ADVANCED · ENGLISH · FREE

5. Trading models

AB=CD

The geometric model compares movement A-B, correction B-C and possible continuation C-D. In the simplest bullish example, AB = B - A and the projected equal-move point D = C + AB. Signs change for the downward version. Real implementations may use tolerances, filters, confirmations and other levels.

Points must be defined without future data. If B is only confirmed two candles later, the algorithm could not know its final position at the peak itself. Limit correction depth and waiting time so the rule does not become an endless search for a convenient D.

Practice geometry. A = 90, B = 100, C = 95; therefore AB = 10 and D = 105. This is a geometric projection level. It is not a probability of reaching it, an automatically valid target or an exact description of LOGRI's current algorithm.

Breakout and Retest

For a breakout, define the original range, crossing criterion and earliest available entry time. A volume filter, if used, must compare against a baseline known then. A breakout on the close followed by execution at the next candle's open requires accounting for a possible price gap.

For a retest, define the permissible return depth, maximum delay and condition for holding. A repeat touch many hours later may take place in a different context. A "touch count" is ambiguous without a definition.

Invalidation must occur before a scenario can be reinterpreted to fit any outcome. Examples include no retest within K candles; a close back inside the range; the target area passing before entry; or price exceeding the permitted deviation. These are separate outcomes useful for research.

Impulse → Accumulation → POC Retest

This research sequence assumes a directional move, followed by bounded volume accumulation and a return to the resulting area's profile. Each stage needs measurable conditions and its own confirmation time. You cannot build POC from a future accumulation phase and then draw entry at its beginning.

The current LOGRI product has its own POC research implementation. Its specific filters, statuses, level calculations and data are the source of truth for the product. The course's general outline explains the research logic but does not replace the implementation or promise profitability.

Select the stop from hypothesis invalidation, then size from the budget. If the minimum contract does not fit the risk after calculation, rejecting the trade is preferable to moving the stop somewhere meaningless. Similar signals in correlated assets are not necessarily independent opportunities.

Exercise. Create a model card with seven fields: input data; context; signal; first available price; invalidation; exit; costs. Give it to someone else. If their markings differ, clarify the rule before testing, not after seeing the results.