This White Paper presents a trading simulation where use of predictions of high and low prices significantly outperforms a passive “buy and hold” strategy. The simulation or backtest focuses on Microsoft’s widely-traded stock (MSFT) January 2006-July 2020, a period when shares of Microsoft rose from $26.84 (1/3/2006) to $201.30 (7/24/2020), or by a factor of 7.5. By comparison, long trades based on forecasts of two-trading day high and low prices produce cumulative gains of 30.73 over this period.
Predictions of high and low stock prices support the classic trading principle – buy low and sell high. The feasibility of such predictions is perhaps not generally recognized, but is compatible with finance theory – including “efficient markets” and “rational expectations.” Through its R&D program, Information and Price Dynamics has streamlined algorithms for real-time calculation of such high/low (H/L) forecasts.
Key metrics of the MSFT trading simulation are as follows:
The trading strategy involves transacting 418 trades over 3638 trading days. The maximum holding period is two trading days with many trades conducted within a single day. Trades are always initiated in a day in which no other trades are initiated or concluded, so holding periods do not overlap.
The cumulative gain is a factor which can be multiplied into an initial capital devoted to trading this stock to determine the final capital position. Thus, a $100,000 investment in MSFT at the beginning of 2006 grows to $3,027,000 by the end of the study period in late July 2020.
A high proportion of these trades are profitable (94%) with the average gain from a trade being 0.8 percent. The maximum gain achieved in the study period is also more than double, in absolute terms, the maximum loss from a trade.
The proportion of profitable trades and asymmetry between maximum gain and minimum loss suggests buying on margin. Assuming 10 percent annual loan costs from buying on 30 percent margins, a $10,000 investment in MSFT at the beginning of the study period in 2006 could be parlayed into more than $6,000,000 by late July 2020.
A trading strategy is a set of rules for entering and exiting trades. For this simulation, positions are established when: (1) the predicted two-day low price for MSFT falls within the price range on the first trading day of the two-trading-day forecast period, and (2) the predicted two-day high does not fall within the first day price range prior to the predicted low being within in the daily price range.
A long position is exited when the predicted two-day high price falls within the daily trading range of either the first or second trading day in the forecast horizon, providing the entry conditions are met. The default is close of the position at the closing price of the second trading day of the relevant two-trading day period.
The focus of the trading strategy on buying the low on the first day of the two-day forecast horizon is motivated by a simple concept. When stock prices trend up over a period, the period low tends to be in the initial part of that period.
For Further Information