> For the complete documentation index, see [llms.txt](https://docs.thekingfisher.io/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.thekingfisher.io/trading-workflows/alpha-research-with-kingfisher-widgets.md).

# Alpha research with Kingfisher widgets

Use Kingfisher and AGGR widgets to form, compare, and test crypto market hypotheses without treating modeled data as a trade signal.

If you are researching crypto trading alpha, use the widgets to make a market hypothesis specific, compare it with independent evidence, and record what would prove it wrong. Kingfisher surfaces market data and modeled views; no widget by itself measures, verifies, or guarantees an edge.

In this guide, **alpha research** means investigating a repeatable market advantage. A credible result still needs a defined benchmark, risk controls, trading costs, and evaluation on observations that were not used to form the hypothesis. The workflows below help you ask better questions; they are not entry, exit, or position-sizing instructions.

## A repeatable research loop

1. **Write one falsifiable question.** Example: “When modeled liquidation exposure is nearby, does executed flow and price response differ from otherwise similar periods?” Avoid starting with “the map predicts a squeeze.”
2. **Fix the comparison scope.** Record symbol, venue or exchange set, timeframe, refresh time, and each widget’s mode. Do not compare unlike settings as if they were the same sample.
3. **Read the primary view, then challenge it.** Use another data source or widget that answers a different question. A second visualization of the same inputs is not independent confirmation.
4. **Define observable confirmation and invalidation.** Use displayed price, trades, flow, or a change in the model. State what would weaken the idea before reviewing the outcome.
5. **Keep a timestamped log.** Record the view, settings, hypothesis, evidence for and against it, and the eventual outcome. Do not select examples only after the move.
6. **Evaluate separately from discovery.** Compare many observations, include fees, slippage, funding and execution assumptions, and reserve a later sample for validation. Dashboard screenshots alone do not establish profitability.

### A simple observation worksheet

| Field            | Record                                                                |
| ---------------- | --------------------------------------------------------------------- |
| Question         | One claim that could be contradicted by later data                    |
| Scope            | Market, venue set, timeframe, timestamp, modes and model inputs       |
| Evidence sought  | The specific displayed behavior that would support the question       |
| Counter-evidence | What would weaken or invalidate the interpretation                    |
| Follow-up        | Which independent widget or source checks the same event differently  |
| Outcome          | What happened after the observation, including cases that did not fit |

## Workflow 1: investigate liquidation exposure around price

**Question:** Does price behavior near a modeled liquidation zone coincide with observable trades or flow changes?

1. In [LiqMap](https://docs.thekingfisher.io/products/liquidations/liquidations-maps.md), select a symbol, exchange mode, and leverage filter. Compare a single exchange with MULTI, or switch to MULTI\_MARKETS for several symbols. Note the setting and time.
2. Use [LiqRatio](https://docs.thekingfisher.io/products/liquidations/liqratio.md) to compare long, short, and total exposure in the same leverage mode. A ratio does not locate liquidations; it is a separate market-level view.
3. Inspect the same market in [Liquidation Heatmap](https://docs.thekingfisher.io/products/liquidations/liquidation-heatmap.md). Keep its symbol, leverage mode, intensity view, and time window consistent when comparing snapshots.
4. Check the relevant price interval in [AGGR Chart](https://docs.thekingfisher.io/tools/chart) and [AGGR Trades](https://docs.thekingfisher.io/tools/trades). Ask whether actual price and executed trades changed near the modeled area, or whether price moved through it without a notable response.
5. Record both supporting and contrary examples. A modeled bar is not a resting order, guaranteed liquidation, or price target; a reaction at one level does not show that the map caused it.

**Useful counter-checks:** Funding & OI for market context; Aggregated Orderbook for visible resting liquidity; Trades for prints. Each has different inputs and limitations.

## Workflow 2: test how options scenarios change a market interpretation

**Question:** Is an options-flow interpretation stable across reasonable spot, implied-volatility, and time-window assumptions?

1. In [GEX Charm Tide](https://docs.thekingfisher.io/products/options-alpha/gex-charm-tide), note the instrument and forward-window setting. Compare the base case with one spot or IV scenario change at a time.
2. Read the daily bars and cumulative line together. A cumulative endpoint can hide offsetting days, and the scenario controls change the calculation rather than the future market.
3. In [GEX Pin Corridor](https://docs.thekingfisher.io/products/options-alpha/gex-pin-corridor), inspect the current-spot marker, nearby rows, hinge, settlement markers, and displayed forward days. Change the visible corridor or resolution to check whether a pattern is local to the chosen view.
4. Compare [GEX Vol-Beta Factors](https://docs.thekingfisher.io/products/options-alpha/gex-vol-beta-factors) with its own selected historical range. Read β+ and β− separately and treat κ/persistence as a description of the sampled response, not a forecast.
5. Check price and actual flow independently. If the interpretation changes when a single model input moves slightly, record that sensitivity rather than presenting one scenario as a certain outcome.

**Useful counter-checks:** GEX+ for the widget’s current exposure view, Historical GEX for time context, and the market’s actual price/volume. Options widgets are conditional models; model brightness, a hinge, or settlement marker is not proof of a future pin or directional move.

## Workflow 3: distinguish visible liquidity from executed flow

**Question:** When visible depth changes near price, does executed trading confirm that the market responded?

1. In [Aggregated Orderbook](https://docs.thekingfisher.io/products/money-flow/aggregated-orderbook.md), set the market and venue set first. Note price granularity, visible levels, and update interval; then compare the spread and depth around the mid.
2. Revisit a notable level over several updates. Record whether it persists, moves, or disappears as price approaches. A single displayed wall can cancel or move.
3. Inspect the same interval in [AGGR Trades](https://docs.thekingfisher.io/tools/trades) or [Trades Lite](https://docs.thekingfisher.io/tools/trades-lite), then compare with [Toxic Orderflow](https://docs.thekingfisher.io/products/money-flow/toxic-orderflow) using the same market and interval. In Toxic Orderflow, compare exchange lines and Chart/Summary modes instead of collapsing disagreement into one score.
4. Use [Counters](https://docs.thekingfisher.io/tools/counters) or [Stats](https://docs.thekingfisher.io/tools/stats) only after checking their interval, mode, and units. Align timestamps before drawing a conclusion.
5. Write down the alternative explanation: hidden liquidity, cancellations, venue mismatch, spread changes, or a delayed feed can make the snapshots disagree.

Visible size alone does not establish support, resistance, absorption, spoofing, or trader intent. Repeated persistence and actual trade response are stronger observations, but still do not establish a profitable rule.

## Workflow 4: compare crowding and activity without confusing the measures

**Question:** Do price, open interest, funding, and traded-volume splits tell a consistent story over the same interval?

1. In [Funding & OI](https://docs.thekingfisher.io/products/money-flow/funding-and-oi), select the market chips and timeframe. Use the synchronized crosshair to compare Open Interest, Funding Rate, and OI/Volume at the same timestamp. Record whether values are absolute or percentage-based.
2. In [Long vs Shorts](https://docs.thekingfisher.io/products/money-flow/long-vs-shorts.md), record the selected time window and exchange set. Compare exchange rows with the aggregate; aggregate percentages are volume-weighted and represent classified traded volume, not an open-position ledger.
3. Use [LiqRatio](https://docs.thekingfisher.io/products/liquidations/liqratio.md) for its distinct modeled exposure split and [LiqMap](https://docs.thekingfisher.io/products/liquidations/liquidations-maps.md) if you need a location view. Do not treat open interest as direction or infer which side opened positions from OI alone.
4. Compare the readings with price and executed trades. If the widgets disagree, preserve the disagreement and investigate venue, units, update time, and measurement definitions.

Funding conventions differ by venue. A crowded-looking snapshot is a question to investigate, not evidence that a squeeze or reversal must follow.

## Workflow 5: find relative-strength rotation, then check absolute behavior

**Question:** Is relative leadership broad and persistent, or limited to a few assets and one timeframe?

1. In [RSI Rotation](https://docs.thekingfisher.io/products/money-flow/rsi-rotation), choose a timeframe and start with all tickers. Use the quadrant filters to explore Leading, Lagging, Improving, or Weakening names; inspect a ticker’s trail as well as its current quadrant.
2. Compare with [RSI Heatmap](https://docs.thekingfisher.io/products/money-flow/rsi-heatmap) using its displayed timeframe and scale. Then check [Markets](https://docs.thekingfisher.io/tools/markets) or [Markets Surge](https://docs.thekingfisher.io/products/money-flow/markets-surge) for the corresponding market-wide view.
3. Change to a slower timeframe and check the underlying price/volume chart before interpreting relative rank. Relative strength can improve while an asset’s absolute price still falls.
4. Log whether leadership persists across windows and how many assets participate. A point’s quadrant is relative to the selected peers; it is not a prediction.

## Workflow 6: examine a basket or relationship without mistaking correlation for P\&L

**Question:** Does a basket relationship remain coherent across components and historical windows?

1. In [KF Index](https://docs.thekingfisher.io/products/money-flow/kf-index.md), begin with a small basket. Inspect signed weights and component overlays before adding more assets.
2. Compare Corr/Beta across more than one historical window and inspect whether one constituent dominates the path. Correlation and beta describe a sample; they do not show that a hedge will work in the next regime.
3. Review any displayed funding or exposure estimate as a cost/risk check, not as a guaranteed fill or return.
4. If available through Bazar, inspect [BetaPlay](https://docs.thekingfisher.io/products/money-flow/betaplay.md) as a separate view.
5. Validate a candidate relationship with an independent return calculation that includes fees, slippage, funding, execution rules, and a later holdout period. A chart relationship alone is not an alpha result.

## Related guides

* [Trading with the Kingfisher](https://docs.thekingfisher.io/trading-workflows/trading-with-the-kingfisher.md)
* [Liquidation map reading example](https://docs.thekingfisher.io/trading-workflows/liquidation-map-reading-example.md)
* [How to read a crypto liquidation map](https://docs.thekingfisher.io/trading-workflows/how-to-read-a-crypto-liquidation-map.md)
* [AGGR Widgets](https://docs.thekingfisher.io/tools/aggr-widgets.md)
* [Kingfisher Widgets](https://docs.thekingfisher.io/products/liquidations.md)
