How Many Dividend Stocks Should You Own? 20 to 30, and Why

Twenty to thirty individual dividend stocks, spread across at least eight sectors and more than one country. If you would rather not follow that many companies, a broad dividend ETF core plus ten stocks you know well gives the same protection with less work.

The number is not arbitrary. It is where two different risks both fall to a level you can live with: the risk of losing capital because one company fails, and the risk of losing income because one company stops paying. The first is well studied and the answer is around twenty to thirty. The second is arithmetic, and it says one cut should never remove more than about 5 percent of your dividends.

Two risks, one number

A share price carries two kinds of risk. Market risk is what every stock shares: a recession, a rate rise, a crash. No number of holdings removes it. Single-stock risk belongs to one company: a fraud, a failed product, a bad acquisition. That part shrinks as you add holdings, and the well-known finding is that most of it is gone by the time you own twenty to thirty stocks in different businesses. Going from five to twenty removes most of it. Going from thirty to sixty removes very little more.

An income portfolio has a second risk that price studies do not measure. Your dividends are a list of separate decisions by separate boards, and any one of them can be suspended on a Tuesday morning. What happens to your income when one company stops paying has a cleaner answer than what happens to your capital, and it is the one that sets the floor.

What one cut costs, by portfolio size

Suppose every holding pays the same dividend. A single suspension then removes one share of your income, and the share depends only on the count.

A cut is rarely a full suspension, so the real loss is often half these figures. But cuts cluster: 2009 and 2020 both produced several in the same portfolio in the same year, which is why the third column matters. With five holdings, a bad year can take 40 percent of your income. With twenty it costs 10 percent, and with thirty, under 7 percent.

That is the case for the lower bound. Below fifteen or so, a single board decision is a pay cut you will notice. Above thirty, the extra names buy very little on this measure and cost time on every other.

The two income caps

The table turns into two rules you can check against any portfolio, whatever its size:

• No single holding above 5 percent of your dividend income. One suspension then costs you at most a twentieth of your income, which most budgets absorb. With twenty equal payers you are exactly at the cap; with thirty you have room for a few larger positions.
• No sector above 20 to 25 percent of your dividend income. Cuts come in sector waves, not one at a time. A cap at a fifth to a quarter means the worst wave in living memory (banks in 2009, energy and travel in 2020) would have cost you that much and no more.

Both caps are written in income, not in value. That distinction is where most concentration hides.

Weight by income, not by value

Your broker shows each position as a share of portfolio value. For an income investor that is the wrong denominator, because high yielders carry more of the income than their size suggests.

Take a 100,000 portfolio with an overall yield close to 3.3 percent. One holding is 5 percent of value, so 5,000, and it yields 6 percent: it pays 300 a year. The other 95,000 yields 3.2 percent and pays 3,040. Total income is 3,340, and the 6 percent yielder is 300 of it, which is 9 percent of your dividends from a position that looks like 5 percent of your portfolio. Cut that one dividend and you lose almost twice what the value weight implied.

The effect compounds at the top of the list. The three or four highest yielders in a typical income portfolio are often 12 to 15 percent of value and 25 to 30 percent of income, and by the logic of yield they are also the payers the market trusts least. Work out each holding's share of income and apply the 5 percent cap to that number.

Spread across sectors and countries

The reason for the sector cap is that dividends fail together. In 2008 and 2009 it was the banks: most large US, UK and eurozone lenders cut or suspended, and in 2020 the European Central Bank and the Bank of England asked theirs to suspend again. In 2020 the wave hit energy, with the largest European oil majors cutting for the first time in decades, and travel, where airlines, hotels and cruise lines stopped paying altogether. A portfolio built on "safe, high-yield banks and oil" in 2007 lost most of its income twice in twelve years. A portfolio with eight or more sectors lost a slice each time and kept the rest.

Eight of the usual eleven sectors is the practical minimum: consumer staples, healthcare, utilities, financials, industrials, energy, technology, communication, consumer discretionary, materials and real estate. The cyclical ones (energy, materials, banks) deserve the smaller shares.

Country spread matters more for a European reader, because home markets are narrower. UK dividend income is dominated by a handful of names in banking, oil, mining and tobacco; the Swiss market is three or four large payers and a long tail; a eurozone investor who buys only local blue chips ends up heavy in banks, insurers and carmakers. Spreading across the US, the UK, the eurozone and Switzerland also spreads the withholding you pay at source (15 percent on US dividends with a treaty, 25 percent in France, 35 percent in Switzerland), but it is the income spread that protects you.

The cost of too many

Diversification has a ceiling as well as a floor, set by attention rather than arithmetic.

Sixty holdings you never review is a common end state. Each was bought for a reason that has not been re-checked since. Dividend cuts land on the names nobody looks at, and a portfolio that large has forty of them.

Overlap is the second cost, and it is mostly invisible. A broad dividend ETF, a high-dividend ETF and a dozen stocks bought from the same screen often share their top names: the two funds hold the same twenty companies at slightly different weights, and your picks are ten of those twenty. Sixty lines on the statement, fifteen companies doing most of the work, and a concentration your broker does not show because it counts lines, not companies. Look through the ETFs before you count.

The third cost is time. Twenty companies at four results a year is eighty sets of results, a weekend a quarter. Sixty is a part-time job, and for most people it will not get done.

Three shapes that work

Which fits depends on the size of the portfolio and the time you want to give it.

• ETF core plus ten stocks. Seventy to ninety percent in one or two broad dividend ETFs, then ten companies you understand and want more of. The ETF supplies the sector and country spread and takes the look-through count well past thirty; the ten stocks are where you learn. Right for a beginner, or for anyone under about 50,000, where ten positions of 3,000 to 5,000 are the smallest that make sense after dealing costs.
• Twenty to twenty-five stocks across sectors. No ETF, eight to ten sectors, three or four countries, no holding above 5 percent of income and no sector above a quarter. The classic income portfolio, for someone with 100,000 to a few hundred thousand who enjoys the work and will do a quarterly review.
• Thirty or more for a large portfolio. Above several hundred thousand, a 3 percent slice of income is still a real sum, and thirty to forty names let you hold the cyclical sectors at small weights without leaving them out. Past forty, ask whether the extra names are diversification or accumulation. An ETF sleeve for the sectors you know least beats another ten stocks.

In all three, the count is a means; the two income caps are the test.

Where a portfolio tracker fits

A portfolio that met every cap on the day you built it drifts: winners grow into oversized positions, raises and cuts shift the income weights. The hard part is seeing that when the positions sit at three brokers that show weights by value only, and the ETF's holdings are on a factsheet somewhere else. A portfolio tracker such as Cadances pulls every account into one portfolio and lets you set the caps as allocation rules (no single holding over 5 percent, no sector over 25 percent, at least 20 positions) that read ok, near or breached as the portfolio moves. It splits the income you received by sector and by country, lists your top payers with their share of the total, and shows the top holdings inside a US-listed ETF so the overlap is visible. The rule stays yours; the tracker is what checks it while you are doing something else.

Questions people ask

How many stocks should a dividend portfolio have?

Twenty to thirty individual stocks across at least eight sectors, or ten stocks on top of a broad dividend ETF core. Below fifteen, one suspension is a noticeable pay cut. Above forty, the extra names add little protection and a lot of review.

What is the ideal number of dividend stocks for a small portfolio?

For under about 50,000, an ETF core plus five to ten individual stocks. Dealing costs and minimum position sizes make twenty-five separate holdings inefficient at that size, and the ETF supplies the spread the stocks cannot yet.

How do I know if my dividend portfolio is too concentrated?

Work out each holding's share of your dividend income, not its share of value. If any one is above 5 percent, or any sector above 20 to 25 percent, you are concentrated even if the count looks fine. High yielders carry far more of the income than their value weight suggests.

How should a dividend portfolio be split by sector?

Across at least eight of the eleven sectors, with none above a quarter of income and the cyclical ones (energy, materials, banks) at the smaller weights. Consumer staples, healthcare and utilities can carry more because their dividends held in 2009 and 2020 while banks, oil and travel cut.

Does a dividend ETF count as diversification?

Yes, on a look-through basis: a broad dividend ETF holds far more than thirty companies across most sectors. The catch is overlap. If your individual stocks are the ETF's own top holdings, you own them twice, and the true concentration is higher than the line count suggests.