Success of any trade depends on the ability to spot and profit from market swings associated with prices {x_t: 1, 2, …, N} of a stock.  In this paper an optimal technical trading rule (OTTR) is proposed to identify profitable positions for ‘when to buy and when to sell’ to help all traders who live and die with minute-by-minute price data. Furthermore a trading rule G_SL(t) that assigns selling positions with  an upper level price and buying positions with a  lower level price is formulated  by monitoring the ratio series R(t)=MA_S(t)/MA_L(t) where, S < L with MA_S(t) and MA_L(t) as simple moving averages (MAs) computed from the stock series {x_t} under study.  We denote  the mean and standard deviation measures of the R_SL(t) series by ‘m’ and ‘s’ respectively and the upper level positions (ULPs) are selected above the mean at time ‘t’ if (R_SL(t) > m+ks,  R_SL(t-1) <m+ks ) and lower level positions (LLPs) below the mean are chosen at time ‘t’ if (R_SL(t-1) > m-ks,  R_SL(t) <m-ks ), defining a trading rule G_SL(t). A combination (S^*, L^*, h^*) that maximises the total expected profit P_SL(t, h) over the positions determined by the OTTR is selected as the ‘Optimal technical trading rule (OTTR(S^*,L^*, h^*))’  for this investigation. To implement the proposed methodology pertaining to this rule, a training data set and testing data set are simulated and an appropriate model is fitted by hybrid-Auto Regressive Integrated Moving Average (hybrid-ARIMA) and Artificial Neural Network (ANN) methods. Using the estimated values of the parameters by hybrid-ARIMA and NN methods, predictions are made for testing data set. From these predicted values, OTTR(S^*, L^*,h^*) for both hybrid-ARIMA and ANN approaches are  obtained and the corresponding maximum profits are compared.

KEYWORDS: ARIMA model; ANN model; MA values; predicted prices; OTTR R(t) ratio; Positional profit.