How Much of Your Portfolio Should Be in One Sector?

A broad market index fund is not sector-neutral. It is market-cap weighted, so it holds each sector in proportion to that sector's share of the market's value, and those proportions are large and uneven. Information technology has been the largest sector in the S&P 500 by a wide margin in recent years, with financials, health care, and consumer discretionary among the next tier, and utilities, materials, and real estate among the smallest. Those weights float with prices, so they are worth reading current rather than recalling.

This is the correct starting point for the sizing question. If you own a broad US index fund, you already have a substantial technology position, a meaningful financials position, and so on, without having made a single sector decision. The question is never “how much should I put in this sector” from zero; it is always “how much do I want in total, given what I already carry”.

Adding a sector fund on top of a broad index fund does not diversify into that sector, it concentrates further into it. The index fund already holds every company in the sector at index weight. The sector fund holds a subset of those same companies, usually the same megacaps at the top, at much heavier weights. Your true sector exposure is the sum, and the individual leaders appear twice, so the top handful of names in the sector can end up carrying an outsized share of the entire portfolio.

The arithmetic is straightforward and rarely done. For each fund you hold, multiply its share of the portfolio by its weight in the sector; add any single stocks in that sector held directly; sum the results. That total is the number that carries the risk. Sizing a new sector position against the new piece alone, rather than against the total after adding, is the most common way portfolios end up concentrated by accident. How to check your portfolio concentration walks the calculation, and how to find overlap in your ETFs covers the fund-level version.

The core-and-satellite framing is the most common way investors organize this, and it settles the sizing question more cleanly than any percentage does. The core is the permanent base: broad, diversified, held indefinitely, and not traded around a view. Satellites are deliberate tilts expressing a specific thesis, sized smaller, reviewed on a schedule, and understood to be the part of the portfolio that can be wrong without derailing everything.

A sector fund is almost always a satellite. Treating it as a core holding means accepting that the portfolio's long-run outcome depends on one industry, which is a different proposition from a diversified base with a tilt on top. The practical value of naming the role first is that it converts an unanswerable question (“what percentage in semiconductors”) into an answerable one (“how much of this portfolio am I willing to have riding on a single industry thesis”). The core-and-satellite structure covers the wider construction, and asset allocation models sets the level above it.

Concentration is not a label, it is a measurable change in the range of outcomes. Raising one sector's weight increases how much of the portfolio's movement is explained by that sector's fortunes, in both directions. A sector tilt that works pulls the whole portfolio up more than a diversified holding would; one that does not pulls it down the same way. Sector funds are also more volatile than the broad market on their own, because they hold fewer companies with correlated businesses.

The most useful sizing test is not statistical. It is to take the total post-addition weight and ask what happens to the whole portfolio if that sector falls by half, which sectors periodically do. If the resulting portfolio decline is one you would sit through without selling, the size is workable. If it is not, the size is too large regardless of how good the thesis looks, because a position you abandon at the bottom does not get to be right later.

Sector labels imply more separation than exists. Semiconductors sit inside the technology sector, so a chip fund and a broad tech fund are largely the same exposure at different intensities. Energy and materials both respond to commodity cycles. Banks, insurers, and real estate all move on interest rates. Industrials and transport share the same economic cycle. Two tilts into correlated sectors are much closer to one larger bet than to two independent ones.

The practical adjustment is to size correlated tilts as a group rather than individually. Three separate positions that each look modest can add up to a single dominant exposure, and this is a frequent way a portfolio that looks diversified by line count is not diversified by driver. How to analyze your portfolio's diversification is the deeper treatment of that gap.

Sector positions do not stay the size you set them. A tilt that performs well grows as a share of the portfolio automatically, so a modest position can become a dominant one purely through success, with no decision made anywhere along the way. This is how most accidental concentration happens, and it is invisible unless something forces a look.

A rebalancing band is that forcing function: a tolerance around the target weight that triggers a review when the position drifts outside it, checked on a set schedule rather than continuously. Wider bands mean fewer trades, lower costs, and more drift; narrower bands mean tighter control, more trading, and, in a taxable account, more realized gains. Neither is correct in the abstract. What the band does is convert silent drift into a conscious choice, which is the whole point. How to rebalance your portfolio with AI covers the mechanics of running that review.

The analysis above is the same everywhere, but the cost of acting on it is not. In a tax-advantaged account, trimming a sector position that has run costs nothing in tax, so bands can be enforced mechanically. In a taxable account, trimming a winner realizes a gain, which is a certain cost against an uncertain benefit, and that argues for wider bands, for directing new contributions toward underweight areas instead of selling, and for being more careful about the size in the first place.

That asymmetry is worth planning around before the position exists, because a large embedded gain removes options later. This is general information rather than tax advice, and the wider framing is in how ETFs are taxed.

Working these in order tends to produce a number, which is why the framework substitutes for the percentage nobody can supply for you. It also explains why two thoughtful investors land in very different places: they have different starting weights, different roles in mind, and different tolerance for the drawdown in the fourth row.

Someone holding only a broad index fund. The starting weight is the index weight, and that is already a real sector position. The first useful step is reading the current sector weights of the fund they own, not choosing a tilt.

Someone who works in the sector. Salary, equity compensation, and job security are already exposed to the same industry, so the investment position is not the whole exposure. Human capital is a real, correlated holding, and investors in this position commonly account for it when sizing.

Someone with a strong thesis. The satellite framing is the relevant one. The size question becomes how much of the portfolio can be wrong at once, and the rebalancing band becomes the mechanism that keeps a working thesis from silently becoming the entire portfolio.

Someone drawing on the portfolio soon. Sequence matters more than average return, because a large sector drawdown at the wrong moment is realized rather than waited out. That generally pushes sizing conversations toward the total post-addition weight and away from the upside case.

Someone who already holds several tilts. Correlation is the binding constraint. Adding a fourth position that shares drivers with the existing three changes the portfolio's concentration more than its line count suggests, and how to invest in sectors covers how the sector map fits together.

The hard part of this question is not the judgment, it is the arithmetic that has to happen before the judgment. Decomposing every fund you hold into sector weights, combining accounts, adding individual stocks on top, and then modeling what a new position would do to the total is genuinely tedious by hand, and it is the step almost everyone skips. Skipping it is why so many portfolios turn out to be concentrated in a sector the owner never consciously chose to overweight.

Walnut is the assistant for that work. With your brokerage connected read-only, you can ask through Claude, ChatGPT, or the built-in assistant what your true weight in a sector is across everything you hold, how much of it comes from funds you think of as diversified, and how a new position would change the total. Any order waits for your explicit approval. Walnut is not an investment adviser and does not tell you what your weights should be; it makes the current ones visible so the sizing decision is an informed one.

The percentage is the wrong unit to start with. Start with your real current weight, including the sector exposure hiding inside your broad index funds, because that is the number the decision is actually being made against. Then name the role: a core holding and a satellite tilt are sized by different rules and should not be confused.

From there, size against the total weight after adding rather than the new piece alone, treat correlated sectors as a single bet, and set a band so drift triggers a review instead of accumulating silently. What number that produces is specific to your portfolio, your timeline, and how much of a drawdown you can hold through, which is why this page describes the mechanics rather than prescribing a figure.