Mdna Tone Delta
In plain terms
Read each year's 10-K through a finance-specific positive/negative word filter. Companies whose tone got more positive year-over-year tend to outperform after the filing date; those whose tone got more negative tend to lag.
How it works
Loughran-McDonald 2011 JF demonstrate the *tone* of 10-K filings — positive vs negative finance-specific lexicon — predicts forward returns. Distinct from filing_text_delta's *uncertainty* lexicon. Effect: bottom-decile tone-delta underperforms top-decile by ~2-3% over 4 weeks, persisting ~12 weeks. Today consumes the cached Item 1 (Business) text; switches to Item 7 (MD&A) when the EDGAR pipeline emits it (tracked in docs/alpha-research/proposals/).
Live results
0 times picked on its own · 128 times inside a blend (124 beat the stock) · updated 2026-06-06Data dependencies
- SEC 10k business text
A data feed this strategy reads, refreshed on its normal schedule.
Expected edge
- Reported return
- ~2-3% over 4 weeks post-filing in the 1994-2008 source paper
- Tested over
- 1994-2008
Loughran-McDonald 2011: ~2-3% over 4 weeks; ~1-2% OOS post-publication.
Related families
Counts uncertainty/risk words ('may', 'could', 'challenging') in 10-K Item 1 (Business) YoY. A spike means management is privately more worried — bearish.
If a company's 10-K barely changes year-over-year, the business is boring-and-steady and outperforms. Big text changes signal hidden bad news.
The short side of 'lazy prices': firms that rewrote their 10-K heavily YoY underperform by 188 bps/month. Most of the lazy-prices alpha lives on this short leg.
Measures whether management's language on the earnings call got more positive or negative quarter-over-quarter. Tone drift predicts price drift.
Explore Mdna Tone Delta on alphactor.ai
See which tickers this family is currently firing on, with live signals and rankings.