July First Week: ‘Sentiment’ Drives KOSPI Past 3,180
- 오리 오리
- Jul 13
- 3 min read
Updated: Jul 14
1. Introduction: Piercing Market Psychology with Data
The paradigm of financial analysis is shifting. Gone are the days of relying solely on financial statements and macroeconomic indicators—today, investors are turning to alternative data such as social-media buzz, satellite imagery, and news headlines in search of actionable signals. At the heart of this revolution is the effort to quantify market participants’ collective mood—the so-called sentiment—a factor long regarded as unmeasurable.
This report extends that effort. Focusing exclusively on the period June 27–July 10, 2025, we conduct a deep dive into a proprietary News Sentiment Index derived from headline data, without recourse to any outside information. Our goal: determine whether systematically scored sentiment data can meaningfully explain—and possibly help forecast—movements in the KOSPI, the VIX “fear gauge,” and the USD/KRW rate.
2. The Market Thermometer: Tracking News Sentiment
Our News Sentiment Index scores each headline by matching positive tokens (e.g. “boom,” “rebound”) with +10 to +20 weights and negative tokens (“decline,” “slump”) with –10 to –20. Summing these per day yields the Daily Sentiment Score—a gauge of the information environment’s “temperature.”
A quick line chart would show:
Consistently positive territory (all >0),
Sharp spikes (0.44 → 1.16 → 1.68) indicating bursts of strongly bullish headlines,
Volatility in “mood” rather than a steady trend—implying a market sensitive to discrete news events.
The 1.68 peak on July 3 serves as our anchor point: the day of highest optimism, and a critical reference for how price, volatility, and FX responded.
3. The Dance of Sentiment and Key Metrics
3.1 KOSPI vs. Sentiment: Coincident, Not Leading
Overlaying KOSPI’s close with daily sentiment reveals:
June 30: Sentiment at trough (0.44), yet KOSPI +0.52%—shows sentiment ≠ absolute driver.
July 1–3: Sentiment jumped from 1.16→1.68 and KOSPI rallied from 3,089.65→3,116.27 (+1.34% on July 3). Here they rose together, highlighting a strong coincident relationship.
July 4: KOSPI fell –1.99%, but no sentiment data to confirm whether headlines turned negative or profit-taking prevailed—an unavoidable data gap.
Conclusion: The index functions best as a trend-confirmation tool rather than a pure leading indicator—bullish headlines and rising prices feed into each other in real time.
3.2 VIX & USD/KRW: Risk Appetite and Capital Flows
VIX (Fear Gauge): As sentiment climbed 1.16→1.68 (July 1–3), VIX eased 16.83→16.38, confirming a classic risk-on phase.
USD/KRW: Paradoxically, during that same risk-on window, USD/KRW rose 1,355.59→1,357.10 (won weakened).
Possible explanations:
Global dollar strength outweighed domestic optimism.
Headline bias toward export-boosting stories (“semiconductor export boom”) that favor a weaker won.
Participant bias: Our index may more accurately capture domestic investor mood, not global flow dynamics.
5. Conclusion: Promise and Limits of News Sentiment
Our proprietary News Sentiment Index demonstrated clear value as a real-time barometer of market mood—closely tracking KOSPI and VIX during July 1–3’s rally. Yet its short sample, data gaps, and bias toward domestic headlineslimit its role as a standalone forecasting tool.
Key lessons:
Correlation ≠ causation: Positive headlines and rising prices feed on each other.
Segment the audience: Domestic vs. global sentiment can diverge, especially in FX.
Expand and refine: Longer histories, lag/lead analysis, and source segmentation (specialist vs. general press) are next steps.
In short, quantifying sentiment opens a unique window into market psychology—but must be wielded alongside traditional metrics and a clear understanding of its scope and biases.
Comments