Stock Strategies
Explore 20 Quantellect stock strategies: 15 open development backtests with measured results and 5 blocked research records with their reasons.
Stock backtests are historical and hypothetical. Rules were chosen on the history tested; no independent holdout or live results. Backtests do not predict future results.
Keeps an annual top-ten dollar-volume roster and evaluates its reduce-only volatility exposure monthly.
Blends cross-sectional 12-minus-1-month momentum and Parkinson range-volatility ranks; monthly top thirty.
Ranks the sum of raw overnight log returns over 252 sessions; weekly top ten at the following open.
Ranks distinct disclosed buyers over 60 days; at least three buyers; excludes tickers with more sale dollars than purchase dollars; top ten.
Ranks lagged-market-beta residual return divided by daily range on high-volume days; monthly top twenty.
Ranks raw intraday responses following material overnight gaps; monthly equal-weight top ten.
Ranks upside through-origin market beta minus downside beta over 252 sessions, using SPY; monthly top ten.
Removes the 25 most-traded admitted tickers over 365 days, then ranks 90-day purchase clusters with at least two buyers; top five.
Ranks year-over-year trailing dividend-per-share growth; four payments in each year, 500 days of history and growth from 3% through 100%; top ten. Dividend amounts are signal inputs only.
Ranks acceleration in distinct disclosed buyers over 30 versus 90 days; at least two short-window buyers and velocity at least 1.5; top five.
Ranks reconstructed filing-date snapshots by distinct buyers in the last 14 days; at least two buyers, top five. Unknown plan status is included by the inherited metric convention.
Restricts native disclosures to Senate rows; at least two buyers over 90 days, net-sale filter, top five.
Ranks the buy share of two-sided open-market dollars over 120 filing days; at least $250,000 total, $50,000 purchases and 50% buy share; top ten.
Ranks role-weighted open-market purchase dollars over 90 filing days; at least two insiders, $100,000 total and $10,000 per trade. Officer/director/other multipliers are 1.5/1/0.5; top ten.
Ranks CEO/CFO purchases disclosed in the last 90 days, minimum $100,000, with positive 30-day net flow; top five. Unknown plan status is included by the inherited metric convention.
Quantellect Congress Party Flow is designed to compare disclosed purchasing and selling activity across political-party groups.
Blocked — no backtest resultsWe need historical party records and the original time-window rules, using only disclosures available on each date.
Quantellect Adaptive Fund Follow is designed to identify a group of strong institutional portfolios and then select stocks shared by several of those funds.
Blocked — no backtest resultsWe need the original fund filings and their amendments, with when each became available.
Quantellect Institutional Dollar Flow is designed to compare consecutive institutional filing snapshots and find stocks with growing reported market value across a broad set of tracked funds.
Blocked — no backtest resultsWe need comparable fund-filing snapshots for consecutive quarters, including late filings and amendments.
Quantellect Earnings Surprise Drift is designed to rank stocks after an earnings result differs materially from the analyst consensus that existed before the announcement.
Blocked — no backtest resultsWe need actual earnings announcement times and the analyst estimates available before each announcement.
Quantellect Dividend Growth Long/Short is designed to compare companies with strengthening dividend records against companies with weaker dividend characteristics, taking long and short positions under a pinned rebalance rule.
Blocked — no backtest resultsWe need validated short-selling calculations and historical share-borrowing availability, costs and dividend obligations.
How the board works
Evidence classes
Every number carries the evidence class it came from. The 15 completed reports are development backtests: the rules were chosen on the same history they were tested on, so the results are a starting point, not proof. No sealed historical holdout has been registered and prospective paper trading has not started for any of these strategies; when a paper record exists it will appear as its own panel, never blended with the backtest. Blocked strategies have no completed report and carry no badge.
Every published report keeps its original results
A report is published from a fixed file with a fingerprint. Corrections and new runs create a new version; older versions stay readable and point to their successor. Nothing is edited in place, no result is deleted, and the catalogue number of a strategy never changes.

