Automatic Asset Forecasts
Our automatic projection model aims to replicate this process, confirm it on a cross asset basis and draw price projections from it:
The model compares MJT oscillator series to an idealized model (similar to the canvases above). It scans this theoretical model to find the best fit to the current situation. It proposes two alternative (a best fit and a second best fit) per oscillator series (lower and upper rectangles) and then weighs these according to trend and risk/reward metrics into an average model projection. It then translates these oscillator projections into price projections using historical volatility as a norm (i.e. using our historical volatility measure “Delta”, 0.158 in the graph above – middle rectangle, right-hand side).
In order to achieve cross asset robustness and avoid esoteric interpretations, the model then performs the same analysis on the 5 best current Analogs in terms of price, our envelopes and our oscillators compared to the symbol which is being analyzed (chosen from 230 possibilities across all asset classes and relatives) as well as on the 5 best Analogs to the symbol in a specific set of 13 Key Markets drivers (e.g. S&P500, US10Y, GOLD, …). It then averages all these related projections to a consensus projection (e.g. 50% weight on the symbol which is being analyzed, 25% weight on the 5 Best Analogs, 25% weight on the 5 Key Market Drivers showing the best analogy to the initial symbol).
It performs the same analysis on both a Weekly and Daily basis and weighs these (50%/50% for now), so that as much as 88 single projections are then combined into a final one (i.e. 2 projections on our medium term oscillators, 2 projections on our long term ones, both on a Weekly and Daily basis, for the initial symbol, 5 Analogs and 5 Key Market Drivers). The result is then expressed using MJT’s envelopes:
a) Early identification and monitoring of the Trend direction
Using two standard deviation envelopes over two different calculation periods delivers very robust stress signals and helps reduce the fat-tail flaws often associated with normally distributed statistical methodologies.
b) Anticipation and confirmation of relevant inflexion points (Timing)
Timing Oscillators
To anticipate and confirm the timing of probable top and bottom zones, MJT uses timing oscillators. The long term oscillators are the drivers of the timing analysis. The medium term oscillators are calculated over a period half as long and use a slightly different formula. These serve as confirmation tools for the long term oscillators as well as monitoring instruments for shorter / intermediate trends.
Each oscillator (blue, red or black) is meant to capture a specific time cycle.
Their relative positioning delivers specific situations ( so called "Cases") to always know where you stand within either an uptrend or downtrend. The long term oscillators (lower rectangle) serve as the basis for the automatic analysis. The medium term ones (upperrectangle) are used as confirmations and to fine tune sub-segments of the model.
Please find below the ideal uptrend and downtrend model ( oscillators, derived model and case sequence) :
Ideal Uptrend Model : (left to right) from an oscillator black bottom (usually a Low Risk or a Case 2), the oscillators and prices will start moving up. An uptrend is confirmed once a red top can be made above a blue one. The correction down that follows delivers a buying opportunity (“Resume Uptrend”) followed by an intermediate top (Case 3). A new period of consolidation down or sideways then starts, ending with a Case 5 acceleration up towards an important top (usually a High Risk or a Case 1).
Ideal Downtrend Model :(left to right) from an oscillator black top (usually a High Risk or Case 1) the oscillators and prices will start moving down. A downtrend is confirmed once a red bottom can be made below a blue one. The correction up that follows delivers a selling opportunity (“Resume Downtrend”) followed by an intermediate bottom (Case 4). A new period of consolidation up or sideways then starts, ending with a Case 6 acceleration down towards an important bottom (usually a Low Risk or a Case 2).
Overall, the MJT system, which was first developed in the 1970s, brouhght online in 1982 and can be applied to any asset class, any time frame, to relative charts or to portfolios with many positions.
c) Automatic assessment of possible price Targets and hence of risk/reward ( Target )
The larger envelope is also used to monitor volatility i.e. the width of the larger envelope, labelled “delta” is MJT’s measure of volatility.
Delta can be used over all time horizons to calculate price targets and hence capture the risk/reward profile of any investment decision at any given point. The width of the larger envelope changes with every frequency and so does delta – the larger the frequency, the greater the width of the larger envelope and the larger the delta. Targets, which are confirmed over two to three investment horizons, usually offer extremely powerful support and resistance levels.
Corrective Target Range
The Corrective target range is calculated using "delta" (Δ: larger envelope width and volatility measure) multiplied by a corrective factor of 0.5x to 0.8x and subtracted/added from the relevant top or bottom (usually but not always MIN and MAX of the graph).
The Corrective Objective is considered "Done" once the price leaves the Corrective range. Prices will probably resume the previous primary trend. A move beyond this corrective range could confirm a change in this primary trend.
Impulsive Target Range
The Impulsive target range is calculated using "delta" (Δ: larger envelope width and volatility measure) multiplied by an impulsive factor of 1.3x to 1.7x and subtracted/added from the relevant top or bottom (usually but not always MIN and MAX of the graph).
The Corrective Objective is considered "Done" once the price leaves the Corrective range. Prices will probably resume the previous primary trend. A move beyond this corrective range could confirm a change in this primary trend.
The Impulsive Objective is considered "Done" once the price overruns the target range, i.e. the price move carries on beyond 1.7x "delta".
In rare occasions, when the primary trend is very strong, a Super Impulsive target range (I2) is calculated using a super impulsive factor of 2.3x to 2.7x. This Super Impulsive target range is envisaged starting at above 2.0x delta and is considered done above 2.7x delta.
Example: Weekly chart of S&P 500 in early November 2016
