swashq
A few metrics I'd want to see:
- Revenue quality: Recurring vs one-time. A company with 80% recurring revenue is worth more than 100% transactional at same margins. - ROIC trend over 5-10 years, not just current. Declining ROIC often precedes margin compression. - FCF yield vs earnings yield. Big divergence = accounting shenanigans or capex cycle. - Insider buying/selling. Insiders are often early on inflection points.
For workflow: ability to save screens and get alerts when new stocks qualify would be sticky.
What data source are you using for fundamentals?
Window functions (tests optimizer): SELECT , ROW_NUMBER() OVER (PARTITION BY category ORDER BY date DESC) FROM large_table
Large aggregations (tests hash tables): SELECT category, COUNT(), AVG(value), PERCENTILE_CONT(0.5) WITHIN GROUP (ORDER BY value) FROM large_table GROUP BY category
Self-joins (tests join algorithms): SELECT a., b. FROM table a JOIN table b ON a.foreign_key = b.id WHERE a.date BETWEEN '2024-01-01' AND '2024-12-31';
String operations (often overlooked): SELECT * FROM table WHERE column LIKE '%pattern%';
Also test with varying data sizes (1K, 100K, 10M rows). Some engines optimize differently at scale.
Both are low-effort but compound, you build intuition without committing to deep analysis.
Add market context to the coaching. Right now most finance apps show spending in isolation. But "should I invest?" is different when: - S&P is at ATH vs down 20% - Fed just raised rates vs cutting - Your industry is laying off vs hiring
A few data points that could make the coaching smarter: - Historical market returns (so "invest in index funds" has numbers behind it) - Inflation rate (to show why cash loses value) - Average savings rate by age bracket (peer comparison)
The "memory" system you built could remember their risk tolerance AND market conditions when they made past decisions, then reference both.