The Stockpulse Brief, August 2026

In this edition we highlight the following topics: new Korean retail sources go live, prediction market and GitHub data get a proper structure, and a white paper on trading earnings with social signals. Plus, DIE ZEIT (large German newspaper) asked us what Trump’s Truth Social posts are actually worth to markets.

1. Korean Retail Chatter: New Sources, More Signal

Data · New Sources

Social media usage runs far higher across much of Asia than in Europe or the US — and financial social platforms are no exception. Over the past few months we’ve extended our coverage to the busiest of them in South Korea.

July’s chip rout showed why this matters. Korean retail investors had piled billions into leveraged ETFs on Samsung Electronics and SK Hynix; when semiconductors corrected, those bets unravelled, and regulators responded by halting new listings and tripling the minimum retail deposit to about $22,000. Conversations that loud leave a trace long before they show up in price.

The sources we’ve added all show heavy retail engagement — several now rank among our highest by post volume. A selection of what we’re tracking: Toss Invest, Naver, and Paxnet.

It’s ticker-mapped, and the discussion isn’t limited to Korean names. Nvidia, Micron and other US and European stocks come up constantly. Alongside real-time sentiment, historical data from these sources is available for deeper analysis.

This is another step towards our goal to provide the broadest and global coverage of social platforms for financial discussions.

2. More Alternative Data Sources

Data · New Perspectives

We mentioned in earlier newsletters that we’d started monitoring prediction markets and GitHub. Neither source is new, but both have become considerably more relevant for quantitative analyses over the past few months.

The questions we’ve had about them have surprised us — not just how many, but who they came from. Interest has been much broader than we expected.

So we’ve built more structured views on the data. The raw volume from prediction markets and GitHub repositories is large enough that it needs heavy aggregation before it’s usable. One example: periodic snapshots of the most relevant fields, which make rankings quick to build and surface trends that the raw data hides.

We’ve also collected a good amount of history, particularly from prediction markets, and that data is ticker-mapped. Please reach out if you’d like more detail, or just reply to this email.

3. A Social-Media Signal Overlay On Earnings Releases

New White Paper

Corporate earnings announcements are among the most information-dense, highest-variance events in the equity calendar. In the days surrounding a release, retail and professional attention concentrates on a single name, and that attention leaves a measurable footprint across social and news media in the form of message volume (buzz) and directional tone (sentiment). The hypothesis is built on the premise that abnormal pre-earnings buzz and sentiment carry predictive information about the direction of the post-announcement move, and that this information can be harvested systematically — but only when calibrated to the idiosyncratic behaviour of each individual security.

We developed a strategy which trades the Nasdaq-100 universe around scheduled corporate earnings events, taking directional (long or short) positions that are triggered by abnormal social-media buzz and sentiment patterns identified on a per-security basis.

The return stream is attractive on a risk-adjusted basis: a Sharpe ratio of 1.56 and a Sortino ratio of 2.28 are paired with a contained maximum drawdown of −24.9% and a Calmar ratio of 1.42. Roughly 63% of calendar months were positive, and the strategy maintained a balanced long/short posture (273 long versus 258 short trades) — a structural property that reduces directional dependence on the broad equity beta.

Please reach out if you want to read the full analysis or simply reply “Earnings White Paper” to my email.

4. Are $100k Per Month Worth It To Get Early Access To Trump’s Tweets On Truth Social?

News Coverage

We helped German newspaper DIE ZEIT to get a better understanding of an answer.

We are tracking the platform Truth Social already for quite some time, with special focus on Trump’s posts.

For DIE ZEIT we looked closer at 2,558 tweets of Trump since January 2025. We systematically processed these posts with LLMs. 103 posts were classified as financially relevant. We calculated a sentiment score for each of them with a potential impact on the price developments of the S&P 500 and oil.

The relationship was staggering in quite a few cases; not really surprising probably.

The German newspaper featured an article on August 6th in its print and online edition. You can read the German article here.