Broken supports and shorts: from a price level to a testable rule

From a local low to a confirmed support

The general definition and uses of support are covered in the existing support-and-resistance guide linked below. In this particular study, the detector identifies a local low relative to surrounding candles, waits for subsequent candles to close and groups nearby lows into an average level.

The supplied confirmation delay is two four-hour candles, or eight hours. The time when a low forms is therefore not the time when the support first becomes available to the strategy. The exact grouping tolerance and complete neighborhood settings are not supplied; they must not be invented.

What is a broken support, and how is it detected?

For the supplied backtest, a support breaks on the first four-hour candle that closes below an already confirmed level. A candle trading below the level but closing above it does not satisfy this definition. Detecting a break requires comparing a completed candle’s close with the support that was available at that time.

A broken support can provide a consistently defined event for historical testing. It does not prove that price must keep falling. In the model, a later close back above the broken support is precisely the stop condition for a short still open.

What is a short, and what determines its result?

A short is a position whose price result benefits from an exit below the entry and loses when the exit is above the entry. The report expresses that result as (entry − exit) / entry, stated as a percentage. This describes the price result in the simulation before its excluded costs; it is not a return on a specified margin account.

Understanding a short helps distinguish a downward-price hypothesis from a profitable outcome after entry and exit rules. The supplied study uses Binance perpetual-futures prices and simulated positions. It does not demonstrate real trades or specify the margin, leverage or account configuration needed for live execution.

How the supplied model opens a short

First, a support must be confirmed. Next, a completed candle must close below it. The model then locates the highest still-active support below the prospective entry and sets its target 0.25% above that lower support. It enters at the breaking candle’s close only when the target is at least 0.25% below that entry.

If there is no lower support, or the target is too close, the model skips the break. It opens only one short per support, even if the level is later recovered and broken again. Several different supports broken in the same candle can nevertheless create overlapping positions. These are backtest rules, not instructions to place an exchange order.

Target, closing-price stop and execution order

The target remains fixed at entry. Earlier supports from the preceding 90 days can supply a target but do not open their own shorts in this study. A later change to a grouped support’s average does not move an existing target.

The model counts a target fill when a candle’s low reaches its price, checking that condition before the stop. If the target has not been reached, the first candle closing back above the broken support exits the position at that close. The stop therefore does not promise an exit at the support itself or cap the intrabar move at a specified amount.

How this information can be used responsibly

The sequence separates detection, entry eligibility and exit. It lets a reader see which breaks were traded, which were excluded and how the assumed exits affected results. Maximum favorable movement is different from the result at exit, and adding overlapping trade percentages is different from measuring a funded portfolio.

The linked example contains 128 simulated trades in one 365-day window. Its assumptions omit commissions, funding and slippage, and touching a limit price may not produce a real fill. Changing the parameters after observing the same sample can fit the strategy to that sample. Historical similarity does not guarantee the same future outcome.

Application and source

Bitcoin support breaks: 128 simulated shorts sum to -16.13%.

Definitions from the RavenInvestor report dated October 27, 2026 at 18:43 UTC and its original chart. Declared data source: Binance BTCUSDT aggTrades.

Educational financial information, not personalized investment advice or a recommendation to open a short. Simulated historical results do not guarantee future results.