Daily futures price data over the 1985–2004 period are used to evaluate performance of technical trading rules. The data set consists of open, high, low, and closing prices for the 12 futures markets analyzed by Lukac et al. (1988) and 5 additional markets. These actively traded markets are selected from each major category of futures contracts, i.e., grains (corn, soybeans, and wheat), meats (live cattle and pork bellies), softs (cocoa, sugar, and lumber), metals (silver and copper), an energy (crude oil), currencies (the pound, the mark, and the yen), interest rates (treasury-bills and the Eurodollar), and an equity index (S&P 500). The price data for the 17 futures markets is provided by the Commodity Research Bureau, Inc.
Because Lukac et al. (1988) investigated the 1975–1984 period, this study considers the 1985–2004 period as the full sample period, with the exception of two financials: the mark (1985–1998) and treasury-bills (1985–1996).
The full sample period is divided into two 10-year subperiods, 1985–1994 and 1995–2004. Out-of-sample performance of the best trading rule identified in the 1985–1994 period, therefore, is evaluated as well as in-sample performance in each subperiod and the full sample period.
It is important to incorporate accurate daily price limits into the trading model because for certain futures contracts, price movements are occasionally locked at the daily allowable limits. Trend-following trading rules typically generate buy (sell) signals in up (down) trends. Thus, if a trading signal is triggered on the day a price limit move occurs, then a buy (sell) trade is typically executed at a higher (lower) price than that at which the trading signal was triggered. This may result in seriously overstated trading returns if the trade is assumed to be executed at the limit ‘locked’ price levels. For contracts having daily trading limits, therefore, no position is taken or closed out if the high price equals the low price or both of the prices equal the closing price (i.e., lock-limit days), or if the contract’s opening price is above or below the daily allowable limit. The history of daily price limits for each contract is obtained from exchange statistical yearbooks and the annual Reference Guide to Futures/Options Markets and Source Book issues of Futures magazine.
The trading model is similar to that used in Lukac et al. (1988). The trading model typically consists of input data, technical trading systems, performance measures, optimization methods, and other relevant assumptions. Daily futures prices are used as input data. To obtain a price series that reflects the most important market characteristics, this study uses ‘dominant contracts’that have the highest open interest (Dale & Workman, 1981). Dominant contracts are identical to nearby contracts in all but 4 of the 17 markets, i.e., corn, soybeans, copper, and silver.[2] The current dominant contract is assumed to roll over to the new dominant contract on the second Tuesday of the month preceding its delivery month.
Performance criteria adopted are the mean net return and the Sharpe ratio. Although defining a rate of return in the futures market may be problematical because there is no initial investment except for a margin deposit, recent technical analysis studies about futures markets tend to use the continuously compounded (log) return per unit (Kho, 1996; Sullivan et al., 1999; Szakmary & Mathur, 1997). The continuously compounded daily gross return on a technical trading rule j at time t can be calculated by:
rjg,t1 [ln(Pt1) ln(Pt)]Sj,t (1)
Уважаемый посетитель!
Чтобы распечатать файл, скачайте его (в формате Word).
Ссылка на скачивание - внизу страницы.