Fibonacci Bollinger Bands: Identify the Formula Before Testing
Fibonacci Bollinger Bands is a name used for different indicator implementations. Before interpreting a 0.618 or 1.618 band, identify the script's author, center line, volatility measure and multiplier. A shared name does not make two formulas interchangeable.
A documented ATR-based implementation
Shizaru's Bollinger Bands Fibonacci ratios describes a moving-average center with Wilder-smoothed ATR and three multipliers: 1.618, 2.618 and 4.236. Its published description specifies three upper and three lower bands. It is listed as an open-source script.
For center M, smoothed range A and multiplier r, the described construction is:
- Upper band = M + r × A.
- Lower band = M - r × A.
For an illustrative M of 100 and A of 2, the 1.618 pair is 103.236 and 96.764. These are arithmetic values, not observed support and resistance or a trading result.
A standard-deviation version instead uses a deviation measure in the width term. Some variants divide a base width into fractions such as 0.236, 0.382, 0.5, 0.618, 0.764 and 1.0. In that design, the outer 1.0 level means the full configured width; it is not equivalent to 1 ATR or 1 standard deviation unless the formula makes it so.
Do not merge settings from different scripts
| Item to inspect | Why it changes the result |
|---|---|
| Center line | SMA, EMA and VWMA weight the observations differently |
| Source | Close, typical price and other sources produce different inputs |
| Volatility measure | ATR measures ranges; standard deviation measures dispersion around a mean |
| Base multiplier | A fractional ratio may multiply an already scaled width |
| Lookback and smoothing | These determine response and initialization |
| Offset and requested timeframe | A drawing shift is not early knowledge; an open requested bar can change |
The common 20-period, two-deviation Bollinger setup is a convention, not an immutable formula. A 200-period VWMA is not automatically a better choice for crypto. Volume weighting also depends on the feed's volume meaning and availability.
What band behavior actually tells you
A price above an upper band is far above the chosen center relative to the chosen width measure. It does not establish that the next trade should be short or long. A mean-reversion strategy fades the move; a continuation strategy follows it. Both need separate rules.
Narrower bands show a smaller calculated width. They do not determine the direction or timing of a future breakout. A Fibonacci label does not supply evidence that market participants will defend the level.
Adding RSI or a volume filter changes the strategy and the sample of trades. It does not independently confirm profitability. Ordinary RSI and Mansfield relative strength measure different things and should not be treated as interchangeable confirmations.
Set up a reproducible TradingView comparison
- Open the exact publication, record its author and version, then add it to the chart.
- Record the symbol/feed, timeframe, source, length, smoothing and every multiplier.
- Inspect several band values against the formula. Include the start of available history and an open bar.
- Keep plotting offsets explicit. Test any higher-timeframe behavior around its bar close.
- Save the configuration before comparing an alternative implementation.
The linked Shizaru script is an indicator. A separately published Bollinger Bands Fibonacci Ratios Strategy is by The-4xdev-company; DdgNyfMY is its publication ID, not the author's name. The author describes selectable bands and reverse-buy/sell options. Do not assume it implements an RSI filter merely because another article describes one.
Define the trading experiment
Choose either a breakout or a reversion hypothesis before looking at results. For example, a breakout experiment might require a completed close above one specified band and place an entry at the next permitted execution. State the exit, size, stop behavior and maximum simultaneous positions separately.
If adding RSI, state the period, threshold and evaluation bar. Compare the filtered and unfiltered rules on the same data, with costs and a reserved evaluation interval. Save the code version and trade list; report sample size, payoff distribution and drawdown as well as win rate.
No BTC/USDT backtest, 58% win rate or 1.8-to-1 realized payoff is established here. Historical results from an author or a different script do not validate a new combination. TradingView's strategy documentation explains fill assumptions that must be considered in the test.
Pineify's coding workflow can help express a specified formula and strategy. Review the generated source and test its behavior; code generation does not establish that the selected band predicts a reversal.