Dividend Harvesting: Rotating Ex-Dates for Income

Dividend harvesting usually means rotating a pool of money through dividend stocks so that you own each one over its ex-date, sell, and move to the next name on the calendar, collecting a payout most weeks. Some people use the same words for simply holding dividend payers and living on what they pay. The rotating version is the one that needs arguing about, and the case against it is fairly simple: each rotation is a full dividend capture, so a year of harvesting repeats the ex-date drop, two spreads and an unfavorable tax treatment many times over, and a basket of the same payers left alone usually keeps more.

What one rotation carries

Each hop in the rotation is the trade covered in dividend capture strategy. You buy before the ex-date. You sell after it. The share price tends to open lower by roughly the dividend, and you cross the spread both ways. Under T+1 each purchase has to go in no later than the session before that stock’s ex-date, so the dividend calendar sets your trading days for you.

Harvesting multiplies that. Run 20 rotations on a 50,000 pool and you have bought 1,000,000 of stock and sold about the same over the year (20 × 50,000 = 1,000,000). That is a great deal of buying and selling to collect payouts the market marks down on each ex-date. Measured by volume, the strategy produces far more turnover than income.

A year of rotating against a year of holding

Every input below is hypothetical. A 50,000 pool makes 20 rotations in a year. Each capture pays an average dividend of 0.8% of the position, each ex-date drop averages 0.75%, and the round-trip spread is 0.10%. The comparison is a basket of the same kind of payers yielding 3% and held all year. Tax rates are a hypothetical 24% on ordinary dividends and 15% on qualified ones.

Line	Rotation	Holding
Dividends collected	8,000	1,500
Ex-date drops realized	7,500	none
Spreads	1,000	25
Tax on dividends	1,920	225
Result	-2,420	1,250

The rotation collects more than five times the dividends and finishes behind. In this setup each capture loses 25 before tax, because the dividend clears the drop by 0.05% while the spread costs 0.10%, and the tax on 8,000 of ordinary dividends lands on top.

There is a partial offset. The rotation also books 8,500 of short-term losses from the drops and spreads (7,500 + 1,000). If you had short-term gains elsewhere for them to absorb at the same hypothetical 24%, they would be worth 8,500 × 0.24 = 2,040, and the rotation would finish at -2,420 + 2,040 = -380. Still behind.

The comparison flatters holding in one respect. The held basket goes ex-dividend too, and its price takes the same kind of drop on each of those dates. Those drops stay unrealized inside a position kept for its businesses, and over a year the basket’s price will move for reasons no example can set in advance, so the table leaves the holder’s price change out.

The tax and wash sale trail

Rotation almost never satisfies the qualified-dividend test. That test needs more than 60 days of ownership out of the 121 days that start 60 days ahead of the ex-date, per IRS Publication 550, and a harvester owns each name for days. Every dividend is ordinary. Every exit is a short-term gain or loss.

Wash sales are the quieter problem. A rotation calendar that sends you back into a stock within 30 days of selling it at a loss can trigger the rule, and so can buying something substantially identical to a name you just sold at a loss. Where similar ends and identical begins is not always clear. The disallowed loss moves into the basis of the new shares, which is manageable until the rotation is repeated so often that nobody can follow the chain.

Then there is the paperwork. Twenty round trips a year means forty trades, each a line on Form 1099-B, and every wash sale adjustment has to be traced back to the sale that caused it. How to read Form 1099-B shows where those adjustments sit. Brokers only flag repurchases of the identical security in the same account, so a harvester running the calendar across two accounts reconciles the rest by hand.

Testing your own version

The inputs above are invented, and yours will differ. The check is quick with real figures. For each rotation, write down the dividend, the drop at the open and the spread you paid, all as a percent of the position, then work down a short list:

• Dividend minus drop.
• Then minus the spread.
• Then minus the tax.

When the first line is barely positive, the second usually turns it negative, as it did in the example. A rotation that pays needs dividends that clear both the drop and the spread by more than the tax takes, over and over, and nothing about the calendar makes that more likely. Keep the log for a full year before judging. A few good mornings early on say little about the next forty trades.

The case for rotating, taken seriously

The strongest objection is about need. Someone living on portfolio income wants money arriving often, and a payout most weeks feels like a paycheck. That pull is real. A dividend lands as cash, while a holder who sells a slice of shares to raise the same amount feels like they are spending capital.

The answer is that the rotation is spending capital too. Each ex-date drop is mostly the dividend coming out of the share price, so a weekly payout from rotation is largely your own money returned in small pieces, with spread and ordinary tax taken on the way. A holder can get regular cash by owning payers whose payment dates fall in different months, or by selling a small amount on a schedule, and both keep the turnover low.

Steadiness has value, and some people will pay for it knowingly. The figures above show roughly what that steadiness costs under one set of assumptions.

Where the argument stops

In a tax-advantaged account the ordinary-income problem disappears, which is the angle taken in dividend capture in an IRA. The arithmetic changes, but only the tax line goes away. The example still ends at -500 before any tax, because the drops and spreads were enough to sink it.

Special dividends are another exception. The price reaction to a large one-off payment may not follow the usual ex-date pattern, and a trader who reads one of those situations well is making a call on a specific event.

Margin makes everything worse. Borrowing to run the rotation adds interest on every day the position is open, and the intraday margin requirements mean a shortfall after any purchase during the day can count as a deficit. On a trade that already nets close to zero, the interest alone can decide the year.