How Opinion Alpha scores finance influencers
Last updated 2026-11-13
Opinion Alpha scores an influencer by simulating what would have happened if you had followed their public stock opinions: each opinion becomes a position on the next US trading day, is held for about three months, and the result is compared with the S&P 500 (SPY). Questions about the method: support@opinionalpha.app.
How is a stock opinion extracted?
Every new video is transcribed and every post is read, and each stock opinion in it is recorded with five things: the ticker (including non-US listings), the direction (bullish or bearish), the horizon (short, medium or long term), how firm it is (an explicit call, a leaning, or a conditional view such as “if it breaks 200”), and the quote in the influencer’s own words with a link to the source.
A stock mentioned only in passing, with no view on it, is not an opinion and is not scored.
How does an opinion become a simulated position?
The position opens on the next US trading day after the opinion was published, never on the day itself, so the simulation cannot use a price the audience could not have traded at. It typically closes about three months later. Repeating the opinion keeps the position open.
What are the three ways of following an influencer?
Conviction-weighted holds the stocks the influencer is bullish on, weighted by how often they mention them, so the calls they repeat count more. Long-only holds every bullish call equally. Long-short also shorts the stocks they are bearish on.
The leaderboard and every influencer page can be switched between the three. Conviction-weighted over the last three months is the default.
What do the numbers mean?
Return is the simulated cumulative return over the selected period. Excess return is that return minus SPY’s over the same period. Win rate is the share of completed simulated positions that ended with a positive return. Sharpe is the average return per unit of volatility. Drawdown is the largest fall from a peak to a later low in the simulated portfolio.
How are influencers ranked?
By simulated return over the selected period and way of following, with win rate and Sharpe ratio shown beside it. Periods run from one week to two years. Nobody can pay to be ranked, listed or removed.
What are the limits of this method?
The results are simulations, not anyone’s real trades: they ignore trading costs, taxes, position sizes the influencer actually used, and anything said in private or paid channels. A short period or a small number of opinions makes a ranking noisy; check the number of calls next to it.
Extraction is automatic and reviewed, and it can still misread a ticker, a direction or sarcasm. Every opinion links to its source so it can be checked, and can be reported if it is wrong.
A high simulated return in the past does not mean the next call will work. This is not investment advice.