What 'Smart Money' actually means in the Tapeline Score (and why it's not what you think).

"Smart money" is the most misused phrase in retail finance. Open any trading subreddit, scroll any finance TikTok, look at any newsletter sales page — somebody is selling you "what the smart money is doing." Almost always, what they mean is "what one famous person on CNBC said in a clip yesterday." That's not smart money. That's TV.

Tapeline's Smart Money factor is 15% of the composite score (see the full formula). It's a real number, sourced from real filings, with real lags. This post is the deep dive on what it actually measures and where the limitations are — because a 15% weight in our scoring engine deserves a paragraph more than "trust us, we're tracking the smart money."

The three data sources behind the factor

Smart Money sums to a 0–100 sub-score from three independent data streams, each with its own lag and signal-to-noise characteristics:

1. Congressional disclosures — required by the STOCK Act (Stop Trading on Congressional Knowledge Act, 2012). US House and Senate members must disclose trades over $1,000 within 30–45 days. The signal: when multiple members on relevant committees buy or sell the same name, that's information they plausibly had access to that the market didn't.
2. Elite 13F filings — quarterly position disclosures by institutional managers with over $100M AUM, required by the SEC. Tapeline tracks eight curated managers whose long-horizon track records justify the lag: Buffett (Berkshire), Burry (Scion), Tepper (Appaloosa), Ackman (Pershing Square), Druckenmiller (Duquesne), Laffont (Coatue), Coleman (Tiger Global), Singer (Elliott). Lag: 45 days post quarter-end.
3. Insider Form 4 filings — required by the SEC within 2 business days of any insider transaction (executives, directors, 10%+ owners). The signal: insiders are the only buyers who know more about the company than the market, by definition. Clusters of buying — multiple insiders in the same window — are a stronger signal than single-buyer events.

What "smart money buying" actually predicts

Each data source has a different predictive horizon. Let me walk through the cases that matter:

Congressional buying works best on names where committee members have informational access — defense contractors near a relevant Armed Services committee member, healthcare names near a Finance committee member, regulatory beneficiaries before a relevant ruling. The base rate of edge is small but non-zero; academic studies (Ziobrowski et al., Belmont & Sayers) have shown weak positive alpha on a portfolio basis. The Tapeline signal weights Congressional flow by relevance — committee assignment + filing volume + recency — not raw transaction count.

Elite 13F changes are the strongest absolute signal but the longest lag. When Buffett enters a new position, the position was likely opened 30–90 days before the filing becomes public. By the time you see it, the bigger move has often happened. The Tapeline signal weights 13F deltas not by the absolute position size but by conviction — new positions vs adds-to-existing vs trims vs exits, scaled by portfolio percentage. A new full position in Burry's Scion fund is signal-rich; a 5% trim from a large Tiger Global position is noise.

Insider Form 4 filings have the shortest lag (1–3 business days) and the highest signal-to-noise for cluster events. Single-insider buys are weak — executives buy for compensation reasons, exercising options is mechanical, charity donations get filed too. Multi-insider buys in the same window where the executives have no scheduled compensation event are what the score weights. Selling clusters are downweighted (they can mean tax planning, diversification, or genuine signal — hard to disambiguate).

Why Smart Money is 15% weight, not higher

A natural retail-trader question: if Smart Money is so signal-rich, why isn't it 30% of the composite or higher? Three reasons:

The lags compound. 13F is 45 days late. Insider buys are sometimes timed to compensation cycles that look like signal but aren't. Congressional filings can be 30–45 days late and include trades that have already been unwound. By the time the data is clean and public, much of the move has happened.

It's a confirmation factor, not a leading one. Smart money flow is most useful in confluence with the other factors — when Trend, Relative Strength, and Smart Money all agree, that's the highest-conviction setup. Smart Money alone is late information; combined with leading factors it becomes directional certainty.

Survivorship in the source data. The 8 elite 13F filers we track were selected for long-term outperformance. They're also the most-watched fund managers in the world — their moves are crowded trades by the time the filing publishes. Buffett buying Apple in 2016 was signal; Buffett buying Apple in 2024 was a market price-anchor, not new information.

How the Tapeline score uses it differently from competitors

Most "smart money" scoring in retail tools is broken in one of two ways: either it's a single-source (just 13F, or just Congressional) which misses the confluence signal, or it's opaque (Tipranks' Hedge Fund Sentiment is a Smart Score input but the weighting and the underlying fund list are not published). Tapeline:

• Publishes the 8 elite-fund list on /app/holdings (Premium feature).
• Combines Congressional, 13F, and Form 4 into a single 0-100 sub-score with published methodology.
• Weights the sub-score at 15% of the composite — high enough to matter, low enough not to drown out the leading factors when smart money is late or noisy.
• Surfaces the actual data feeds: the Premium tier exposes the underlying Congressional trades feed at /app/congress and the elite 13F holdings at /app/holdings — not just the aggregated score.

What to actually do with this

Don't treat Smart Money as a trigger on its own. Treat it as a confluence multiplier:

• A 90 Smart Money sub-score on a 40 composite is a value signal — institutions are positioning before the market has rerated it. Worth a watchlist add.
• A 90 Smart Money sub-score on a 75 composite is confirmation — the smart money is in a setup that's already showing up in Trend, RS, and Momentum. Standard signal-of-signals.
• A 30 Smart Money sub-score on a 75 composite is a yellow flag — strong setup, but institutions and insiders aren't confirming. Worth understanding why before sizing up.
• A 90 Smart Money sub-score with no other factor confirming is curious but not actionable. Maybe insiders are buying for a reason the market hasn't seen yet; maybe they're wrong.

The point of breaking out the sub-score is exactly this kind of nuance. The composite gives you a summary; the breakdown lets you read where the conviction actually lives, and where it's conspicuously absent.

You can see live Smart Money sub-scores on any ticker page — e.g. /t/NVDA, /t/AAPL — or filter by it on the live scanner. The full Congressional trades feed and elite 13F holdings are Premium features at /app/congress and /app/holdings; the score weighting itself is free for everyone, every row, every day.