Asset-Specific Sentiment Inversions in Financial NLP
Standard sentiment models assign polarity to text in isolation. In financial NLP, that's often wrong. The same phrase can be bullish for one asset and bearish for another. swik catalogs these systematic inversions — instances of context-dependent polarity — so inference engines can correct for them. This is the core challenge of Aspect-Based Sentiment Analysis (ABSA) applied to financial markets.
35
assets covered
124
inversion entries
CC BY 4.0
open license
47k+
labeled headlines
The Inversion Problem
A sentiment inversion occurs when the naive polarity of a financial headline differs from the actual directional impact on a specific asset's price. These are not random errors — they are systematic inversions rooted in supply chain relationships, macro linkages, and domain knowledge that general language models don't encode.
Solving this requires context-dependent polarity: the same phrase scored differently depending on which asset you're analyzing. This is the financial NLP equivalent of ABSA — and it's why asset-specific sentiment signal correction matters for any serious application in trading or quantitative research.
Active inversion catalog
Live data from the swik community catalog. Updated as entries are confirmed.
"Russia sanctions"
Russia is major aluminum supplier
naive:
negative
→
bullish
"power shortage"
Aluminum smelters shut down = less supply
naive:
negative
→
bullish
"China PMI falls"
Weak Chinese demand hits commodity exports, hurts AUD
naive:
negative
→
bearish
"commodity rally"
Higher commodity prices support commodity-exporting AUD
naive:
positive
→
bullish
"OPEC cut"
Reduced supply = price support
naive:
negative
→
bullish
"inventory build"
Supply glut depresses prices
naive:
positive
→
bearish
"ETF inflows"
Spot ETF inflows represent new institutional capital entering BTC, directly supporting price.
naive:
positive
→
bullish
"ETF outflows"
Sustained ETF outflows signal institutional de-risking and reduce demand for BTC.
naive:
negative
→
bearish
"SEC approves ETF"
ETF approval opens BTC to institutional capital that cannot hold spot crypto directly. Historically triggers significant price appreciation as new demand channels open (e.g. January 2024 spot ETF approval).
naive:
positive
→
bullish
"SEC rejects or sues"
Regulatory action from SEC signals hostile US regulatory environment, triggering institutional risk-off and retail panic selling. Short-term strongly bearish regardless of long-term outcome.
naive:
negative
→
bearish
"bitcoin crashes"
Direct price action — BTC crashing is bearish.
naive:
negative
→
bearish
"bitcoin surges"
Direct price action — BTC surging is bullish.
naive:
positive
→
bullish
"crypto winter"
Prolonged bear market signal — crypto winter implies sustained selling pressure across all crypto assets.
naive:
negative
→
bearish
"exchange collapse"
Exchange failures destroy confidence and force selling across the market (e.g. FTX collapse).
naive:
negative
→
bearish
"exchange hack"
Reduces confidence in crypto infrastructure
naive:
negative
→
bearish
"halving"
Supply issuance cut 50% — historically precedes bull run
naive:
neutral
→
bullish
"halving approaches"
Bitcoin halving reduces new supply issuance by 50%, historically triggering bull cycles 6-12 months later.
naive:
neutral
→
bullish
"rate hike"
Risk-off rotation out of speculative assets
naive:
positive
→
bearish
"El Nino"
El Niño brings drier conditions to West Africa, damaging cocoa crops
naive:
neutral
→
bullish
"harmattan winds"
Dry northerly winds damage cocoa pods in Ghana/Ivory Coast
naive:
neutral
→
bullish
"Brazil frost"
Cold damages coffee trees, reducing future supply
naive:
negative
→
bullish
"off-year production"
Coffee trees produce less every other year (biennial effect)
naive:
negative
→
bullish
"China slowdown"
China accounts for ~55% of global copper demand
naive:
negative
→
bearish
"green energy"
EVs and wind/solar require 3-4x more copper than conventional
naive:
positive
→
bullish
"mine strike"
Supply disruption reduces available copper
naive:
negative
→
bullish
"La Nina"
La Nina typically causes drought in South American corn belt
naive:
neutral
→
bullish
"bumper crop"
Record supply depresses prices
naive:
positive
→
bearish
"drought"
Crop damage reduces supply
naive:
negative
→
bullish
"synthetic alternative"
Cheaper polyester captures market share from cotton
naive:
neutral
→
bearish
"China slowdown"
Germany's auto/industrial exporters depend heavily on China
naive:
negative
→
bearish
"energy crisis"
High energy costs squeeze German industrial margins
naive:
negative
→
bearish
"gas price drop"
Gas peakers set the marginal price in most markets — cheaper gas directly lowers the electricity price ceiling
naive:
negative
→
bearish
"heatwave"
Extreme heat drives air conditioning demand, pushing electricity consumption and spot prices sharply higher
naive:
negative
→
bullish
"mild weather forecast"
Low heating/cooling demand reduces electricity consumption, bearish for spot prices regardless of supply
naive:
positive
→
bearish
"nuclear outage"
Nuclear provides cheap baseload — outages remove large blocks of low-cost supply, forcing expensive gas peakers to set the price
naive:
negative
→
bullish
"record wind output"
High wind generation floods the grid with zero-marginal-cost power, collapsing spot electricity prices
naive:
positive
→
bearish
"solar generation record"
Record solar output during midday hours drives negative or near-zero spot prices in well-connected markets
naive:
positive
→
bearish
"network congestion"
High gas fees = more ETH burned = deflationary pressure
naive:
negative
→
bullish
"ECB dovish"
Dovish ECB signals lower rates ahead, reducing EUR carry appeal and narrowing the rate differential with USD.
naive:
negative
→
bearish
"ECB hawkish"
Tighter ECB policy supports EUR
naive:
positive
→
bullish
"ECB signals cuts"
Rate cut signals from ECB reduce EUR yield premium, weakening the currency against USD.
naive:
negative
→
bearish
"Fed hike"
Higher US rates attract capital to USD, weakening EUR/USD
naive:
positive
→
bearish
"bullish bias"
Analyst bullish bias on EUR/USD signals expected appreciation; market positioning supports upside.
naive:
positive
→
bullish
"dollar firms"
Dollar firmness directly implies EUR/USD weakness — the pair moves inversely to USD strength.
naive:
positive
→
bearish
"dollar index rises"
DXY and EUR/USD have strong inverse correlation — rising dollar index directly depresses EUR/USD.
naive:
positive
→
bearish
"dollar rallies"
Dollar rally = EUR/USD decline by definition. Direct inverse relationship.
naive:
positive
→
bearish
"dollar remains firm"
Sustained dollar strength keeps EUR/USD suppressed; firmness implies continuation of downtrend.
naive:
positive
→
bearish
"dollar strengthens"
EUR/USD is inverse of DXY by definition
naive:
positive
→
bearish
"energy crisis"
European energy crises (gas supply disruption, price spikes) damage eurozone growth outlook and weaken EUR/USD.
naive:
negative
→
bearish
"energy shock"
Europe is heavily energy-import dependent — energy price shocks hit the eurozone economy disproportionately, weakening EUR relative to USD.
naive:
negative
→
bearish
"euro drops"
Direct price action signal — euro dropping means EUR/USD is declining.
naive:
negative
→
bearish
"euro extends losses"
Momentum continuation — euro extending losses signals persistent selling pressure on EUR/USD.
naive:
negative
→
bearish
"euro falls"
Direct price action — euro falling = EUR/USD bearish.
naive:
negative
→
bearish
"euro gains"
Euro gaining ground = EUR/USD appreciating.
naive:
positive
→
bullish
"euro rallies"
Direct price action signal — euro rallying means EUR/USD is rising.
naive:
positive
→
bullish
"euro strengthens"
Euro strength directly translates to EUR/USD upside.
naive:
positive
→
bullish
"greenback strengthens"
Greenback = USD — stronger USD means weaker EUR/USD.
naive:
positive
→
bearish
"risk appetite improves"
Improving risk appetite reduces safe-haven USD demand, allowing EUR to recover against the dollar.
naive:
positive
→
bullish
"risk-off"
Risk-off environments drive capital into USD as the global safe-haven, weakening EUR/USD.
naive:
negative
→
bearish
"Norwegian outage"
Norway is Europe main pipeline gas supplier; outages immediately tighten supply
naive:
negative
→
bullish
"Russian gas supply cut"
Russia supply cuts remove significant European gas supply, driving TTF prices sharply higher
naive:
negative
→
bullish
"high storage levels"
Full storage removes urgency to buy spot gas, suppressing TTF prices heading into winter
naive:
positive
→
bearish
"warm winter forecast"
Reduced heating demand lowers European gas consumption, bearish for TTF
naive:
positive
→
bearish
"BOE hike"
Higher UK rates attract capital to GBP
naive:
positive
→
bullish
"UK recession"
Weak growth = BOE more likely to cut rates
naive:
negative
→
bearish
"central bank buying"
Central bank gold purchases reduce available supply and signal institutional confidence in gold as reserve asset.
naive:
positive
→
bullish
"dollar index rises"
Gold is priced in USD — a stronger dollar makes gold more expensive for foreign buyers, reducing demand. DXY strength and gold price have a strong inverse correlation historically.
naive:
positive
→
bearish
"geopolitical tensions rise"
Geopolitical risk drives safe-haven demand for gold — negative world events are bullish for gold prices.
naive:
negative
→
bullish
"gold drops"
Direct bearish price action.
naive:
negative
→
bearish
"gold falls"
Direct price action — gold falling is bearish.
naive:
negative
→
bearish
"gold rallies"
Direct price action — gold rallying is bullish.
naive:
positive
→
bullish
"gold slides"
Downward price movement — bearish continuation signal.
naive:
negative
→
bearish
"gold surges"
Strong upward price movement — directly bullish signal.
naive:
positive
→
bullish
"rate hike"
Raises real yields, reduces gold's appeal as non-yielding asset
naive:
positive
→
bearish
"real yields negative"
Negative real yields eliminate the cost disadvantage of holding gold versus bonds, strongly supporting gold prices.
naive:
negative
→
bullish
"real yields rise"
Opportunity cost of holding gold increases
naive:
positive
→
bearish
"rising yields"
Higher real yields increase the opportunity cost of holding non-yielding gold. When Treasury yields rise, gold becomes less attractive relative to interest-bearing assets, driving capital outflows from gold.
naive:
positive
→
bearish
"safe haven"
Crisis demand flows into gold
naive:
positive
→
bullish
"strong dollar"
Gold priced in USD — inverse relationship
naive:
positive
→
bearish
"war escalates"
Armed conflict escalation increases safe-haven demand for gold significantly.
naive:
negative
→
bullish
"yields fall"
Falling yields reduce opportunity cost of holding non-yielding gold, increasing demand.
naive:
negative
→
bullish
"port inventory high"
Excess inventory reduces spot buying pressure
naive:
positive
→
bearish
"steel output cut"
Less steel production = less iron ore needed
naive:
negative
→
bearish
"LNG export terminal outage"
Outage reduces US LNG exports — bearish US gas but bullish global LNG spot prices
naive:
negative
→
bullish
"LNG facility restart"
Restart increases export capacity, reducing domestic gas prices and adding to global supply
naive:
positive
→
bearish
"Qatar production increase"
More Qatar supply increases global LNG availability, pushing spot prices down
naive:
positive
→
bearish
"cold snap"
Spike in heating demand
naive:
negative
→
bullish
"mild winter"
Less heating demand = less consumption
naive:
positive
→
bearish
"storage build"
Above-average storage reduces scarcity premium
naive:
positive
→
bearish
"10Y yield rises"
Higher long rates compress growth stock valuations (duration effect)
naive:
positive
→
bearish
"AI boom"
AI tailwind for major Nasdaq constituents
naive:
positive
→
bullish
"EV slowdown"
Battery-grade nickel demand falls
naive:
negative
→
bearish
"Indonesia export ban"
Indonesia is world's largest producer — export restrictions tighten supply
naive:
negative
→
bullish
"yen strengthens"
Strong yen hurts export earnings of Nikkei companies
naive:
positive
→
bearish
"yen weakens"
Weak yen inflates overseas earnings of Japanese exporters
naive:
negative
→
bullish
"demand destruction"
Confirms weakening consumption
naive:
negative
→
bearish
"inventory build"
Rising stockpiles signal oversupply, pushing prices down
naive:
positive
→
bearish
"inventory draw"
Falling stockpiles signal tight supply, pushing prices up
naive:
negative
→
bullish
"strong dollar"
Oil priced in USD — stronger dollar makes it more expensive globally, reducing demand
naive:
positive
→
bearish
"EV adoption"
EVs don't need catalytic converters — reduces palladium demand
naive:
positive
→
bearish
"stricter emissions"
More catalyst loading per vehicle required
naive:
negative
→
bullish
"diesel ban"
Diesel cars use platinum catalysts — fewer diesel = less demand
naive:
negative
→
bearish
"hydrogen economy"
Platinum is key catalyst in hydrogen fuel cells
naive:
neutral
→
bullish
"refinery outage"
Reduced supply of refined product
naive:
negative
→
bullish
"rate hike"
Like gold, higher real yields reduce silver's appeal
naive:
positive
→
bearish
"solar demand"
Silver is key input in photovoltaic cells
naive:
neutral
→
bullish
"Argentina drought"
Argentina is world's largest soy meal exporter
naive:
negative
→
bullish
"China cancels orders"
China buys ~60% of traded soybeans
naive:
negative
→
bearish
"Fed pivot"
Rate cut expectations reduce discount rate, boost multiples
naive:
neutral
→
bullish
"rate hike"
Higher rates = higher discount rate = lower equity valuations
naive:
positive
→
bearish
"recession fears"
Expected earnings decline hurts stock prices
naive:
negative
→
bearish
"infrastructure stimulus"
Government spending on infrastructure = steel demand spike
naive:
positive
→
bullish
"property crackdown"
Less construction = less steel demand
naive:
negative
→
bearish
"ethanol parity"
When ethanol is profitable, Brazilian mills divert cane to ethanol, reducing sugar output
naive:
neutral
→
bullish
"monsoon failure"
India's sugar output falls when monsoon underperforms
naive:
negative
→
bullish
"oil falls"
Lower oil = weaker CAD = higher USD/CAD
naive:
negative
→
bullish
"oil rally"
Higher oil = stronger CAD = lower USD/CAD
naive:
positive
→
bearish
"SNB intervention"
SNB buys USD/sells CHF to cap franc strength
naive:
neutral
→
bullish
"risk-off"
Safe-haven demand for CHF = USD/CHF falls
naive:
negative
→
bearish
"BOJ tightening"
Narrowing rate differential weakens USD vs JPY
naive:
positive
→
bearish
"carry trade unwind"
Unwinding yen carry = buying yen = USD/JPY falls
naive:
negative
→
bearish
"risk-off"
Risk-off flows into yen (safe haven), USD/JPY falls
naive:
negative
→
bearish
"Black Sea corridor"
Export corridor open = more supply reaching market
naive:
positive
→
bearish
"Russia export ban"
Russia is world's largest wheat exporter — ban tightens supply
naive:
negative
→
bullish
How swik corrects for inversions
swik uses a two-layer inference architecture. A base language model provides an initial sentiment reading. The swik inversion catalog then applies asset-specific sentiment signal correction — overriding the naive reading when a known phrase inversion applies. This is what makes swik an asset-specific news analyzer rather than a generic sentiment classifier.
The catalog is community-maintained and open. Every inversion entry includes the phrase, naive polarity, actual directional impact, and economic reasoning. Entries move from hypothesis to active status through community voting: minimum 3 confirmations, 2:1 confirm-to-reject ratio required.
Open data
The full inversion catalog is released under Creative Commons Attribution 4.0 and available on GitHub and HuggingFace. Use it for fine-tuning models, cross-asset spillover analysis, backtesting sentiment signals, or building your own asset-specific sentiment APIs.