Dividend Hunting: dividend capture done with data. A model predicts which stocks recover after they pay, so you collect the dividend and keep your capital.
The Dividend Hunting Strategy
Collect the dividend.
Sell on the bounce.
Repeat tomorrow.
Dividend capture, done with data. The version that finally works, in five minutes a day.
Old-school dividend capture is a coin flip: buy before the ex-date, collect the dividend, and hope the price recovers. Dividend Hunting takes out the hope. A model trained on a million dividends tells you which stocks actually bounce back, and how fast, so you collect the payout and sell on the recovery.
Not day trading, not crypto, not leverage. We publish every pick and grade it the next day, so the whole record is yours to check, line by line.
21 days of the Pick of the Day before the market opens. No card.
Not ready for an account? Get today's Pick of the Day free, by email →
Trading carries risk and past results do not guarantee future returns. This is not financial advice.
Go ahead, ask an AI first
Ask ChatGPT if dividend capture works. It will say no. It is answering the wrong question.
Every AI model and Article about Dividend Capture says the same thing, and it is not wrong: the dividend is already priced in. The stock falls by roughly what you collect, so you pocket the payout, give it straight back on the price, and after tax you end up behind.
All true. And all about the blind, textbook version. That is not what we do.
The AI never answers the one question that matters: after a stock drops on its ex-date, which ones climb back, and how fast? That is The Dividend Hunting Strategy: dividend capture done with data, acting only on the stocks the model expects to recover the same day, backed by 26 years of public, graded results.
What the AI answers
"Does collecting a dividend make money on its own?" No. It is priced in: the stock falls by what you collect, and tax makes it worse. Correct, and not our claim.
What we answer
"Which of tomorrow's dividends will recover the drop on the same ex-date?" That is a prediction problem, and it is our whole product.
How to check us
Take our published picks, score them against what the prices actually did, in any tool you like. The record is open precisely so you do not have to trust our word.
Do not trust our word. Check it by yourself in under a minute.
Open the track record →Do not bother asking Claude, ChatGPT or Perplexity to grade our results. Chat AIs cannot pull clean historical prices, so they will just tell you they cannot check. You do not need them anyway. Every pick is a public stock on a public date, so you can verify it against a neutral chart with your own eyes.
- Take any row from our track record and note the ticker and the ex-dividend date.
- Open a free chart (TradingView, Yahoo Finance, or Google Finance), search the ticker, and set it to daily candles.
- The close the day before the ex-date is the buy price. On the ex-date the stock opens lower by about the dividend.
- See what happened next: back to the buy price is a Correct, part of the way is a Partial, still falling is a Miss. That is the result we published.
- The dividend amount is public on the same page. On a Correct trade you keep it in full, which is the Return we show.
Comfortable with data tools? Export our picks and run them against a real market-data API (Polygon, EOD Historical, or yfinance in a Python notebook) for an exact, automated audit.
Three steps. Five minutes a day.
The mechanism is older than the internet. What changed is that zero-fee brokers, fractional shares, and a model trained on a million dividends finally make the recovery edge work for small capital.
Buy before the ex-date
We surface the stocks going ex-dividend tomorrow that our model forecasts will recover. You buy one, the day before, on your own zero-fee broker.
Collect the dividend
On the ex-date the price drops by roughly the dividend. You are entitled to that dividend, typically 0.7 to 1.5 percent of your position, paid a few days later.
Sell on recovery
When the price recovers, often the same day, you sell. Your capital comes back intact and the dividend is yours to keep. Then you rotate into tomorrow's forecast.
Your capital does the same job over and over. The dividends stack up. The hard part, the only part, is forecasting the recovery.
No hype. Here is exactly how it goes.
Most trades win small. Some lose small. We show you both.
This is not a get-rich button. It is a small, repeatable edge, applied with discipline every day.
The win (most days)
Price recovers, you keep the dividend, and sometimes a little extra on the entry and exit.
The partial
Price only partly recovers, but the dividend still puts the trade in profit overall.
The loss (about 1 in 10 to 15)
Sometimes the stock keeps falling. Losses are usually small, 1 to 2 percent, and we publish them too.
Dividends coming up in the next few days
A live look at the board. Create a free account to see the recovery forecast and probability behind each one.
JSE:STXJGE
Ex-Date: Jul 22
Dividend Yield
Recovery
🔒
Probability
🔒
NSE:GRAPHITE
Ex-Date: Jul 20
Dividend Yield
Recovery
🔒
Probability
🔒
SZSE:001979
Ex-Date: Jul 21
Dividend Yield
Recovery
🔒
Probability
🔒
ATHEX:LAVI
Ex-Date: Jul 20
Dividend Yield
Recovery
🔒
Probability
🔒
BSE:QUESTCAP
Ex-Date: Jul 20
Dividend Yield
Recovery
🔒
Probability
🔒
NSE:BHAGERIA
Ex-Date: Jul 24
Dividend Yield
Recovery
🔒
Probability
🔒
See it work in two minutes
A quick walkthrough: find tomorrow's opportunity, read the recovery forecast, and place the trade.
Why this works now, and did not before
Three things changed. Together, they turned a pro-only tactic into something a small account can run.
Zero-fee brokers
When commissions ate 1 percent per round trip, a 1 percent dividend was pointless. With zero-fee brokers, the dividend is yours to keep.
Fractional shares
A $50 or $500 share price is no longer a wall. Fractional trading lets a small account put its whole balance into the single best forecast, every day.
A model that learns daily
Trained on nearly a million dividends since 1960 and retrained every day, so the forecasts adapt instead of going stale.
The math, before the proof
Why the numbers are not as crazy as they sound
Numbers this big sound too good to be true, and we felt that too. So here is the arithmetic in the open, with nothing hidden.
Same capital, recycled into one dividend a day. No leverage, no compounding assumed. The whole game is forecasting the days the price actually recovers, which is the one hard part, and the only part the model does.
And here is what actually happened
26 years of backtests. Every year positive, before and after tax.
Walk-forward tested across every market we cover (NASDAQ, NYSE, AMEX and beyond): each year forecast only with data available before it, never with hindsight. We put US front and center on purpose, because of all the markets we test it returns the least. So this is our floor, not a cherry-picked peak, and even here it was positive every single year. Dividends only, no compounding. The lines below show the gross result and what is left after tax for three example investors.
Average annual return, 2000 to 2025
Every single year positive, gross and after tax, through the 2008 crash and the 2020 crash. Backtested, dividends only, no compounding.
The three after-tax lines (US, UK, Denmark) are a low-to-high tax spread. Wherever you pay, your net return sits somewhere in that band.
Backtested, but not blind to the future, and now live in the open
Every year above was forecast using only the data available before it, so the model never saw the answer in advance. That is the honest part of a backtest. The newer part you can watch happen: since we went public in September 2025 we have posted every pick before its ex-date and graded it the next day, out in the open, and the live win rate has held at the same level as the 26-year backtest. You do not have to take the history on faith; you can follow the next call in real time.
Backtested and historical results, not a forecast of future returns. We publish the full year-by-year record and every individual pick behind it, so you can verify any number here against the source events. Past performance does not guarantee future results. This is not financial advice.
The part most tools hide
Tax takes a real bite. Here is roughly how much.
These dividends are taxable income, and short holds do not get the friendly long-term rates. We would rather show you the after-tax picture up front than let it surprise you later.
United States
Because the holding period is short, these dividends are non-qualified, so they are taxed as ordinary income. We model about 24 percent, a representative middle bracket. Your real rate runs from 10 to 37 percent.
United Kingdom
On US stocks the US withholds 15 percent at source under the treaty, which is then credited against your UK tax. The total lands near your UK dividend rate. We model about 34 percent, the higher rate; the basic rate is lower.
Denmark
Same 15 percent US withholding, credited against Danish tax. Danish share income above a low threshold is taxed at 42 percent, which is what we model: one of the highest dividend rates in the developed world.
A note on how the credit works, since it confuses almost everyone: the foreign withholding is not an extra tax stacked on top. It is a prepayment that is credited against the tax you owe at home, so for a higher-rate UK or Denmark investor the total is roughly the domestic rate, not the domestic rate plus the 15 percent. The lines above are flat, illustrative effective rates applied to the gross result, not a calculation for your situation. We are a data service, not tax advisors. Please confirm your own position with a qualified professional.
Not in one of these three? Read them as a low, middle and high marker and place your own country between them. The exact rate moves; the conclusion does not, the strategy stays clearly in profit after tax. And in a tax-advantaged account, such as an ISA, IRA, TFSA or pension, the drag can disappear entirely.
Straight answers to the questions you are already asking
Does dividend capture actually work?
The blind version does not. The dividend is already priced in: on the ex-dividend date the stock opens lower by about what it just paid, so you pocket the payout and lose roughly the same on the price, and after tax you are behind. The edge is not the dividend, it is the recovery. Some stocks climb back through that gap and some do not. We forecast which ones recover, and how fast, so the position closes flat or better while the dividend is kept. We tested this across every market we cover for 26 years and every year came out positive. US is the lowest-returning market of them all, and even there the strategy historically grew capital by roughly 130 to 170 percent a year after tax, so you more than double and come close to tripling your money. That is a demonstrated floor when you follow the rules, not a promise about the future.
What is dividend recovery forecasting?
It is the prediction of whether a stock will reclaim the price it loses on its ex-dividend date, and how quickly. A model trained on nearly a million dividends since 1960 scores each upcoming ex-dividend event with a recovery probability. You buy the day before, collect the dividend, and sell when the price recovers, often the same day.
How is this different from ordinary dividend capture?
Ordinary dividend capture buys any high-yield stock before the ex-date and hopes it recovers. We only act when the model forecasts the recovery, and skip the events where it does not. The dividend is the payout. The recovery forecast is the actual product.
Is it still profitable after tax?
In the backtests the strategy stays net positive after tax, but the drag is real and varies by country. A US investor pays ordinary income tax because the holding period is short. A UK or Denmark investor in US stocks has 15 percent withheld at source, which is then credited against domestic tax, so the total is roughly the domestic dividend rate. See the after-tax lines and the country notes above. This is not tax advice.
What happens on a losing trade?
Roughly one forecast in ten to fifteen does not recover and closes at a loss, usually small, around 1 to 2 percent. We publish the losing picks alongside the winners and grade every one the next day. Keeping losses small comes down to entry and exit, which matter as much as the pick. Entry: the model is trained on closing prices, so you buy as close to the market close before the ex-date as you can; holding for the shortest possible time cuts the chance the stock dips while you hold it ahead of the ex-date. Exit: set a take-profit order at your buy price and sell the moment it recovers, instead of waiting and hoping. See our guide to exit strategy →
Do I need a lot of money to start?
No. This strategy is built for small accounts and for building initial wealth. It runs on a small balance because zero-fee brokers remove commissions and fractional shares let even a modest amount go into the single best forecast each day. If anything, the limit is at the other end: a very large account placing a big order can move the price of the stock itself and trigger partial fills, and that slippage eats the edge. As a rule of thumb keep any single position under about 1 percent of the stock's average daily volume, which for the liquid names most forecasts land on means it works well up to roughly $100,000. Start with whatever you have and scale later.
Do I need a special broker?
Yes. You need a zero-fee broker. The whole strategy depends on keeping the full dividend, so a broker that takes a percentage of each transaction will erase the edge on every trade. Fractional shares help too, so a small balance can go into the single best forecast each day. See our guide to choosing a zero-fee broker →
Is it risky to put my capital into one stock at a time?
For a small account it is normal and fine. You are buying a real, liquid company on a public exchange and holding it for a day or so, not a leveraged CFD or a crypto token that can evaporate overnight. One trade a day is also manageable, whereas running several dividends at once tends to burn you out early, before you have the rhythm. Once the cadence is second nature you can split your capital across a few dividends a day, which is actually the smarter move. The point is to show up and do it. If you would rather commit only half your capital to a trade at first, that is fine too.
Get today's Pick of the Day, free
Drop your email and we send you today's Pick of the Day right now, no account needed. After that, a short daily recap so you can watch the strategy prove itself before you commit to anything.
You also get our Dividend Tracker spreadsheet, free. The Google Sheet we use to track every trade: log your buys, sells and dividends, and it works out your P&L, win rate, dividend pipeline and real ROI, charts included.
No spam. One email a day. Unsubscribe any time.
Start free. Upgrade for the forecasts in time to act.
The free plan shows every forecast after it is graded. Premium gets you the next one live, before pre-market, from $9.99/mo.
🎯 Basic
$9.99/mo
- ✓ Today's forecast, before pre-market
- ✓ AI recovery predictions
- ✓ The probability behind each dividend
- ✓ Stock & ETF screeners + watchlists
- ✓ Dividend calendar
- ✓ Public dividend trackers
💎 Hunter
$29.99/mo
- ✓ Everything in Basic
- ✓ Full instrument detail pages
- ✓ Hunting Grounds: custom backtests
- ✓ Pick of the Day at pre-market
- ✓ Private dividend tracker
- ✓ Exports, API, and more →
Every forecast graded in the open. Decide for yourself.
Free account, no card. See today's forecast in minutes, and upgrade the moment you want them live.
Get Started Free →