-1966.
Data Frequencies and Great Events
Koutmos, G. (1998), “The Volatility of Interest Rates across Maturities and Frequencies," Journal of Fixed Income, 8(3), 27-31.
Koutmos, G. and Philippatos, G. C. (2007),“Asymmetric Mean Reversion in European Interest Rates: A Two-Factor Model," European Journal of Finance, 13(8), 741-750.
Larrain, M. and Pagano, M. (1993),“F orecasts from a N onlinear T-Bill Rate Model," Financial Analysts Journal, 49(6), 83-88.
Lo, A. W. and Mackinlay, A. C. (1988),“Stock Market Prices Do Not Follow Random Walks: Evidence from a Simple Specification Test," Review of Financial Studies, 1(1),41-66.
Pilbeam, K. (2001),
I恥S認su肘es，" Managerial Finance, 27(1&2), 111-133.
Poterba, J. M. and Summers, L. H. (1988),“Mean Reversion in Stock Prices:
Evidence and Implications," Journal ofFinancial Economics, 22(1), 27-59.
Said, S. E. and Dickey, D. A. (1984),“Testing for Unit Roots in Autoregressive .
Moving Average Models ofUnknown Order," Biometrika, 71(3), 599-607.
Samo, L. and Taylor, M. P. (1999),“Moral Hazard, Asset Price Bubbles, Capital Flows, and the East Asian Crisis: The First Tests," Journal of International Money and Finance, 18(4),637-657.
Sias, R. W., Starks, L. T. and Tinic, S. M. (2001),“Is Noise Trader Risk Priced?"
Journal ofFinancial Research, 24(3),311-329.
Smallwood, A. and N orrb血， S. C. (2008),“An Encompassing Test of Real Interest Rate Equalizatio丸" Review 01 lnternational Economics, 16( 1), 114-126.
Staikouras, S. K., 2006, “Testing the Stabilization Hypothesis in the UK Short-Term Interest Rates: Evidence from a GARCH-X Model," Quαrter砂
Review of Economics and Finance, 46(2), 169-189.