How to Read Cycles
When reading a CycleProphet chart, there are three primary elements to consider (Turns, Direction and Amplitude):
Turn
A turn is defined as a change in direction of trend. The CycleProphet program maps the historical data cycles to unchanging cycles of time. Then, it removes all time cycles that have more than about 10% randomness. This means the resulting time cycles are ‘mapped’ 90% or better to the historical data. Then, these time cycles are combined into a composite time cycle that projects a map of the future data. Key elements have to do with tops, bottoms and strength of move from one point of time to the next.
Turns do not have to be a complete trend reversal. A Turn can also be nothing more than a change in slope of a trend; an inflection point.
Direction
At each Turn, CycleProphet predicts whether the data will trend higher or lower, following the Turn event. The vast majority of the time, these turns are directly correlated to the actual event (historical and forecast), but a small percentage of the time, the Turns are inversely correlated, meaning that the direction of the trend coming out of the Turn is the exact opposite of reality.
CycleProphet continually looks for these inversions and, if one is encountered, it adjusts the forecast automatically. The charts are updated nightly. Each new day’s closing data provides another data point for CycleProphet to analyze. As inversions are encountered, the future forecast (blue line in the chart, above) is adjusted accordingly.
Amplitude
Amplitude is defined as the difference in value of the cycle-data from one Turn to the next Turn. Amplitude provides insight as to the significance of a Turn. For example, it would be very valuable to know that an upcoming Turn is going to precipitate a 15% move in the market, as opposed to a 1% move in the market.
NEVER LOSE SIGHT OF THE FACT THAT CYCLEPROPHET CHARTS ARE FORECAST MODELS...
NOT FORECAST FACTS!
"Forecasts" are NOT "Facts". You should never consider the CycleProphet forecasts to be reliable enough to use exclusively for trading. You should use the CycleProphet charts for what they are intended... an additional piece of information to consider before making a trading decision.
Background
When it comes to using time cycles to predict future market movements, the following assumptions are used:
- Markets move in cycles, where there is normal ebb and flow of markets moving higher and lower over time.
- It is possible to find a unique set of time cycles that will closely correlate (map) to the actual historical cycles of closing data.
- The higher degree of correlation of the time cycles to actual historical data, the higher the degree of accuracy in predicting future trends of the data.
Mapping of Time-Cycles to Historical Data is NOT "Curve-Fitting"
It is very important to understand that the process of finding the 'best map' of real data to time-cycles is NOT a "curve fit" process. Complex curve fitting algorithms (polynomial curve fitting, least squares curve fitting, regression analysis and nonlinear programming, for example) can produce a better 'curve fit' of the historical data than the typical results of our proprietary time-cycle mapping algorithms. The problem isn't a matter of how good one curve can fit to another... rather, the problem is how to best forecast a future movement in the market. Traditional curve fitting routines, regardless of the sophistication of the algorithms used, tend to be poor choices for extrapolating into a forecast of future events.
CycleProphet uses our proprietary 'mapping' technique, where time-cycles of fixed amplitude and wave length are 'mapped' to historical data. The objective is NOT to find the best curve fit to the data. The objective is to find the best combinations of "time cycles" that, when combined, have the highest coincidence of Turns, Directions and Amplitude.
Working Through an Example
Take a look at the chart above of the S&P 500. It shows 12 months of historical data and 3 months of forecast data. CycleProphet charts are designed to run infinitely into the future, but accuracy degrades the further into the future the charts are run. The CycleProphet charts (market indicators) all show 12 months of history and 3 months of forecast.
The black line represents the actual historical daily closing prices. The blue line is CycleProphet’s forecast of the future daily closing prices. The red line is composite time-cycle data, both historically and into the future.
PRIMARY ASSUMPTION:
IF THE HISTORICAL DATA REASONABLY MAP TO SPECIFIC TIME-CYCLES, THEN ONE CAN ASSUME THAT THE FUTURE DATA WILL REASONABLY MAP TO THESE SAME TIME-CYCLES.
In the chart, above, the 12 months of historical data and 12 months of Time-Cycle matching, are provided for two reasons:
- To give the investor a view of how the market performed during this same time, a year ago. Many investors believe markets have an annual cyclicality to them.
- To give the investor a visual sense of how well the Time-Cycles mapped with the major Turns of the historical data. The assumption being, that the better the map historically, the better the forecast.
At the bottom chart is a table of Turns and Directions as forecasted into the future. This helps the reader by providing specific dates and directions, rather than just visually looking at the chart.
|
