Fib confluence: the one chart idea we use, and the ones we ignore

Runir publishes a small set of fibonacci options levels — specifically, the points where a dealer- positioning wall coincides with a fibonacci retracement of the recent swing. The coincidence is what we publish; neither piece is a signal on its own. The discipline that filters every other charting tool off the page is the same discipline that determines which fib levels are worth tagging.

Why most chart indicators fail the bar

The framework Runir publishes already has a credibility anchor: every wall is computed from current dealer positioning, scored against the next session's close, with a published hit rate. The bar for adding a SECOND indicator is high — anything we add either earns its place by independently corroborating the wall, or it muddies the framework.

Most popular chart tools fail this bar:

• RSI, MACD, oscillators. All derived from price history alone. They tell you what price did, not what's structurally on the other side of where it might go next. Nothing independent corroborates them. Pattern recognition without an anchor.
• Moving-average crossovers. Lag every move they describe. By the time a 50-day crosses a 200-day, the move that mattered already happened.
• Bollinger bands. A volatility envelope around recent price. Useful for mean-reversion intuition, but it's still just a recasting of price history. The bands don't know anything the chart doesn't already show.
• Hand-drawn support / resistance lines. Subjective. Two analysts looking at the same chart draw different lines. A level whose existence depends on who's looking isn't a level we can score.

None of these are wrong as standalone tools for a chart-only trader. But for a framework whose claim is “we publish levels that mean what they mean tomorrow because the math is mechanical and the rate is published,” they don't add anything verifiable. They'd ride the brand's accountability discipline without contributing to it.

Why fib retracements meet the bar

Fibonacci retracement levels — the 23.6%, 38.2%, 50%, 61.8% retracements of a swing — have one property the indicators above don't: they're observed and traded by a large enough cohort of market participants that they create REAL positioning at those levels. The level exists not because it's mathematically magic but because enough traders watch it that the level becomes self-reinforcing.

That makes fib retracements one of the few chart-derived ideas that survives the “is anyone actually doing something there” test. The fib level isn't the signal; the BEHAVIOR of traders who watch the fib level is. That second-order framing is what separates fib levels from RSI — nobody positions around RSI 70; large numbers of traders position around the 61.8% retracement.

The interesting move is when a Runir wall lands AT one of those fib levels. Two independent frameworks — dealer positioning AND chart structure — are pointing at the same price. Each one is a probability tilt on its own; together, they describe a more credible level than either alone could.

What “confluence” specifically means here

A wall gets a confluence tag on the per-ticker page when the wall's strike lands within 0.5% of any of the four standard fibonacci retracements (23.6%, 38.2%, 50%, 61.8%) of the trailing ~60-session swing.

• 0.5% tolerance. Tight enough that “confluence” means actual coincidence, not “in the neighborhood.” A wall at $202 with a fib at $215 is not confluent.
• 60-session swing. Long enough to capture meaningful chart structure, short enough to stay relevant to current positioning.
• Standard four levels only. No esoteric fib variants (78.6%, 88.6%, fan lines, extensions). Three reasons: those levels are less observed by traders, so the "real positioning" property weakens; adding them dilutes the confluence signal; and at some point you can find a fib level near any wall just by adding enough lines.

The tag is binary — confluence present or not. We don't try to grade “stronger confluence vs weaker confluence.” The strike either coincides within 0.5% or it doesn't. Anything more granular adds judgment the framework doesn't earn.

What it doesn't tell you

Confluence is a CREDIBILITY MULTIPLIER on a wall, not a separate level. A fib level alone is not a signal Runir publishes. We're not saying “this fib retracement is support” — we're saying “this wall happens to coincide with a fib retracement, so it's more likely to be defended than a wall at a strike no chart-watcher cares about.” The wall does the load-bearing work; the confluence tag is the small “yes, that one” addition.

Confluence absence is also not a negative signal. A wall without a confluence tag is still a wall. Most published walls don't have confluence; that's the expected baseline. The tag fires when it's worth noticing, not as a vote against levels that don't trigger.

The discipline of not publishing more

There's a temptation, on a chart-derived enrichment page, to ship more indicators. RSI is easy to compute. MACD is easy. Adding three more tools would make the page look thorough.

We don't, because more indicators that don't earn their place make the framework noisier without making the levels better. Every indicator we add is one more thing that has to demonstrate it improves the published rate; if it doesn't, it's just visual noise that creates the impression of analysis without doing any analysis. That's exactly the failure mode of dashboard-stacked services that pile on indicators because indicators sell.

The discipline of NOT publishing things is one of the brand's editorial choices. The per-ticker page renders the wall, the regime, the persistence count, the confluence tag if it fires, and the heat reading. That's it. Five reads, each with a defined role, each earned the right to be there by demonstrating it improves the read.

If a chart-derived idea ever earns its way onto the page — something that demonstrably tightens the published rate when included — we'll add it. Until then, fib confluence is the only chart-derived enrichment that's earned it.

What to do with this

Open today's NVDA page and look for the confluence tag on the wall rows. When present, it appears as a small chip on the row indicating the fib level the wall coincides with. When absent — the more common case — the wall ships without it.

If you came in cold, the call wall piece covers the upper-level mechanic and the put wall piece covers the floor. Fib confluence is an enrichment on those levels, not a separate concept — it makes a credible wall more credible, not a non-wall a wall.

Common questions

Does Runir use Fibonacci levels?

Only as a wall enrichment. When a published wall lands within 0.5% of a standard fib retracement of the trailing ~60-session swing, it is tagged as confluent — two independent frameworks pointing at the same price. A fib level alone is never published as a signal.

Why doesn't Runir use RSI, MACD, or moving averages?

They are derived from price history alone, with nothing independent corroborating them. Fib retracements earn their place because enough traders watch them to create real positioning at those levels — the behavior, not the line, is the signal.