High/Low Forecasts and a New Technical Indicator

To what extent can technical indicators and other predictive tools provide forward guidance for future stock price movements?

On the one hand, finance orthodoxy suggests that stock prices are essentially a random walk. This means, among other things, that the best forecast of tomorrow’s price is the current price of the stock. Markets are efficient. Public information is quickly assimilated and reflected in stock prices. Instead of timing stock purchases, a portfolio approach assures more reliable and higher profits by selecting securities which show negative correlation in their returns, e.g. bonds and stocks.

But one type of price forecasting is widely accepted – this involves anticipation of price volatility.

Volatility involves more rapid, more energetic fluctuations in stock prices. Volatility events tend to cluster, so periods of higher or lower volatility are somewhat predictable.

Research in Information and Price Dynamics has uncovered linkages between volatility and predictions of high and low prices. In this short discussion, we consider how the predicted price range turns out to be a proxy for swings in high or low prices compared to the current closing price. This fact – correlations between the predicted price range and metrics we name UP and DOWN – leads to successful trading over multi-day time frames.

Consider a technical indicator or trading signal called RANGE5* which is the predicted 5-trading-day range, calculated as the difference between the predicted 5-day high and predicted 5-day low prices. RANGE5* has strong correlations with UP (0.52) and with DOWN (-0.474). UP is the ratio of the high price for the next 5 trading days and today’s closing price. DOWN is the ratio of the low price for the next 5 trading days and today’s closing price.

RANGE5* is a leading indicator for volatility, but the immediate point is that it performs well as a trading signal.

To show this, consider daily prices of the SPDR exchange traded fund tracking the S&P 500 called SPY. We develop a trading simulation over a 14-year period from 2007 through late 2019, which includes 3194 trading days. The simulation involves entering a long trade when RANGE5* is above a threshold value, holding that trade while RANGE5* remains higher than the threshold value, and exiting the trade when RANGE5* drops below the threshold. Trades are executed at the closing price on the days when these events occur. Short trades are initiated at the exit of long trades, and long trades are entered at the exit of short trades, so an investor is fully invested with Trading Strategy RANGE5*.

Combining gains from long and short trades, overall gains from reinvesting funds from each trade into the next trade cumulate to a factor of 4.90. When 30 percent capital gains are deducted from profitable trades at the end of each tax year, the overall gain drops to 3.35, still significantly greater than the raw gain from a buy and hold strategy of 2.2.

The following TABLE lists annual rates of return from this Trading Strategy RANGE5*, after deducting 30 percent capital gains taxes from profits each tax year. Capital losses from 2008 and 2017 are carried forward, reducing capital gains taxes in subsequent years. The TABLE also shows annual rates of return for a buy and hold strategy.

Trading strategy RANGE5* can be expressed as step-by-step procedure, as follows.

  1. Develop predictions of 5 trading day high and low prices
  2. Calculate RANGE5*
  3. If RANGE5* > threshold X then buy SPY at the closing price of the trading day
  4. If a long position is held on at the beginning of a trading day, continue to hold the position, as long as RANGE5* > threshold x
  5. If a long position is held on the previous trading day, exit the position at the daily closing price if RANGE5* is less than or equal to threshold x
  6. Institute complementary short positions when a long position is not held at the same price points

Altogether, Trading Strategy RANGE5* results in 198 long trades, and 196 short trades. 74.7 percent of the long trades are profitable, while 51.5 percent of short trades are profitable.

RANGE5* – which is built from predictions of five-trading-day high and low prices – is, accordingly, like other technical indicators. Its fluctuations in value signal profitable trades.

The CHART below shows how RANGE5* tracks the metric UP which measures how much future high prices rise over the current closing price. UP clearly represents some type of potential for long trades.

One caveat is the predictions of high and low prices must be relatively accurate – a focus of research for Information and Price Dynamics.

CONTACT research@infopricedynamicsinc.com  for additional information and access to forecasts