10-year historical analogue projection
Loading...
This model is a historical analogue model. It does not predict the market from fundamentals, news, interest rates, or other external variables. Instead, it assumes that a selected instrument's future price path may partially resemble its price path from approximately 10 years earlier.
For a forecast year, the model selects the same instrument's data from exactly 10 years earlier.
For the current year, the model compares the number of trading days already completed with the corresponding trading-day position in the historical analogue.
For example, if the selected instrument has completed approximately 170 trading days in the current year, the model uses the 170th trading day of the historical analogue as the reference point.
The model calculates how much the historical analogue moved from the equivalent trading-day position until the end of that historical year.
This factor represents the remaining movement observed during the historical analogue period.
The remaining historical movement is then applied to the latest available value of the selected instrument.
Therefore, the model is effectively asking: "What would happen if the remaining part of the current year followed the same proportional movement as the corresponding part of the historical cycle?"
For a year that has not yet occurred, there is no current-year actual data. Therefore, the model starts from the previous year's projected year-end value.
The same process is then repeated for subsequent future years.
Cycle Similarity measures how closely the current year-to-date price path resembles the corresponding historical path.
Both paths are normalized to their starting value and compared trading day by trading day. The model calculates the difference between the two normalized paths and converts the resulting error into a 0–100 similarity measure.
Similarity is a measure of path resemblance, not a probability that the forecast will be correct.
The backtest applies the same 10-year analogue idea to completed historical years where both the actual year and its 10-year historical analogue are available.
The backtest reports how often the historical cycle direction matched the actual direction.
The current incomplete calendar year is excluded from the full-year backtest.
Historical-cycle analysis should be considered a scenario model, not a statistically guaranteed prediction.
A historical pattern can repeat partially, completely, or not at all. Market conditions can change because of company fundamentals, economic conditions, monetary policy, political events, regulation, liquidity, investor behaviour, corporate actions, and unexpected events.
Therefore, a high Cycle Similarity does not mean that the forecast has the same percentage probability of occurring.
Likewise, the backtest measures only the specific historical rule used by this model. A directional match does not establish that the model can predict the magnitude, timing, volatility, or future market regime.
The backtest provides an empirical check of whether the historical 10-year cycle direction has matched the selected instrument's subsequent full-year direction in the available historical sample.
A larger historical sample generally provides more information than a very small number of observations, but past results still do not guarantee future performance.
Current price path → find its 10-year historical analogue → measure the historical remaining movement → apply that proportional movement to the current value → produce a scenario-based year-end projection .
Directional comparison between each year and its 10-year analogue for the selected instrument.
| Year | Analogue | Historical Return | Actual Return | Cycle | Actual | Match |
|---|