Stocks vs Index Funds: Why Your Good Company Falls Too

Stocks vs index funds: one difference weighs more than every fee table put together. A basket buys your company because it meets a rule; you buy it because you have a view. The practical consequence is that part of its price is set by demand for the whole basket rather than by any judgement of the business. That is why a good company can fall alongside weak ones in a week when it said nothing about itself at all.

A basket has a rule, you have a thesis#

An index fund does not rate companies. It checks a condition: does this one sit in the sector, is it large enough, does it trade on this market. If the condition holds, it buys, and it never opens the annual report, because that is not what it was built to do and nobody pays it for that.

The rule can be as mechanical as you like. It can read "the five hundred largest US companies", or "semiconductor manufacturers", or "companies domiciled in emerging markets". Not one of those sentences contains a word about the quality of a business, and none of them is meant to.

A fishing net does not choose fish, only the size of its mesh, and then it hauls up everything that could not slip through.

Say a sector basket holds 30 companies: two excellent, twenty-odd ordinary, and a handful you would put straight back down after reading the annual report. The fund owns all thirty, in proportion to their market values. When one of them rises, its weight in the index rises, so the next time money comes in the fund buys more of that one than of the others. Not because the rise was earned. Because it happened.

The mechanism reinforces itself, and most people only ever watch the pleasant half of it. The other half works identically in reverse: when unit holders start redeeming, the fund has to hand over cash, so it sells all thirty companies at once, in the same proportions, the excellent ones alongside the weak. One order goes out, and it touches thirty businesses nobody compared that morning.

Which is why a price moves between one set of results and the next for reasons that appear in no release.

How much of your price is basket demand#

A mechanism without a scale is a curiosity. So how does anyone know how much money moves this way, when nobody is obliged to admit to it?

Alex Chinco and Marco Sammon went around the problem with a method that suits the subject nicely. Instead of asking managers, they read the footprint the market leaves by itself: they asked how much capital would have to be tracking an index to explain the spike in volume on the day that index changes its members. Since every passive owner has to make the same trade at that moment, the size of the spike gives away how many of them there are.

Their answer was that in 2021 passive owners held 33.5% of the US stock market, while index funds themselves accounted for 16% (Journal of Financial Economics, 2024). Read that figure comparatively rather than on its own: it is twice the number usually quoted, because it adds institutions replicating an index inside their own portfolios and managers who are formally active while quietly tracking the benchmark.

Which means roughly one US share in three sits in a portfolio that has no opinion about it and never will. That portfolio will not sell when management misses a promise, and it will not add when the operating margin gains four points. It reacts to one thing only: whether somebody is creating or redeeming units of the fund.

I do not know what the share is today, because 2021 is the last year in that calculation. I do know that nobody has announced a retreat from indexing since.

The day everyone had to buy#

The cleanest view of the mechanism comes from a day when an index changes its members and every tracker has to make the same trade in the same minute. On 16 November 2020, S&P Dow Jones Indices announced that Tesla would join the S&P 500 before the open on 21 December (announcement). The funds tracking that index were not there to judge the decision. They were there to execute it.

Execution meant more than $80 billion of shares, all of it bought before the start of trading on that Monday, with traders expecting the heaviest volume on the Friday before (Associated Press). Howard Silverblatt of S&P Dow Jones Indices put the scale this way: "Historically, the $21 billion trading for fourth-quarter rebalancing is minor league, but when you add in heavy-hitter Tesla, $82 billion, you end up doubling the historical high, surpassing the $100 billion mark". Which means an ordinary fourth-quarter rebalancing runs to about $21 billion of trading, and one company added four times that on top.

The price at which those shares changed hands came out of an obligation, not out of a judgement.

Tesla is an extreme case and therefore a convenient one: nobody has to argue about whether it happened or how large it was. The same net goes out every quarter on a smaller scale, over companies nobody writes about. Every addition to an index has a deletion facing it, and a deletion means selling done by everyone at once, with no question asked about whether results had just improved. A company leaving an index meets sell orders from every tracker on the same day, though it did nothing to bring them on.

Why a good company falls with the bad ones#

Say you buy a company you have read from the annual report through to the earnings call transcript, and two weeks later it is 15% lower, having published nothing at all in the meantime. Anyone who holds individual companies knows that week. The first thought is always the same: other people know something I don't.

Usually they don't.

Usually money left the basket, the basket sold thirty companies in one order, and yours was thirteenth on the list. Bad news about one firm in the sector is enough, because a unit is redeemed whole rather than in pieces, so the money leaves all thirty at once.

The price that comes out carries information about the flow, and you read it as information about the business. That is the same error as mistaking a headline for an event, only harder to catch: a headline visibly had an author, while a price looks like hard data. Sifting a price like that from the events that genuinely move a company leaves zero to two things a month. Whether one company moved the portfolio or something wider than any of them is what Today's move answers in Taufolio.

This is where correlation without cause comes from. Companies in one sector move together even on days when nothing connects them except membership of the same basket, and the more money arrives through baskets, the more those moves look like one group decision rather than the sum of many. One specific, large fall calls for something else: an hour written out as an order of operations.

Stocks vs index funds: what you know#

Buy the basket and you know the selection rule and not a single business inside it, because nobody checked them on those terms. You know two of the thirty are excellent, and you know four are ones you would not want, and you do not know which are which.

Buy the company and the arrangement inverts: you know the business as well as you have read it, and that knowledge sits in a document with a date on it. About flows you know nothing, and flows are what will move the price for the next quarter.

Both routes leave you with partial knowledge. The difference is where the hole is, and which hole you can close with your own work. Which one can you close?

What this argument does not say#

In theory this mechanism makes a basket the worse idea. In practice it does nothing of the sort, and that is the most important caveat in this piece.

Flows run in both directions. The same mechanism that sold your company alongside twenty-nine others will buy it back when money returns to the basket, and it will not ask about margins then either. Over a long enough stretch those moves cancel, and what remains is how much the company earned and how much of it reached the owners.

Selection costs work that most people will not do, and that is an honest argument for the basket rather than an excuse. Anyone without the time to read filings is making a deliberate decision when they buy the basket: I give up the judgement and take the average of thirty. On the practical side both routes start the same way anyway, on the same account and with the same questions about horizon.

The largest risk here is not in the mechanism but in what you do with it. "It is only ETF flows" is a comfortable explanation for every fall, including the one with a reason sitting on page forty of the report. I think flows may be invoked only after you have checked the documents and found nothing in them. Used in the other order, this mechanism becomes a reason not to check.

A sector basket also carries a risk that selection does not remove: if a whole industry earns badly, the best company in it still earns badly, because the stage of the category is settled before the spreadsheet opens.

Price as information about flows#

Since part of a price comes from demand for a basket, price is a poor instrument for watching a company. Not because it lies. Because it mixes two signals and never says in what proportion it mixed them.

The data free of that contamination sits in the documents: revenue, operating margin, cash from operations, net debt, and what management said on the call. None of those numbers twitches because somebody redeemed units of a fund. Watching a company by its price produces reactions; Monitoring built on documents produces decisions. Sifted, the daily stream leaves zero to two things a month that genuinely move a business.

Three questions before you call it a signal#

When your company falls for no visible reason, work through these three questions in order before you do anything.

1. Did the other companies in the same sector and the same basket fall too? If they did, the move is about the basket, not about your company.
2. Did the company publish anything in those days? If not, there is no document in which the reason could be checked.
3. Which assumption in your thesis does this fall undermine? If none, the fall is a price, not information.

The third question only works if the thesis was written down beforehand and broken into conditions, because a thesis recalled from memory always turns out to fit the price. That is why Taufolio keeps the reason for owning a company apart from its valuation. The Investment thesis is produced automatically from a Full report and splits that reason into a handful of assumptions, each with a condition you can check in a document rather than on a chart. See what such a thesis looks like and where its conditions come from.