Dividend Capture
See how much of a dividend the stock hands straight back on the ex-date, what is left after the drop, and how long past positions took to climb back to breakeven. Every ex-dividend event is measured against the last cum-dividend close, and the same endpoint carries the forward calendar of dividends still to come.
Measure the give-back.
drop_ratio_close divides the ex-day price drop by cash_amount, so you can see how much of the dividend the market took back the same day.
See what is actually left.
net_capture_pct is the percentage kept after the drop, and breakeven_price is the level the position has to reclaim.
Time the exit with history.
days_to_recover_breakeven and the recovered_within_1d through recovered_within_20d flags show how long past events took to get back.
Is buying a stock just before its dividend and selling after it actually worth doing? Pick a few dividends coming up in the next two weeks and backtest how that trade has worked in those names.
Your agent translates this into the forward ex-dividend calendar, each name's historical ex-day drop and net capture, and the recovery odds behind them.
The dividend, the drop, and what is left over.
On the ex-dividend date, a stock starts trading without the right to the upcoming dividend, and it usually opens lower by roughly the dividend amount. Dividend capture is the trade that buys before that date and sells after, hoping to keep the cash while the price comes back. This dataset measures both halves of that trade for every ex-dividend event: how far the price actually fell relative to the dividend, and how long it took to recover.
Each row is one ex-dividend event for one ticker. Events that have already happened carry the measured prices and the recovery outcome. Events still ahead carry the declared schedule with the price fields empty, which is what lets the same endpoint work as a forward calendar.
Each row includes:
- The dividend itself: cash amount per share, frequency, type, and the declaration, record, and pay dates.
- The prices that bracket the ex-date: the cum-dividend close the day before, and the ex-date open and close.
- The drop measured two ways, to the ex-date open and to the ex-date close, each divided by the dividend.
- What is left over: net capture percent, capture yield, and the breakeven price the position must reclaim.
- Recovery: trading days taken to reclaim breakeven, the date it happened, and flags at 1, 3, 5, 10, and 20 days.
The dividend is not free money. On average the ex-day drop takes back 0.87 of every dollar paid out the same session, so a capture trade is a wager on the leftover, not a yield you simply collect.
Find out whether the capture is worth doing at all.
The published study on this dataset covered 82,012 resolved ex-dividend events since 2024. Net capture averaged +0.04% and 45% of events finished underwater, with the ex-day give-back eating the difference. Read plainly, the generic capture trade is close to zero expected value before costs, and knowing that with a number attached is most of what the dataset is for.
The distribution is where the remaining work is. Recovery odds run from 55% within one trading day to 91% within twenty, so the same event is a very different trade depending on how long you are willing to hold it. The forward calendar lets you line upcoming ex-dates up against each name's own history of drop ratios and recovery times before any capital is committed.
- Pull the forward schedule with upcoming=true and future_days, then join each ticker's resolved history to it.
- Rank candidates on drop_ratio_close rather than capture_yield_pct, since the yield is the part you hand back.
- Split resolved events with recovery_status to compare the ones that recovered against the ones that did not.
- Compare drop_ratio_open with drop_ratio_close to see whether the give-back lands at the open or over the session.
- Set holding periods from the recovered_within_5d and recovered_within_20d rates instead of a single average.
Measured from the close before to the sessions after.
The schedule comes from the declared corporate action: declaration_date, record_date, pay_date, cash_amount, frequency, and dividend_type as announced. The price side is measured from daily bars. cum_close is the close on cum_date, the last session that still carries the dividend, and ex_open and ex_close are the ex-date prices. The drop ratios divide that fall by the dividend, so 1.0 means the market took back exactly what was paid out.
After the ex-date each row is tracked forward for recovery_window_days trading days, with bars_observed recording how many sessions have actually been seen so far. recovery_status stays pending until the position reclaims breakeven or the window closes. Rows refresh daily after the close and last_updated dates the refresh. Nothing is published beyond a 120-day forward horizon, and future_days is clamped to that ceiling.
The fields that decide whether the trade nets anything.
The headline fields below are a subset. Every field, with exact types and semantics, is documented in the API reference.
| Field | Type | What it tells you |
|---|---|---|
| date | string | Ex-dividend date, the session the stock starts trading without the dividend |
| ticker | string | Equity ticker symbol |
| cash_amount | float | Cash dividend per share for this event |
| capture_yield_pct | float | The dividend as a percentage of price, before any drop is taken into account |
| cum_close | float | Close on cum_date, the last session that still carries the dividend |
| drop_ratio_close | float | Ex-day drop measured to the close, divided by the dividend; 1.0 means the whole dividend came out of the price |
| drop_ratio_open | float | The same ratio measured to the ex-date open instead of the close |
| net_capture_pct | float | Percentage of the position kept once the ex-day drop is taken off |
| breakeven_price | float | Price the position has to reclaim to break even on the capture |
| recovery_status | string | Whether the event recovered, has not, or is still pending inside its window |
| days_to_recover_breakeven | integer | Trading days taken to reclaim breakeven; recovery_date names the day it happened |
| recovered_within_5d | boolean | True when breakeven came back inside five trading days; matching flags exist at 1, 3, 10, and 20 |
One call returns the calendar and the history behind it.
One key works across the REST API, the hosted MCP server, and this dashboard. Every response is JSON with a { count, data } envelope.
import requests
url = "https://api.alphanume.com/v1/dividend-capture"
params = {
"upcoming": "true",
"future_days": 14,
"api_key": "alp_abc123"
}
r = requests.get(url, params=params)
print(r.json())curl "https://api.alphanume.com/v1/dividend-capture?upcoming=true&future_days=14&api_key=alp_abc123"{
"count": 1,
"data": [
{
"date": "2026-05-08",
"ticker": "KO",
"declaration_date": "2026-04-16",
"record_date": "2026-05-11",
"pay_date": "2026-07-01",
"cash_amount": 0.51,
"frequency": 4,
"dividend_type": "regular",
"annual_dividend": 2.04,
"capture_yield_pct": 0.72,
"cum_date": "2026-05-07",
"cum_close": 70.84,
"ex_open": 70.41,
"ex_close": 70.55,
"price_drop_close": 0.29,
"price_drop_open": 0.43,
"drop_ratio_close": 0.57,
"drop_ratio_open": 0.84,
"net_capture_pct": 0.31,
"breakeven_price": 70.33,
"recovery_status": "recovered",
"days_to_recover_breakeven": 2,
"days_to_recover_price": 4,
"recovery_date": "2026-05-12",
"recovered_within_1d": false,
"recovered_within_3d": true,
"recovered_within_5d": true,
"recovered_within_10d": true,
"recovered_within_20d": true,
"recovery_window_days": 20,
"bars_observed": 20,
"last_updated": "2026-06-08"
}
]
}What this data does not claim.
- Forward calendar rows carry the schedule only. Ex-day prices, drop ratios, and recovery fields stay empty until the ex-date has passed and price bars are observed.
- Drop ratios are raw price differences, not market-adjusted ones. A broad selloff on an ex-date shows up as a larger drop than the dividend caused, and a rally hides part of it.
- The published figures are gross of commissions, spreads, and borrow, and they say nothing about taxes, which fall differently on dividend income than on price gains.
- recovery_status is pending until breakeven is reclaimed or the window closes, so recent events are not a settled sample. recovery_window_days and bars_observed tell you how far along each one is.
- Pro includes full historical access and the forward calendar via upcoming=true. Forward queries use the upcoming ex-dividend schedule rather than a historical observation window.
Asked by researchers, answered plainly.
Why does a stock drop on the ex-dividend date?
From the ex-date onward a buyer no longer receives the upcoming dividend, so the shares are worth that much less cash to whoever holds them, and open orders are adjusted accordingly. The average drop measured here is 0.87 times the dividend rather than 1.0, which is the usual empirical finding: ordinary trading, tax treatment, and rounding all pull the ratio away from a clean one-for-one.
What does a drop ratio above 1 mean?
It means the stock fell by more than the dividend on the ex-date. The ex-date is a normal trading day, so the ordinary move often swamps the dividend adjustment, especially for a small dividend on a volatile name. A single ratio above 1 is usually the market moving, not the dividend. A ticker whose ratios sit above 1 across many events is the more interesting case.
How far ahead does the upcoming ex-dividend calendar go?
future_days sets the horizon and defaults to 7 days, with a hard cap of 120. The forward edge of a date-range query is also clamped to today plus future_days, so an open-ended range will not hand back months of dividends nobody has declared details for yet. Past-only queries are unaffected by the clamp.
What is the difference between recovering to breakeven and recovering to the pre-dividend price?
Breakeven counts the dividend as money already in hand, so the stock only has to climb back to breakeven_price, which sits below the cum-dividend close. days_to_recover_price is the stricter test: the number of trading days until the stock is back at the cum-dividend price itself, dividend excluded. The first measures the trade, the second measures the stock.
Available with Pro. Enterprise for teams.
New REST API and MCP access requires Pro or Enterprise. Public dashboard previews are separate from a subscription. Pro includes the available historical record, current updates, REST, MCP, dashboard exports, and every standard dataset as it launches. Explore the dashboard preview before subscribing; its existing limits still apply.