40

I don't really want to think too much about investments -- just park my surplus money somewhere and be able to come back several years later and see that it has grown. (As this question shows, I'm none too good at following through on this intention, but let's ignore that).

A lot of advice on the internet seems to claim that low-cost index-tracking mutual funds are what I'm looking for. And it does sound like a nice idea, at first. But then I got to thinking: what if everyone did that?

Indexes are reconstituted from time time, meaning that index-tracking fund managers will suddenly need to dump a lot of the stocks that dropped out of the index into the market, and conversely, stocks that were just added to the index will suddenly be in high demand, and therefore expensive.

In fact it seems like a savvy trader (which is to say, not me) could make a return by going short on stocks that are expected to drop out of a popular index soon, and long on stocks that are expected to be added to it. If the organization that defines the index are reasonably open about their criteria, this should be doable without much risk. And since I can imagine doing this, surely there must already be people doing it.

And it seems that the only place those people's profit could possibly come from, is the people who invest in index-tracking funds.

How can tracking indexes then still be a good investment? Is it that they are in fact (despite their prominence in internet-based free advice) such a niche mode of investment that reconstitution events don't influence prices appreciably?

Rodrigo de Azevedo
  • 1
  • 8
  • 22
hmakholm left over Monica
  • 3,370
  • 1
  • 22
  • 24
  • 2
    Related: https://money.stackexchange.com/questions/85706/at-what-point-do-index-funds-become-unreliable – henning Mar 26 '18 at 11:49
  • 2
    Re "...what if everyone did that?", everyone doesn't :-) – jamesqf Mar 26 '18 at 19:26
  • @jamesqf: If you want, you could say that I'm attempting to apply Kantian ethics to investing. A strategy where the only thing that saves its followers from being fleeced is that it has sufficiently few followers, would strike me as philosophically unsound, a kind of freeloading. – hmakholm left over Monica Mar 26 '18 at 19:59
  • 2
    One thing that I didn't notice when skimming through the answers is the fact that index funds do not make these trades all at once. They move blocks in and out across a period of time at irregular intervals. This is intentional and meant to not flood the market with large numbers of orders one way or another. Here's an (old) article which discusses the exact effect you are asking about. – JimmyJames Mar 26 '18 at 21:36
  • @JimmyJames: Which is to say that they don't actually track the composition of the index from day to day but behave more like the "pseudo-index fund" I was imagining in my comment to Harper's answer? If so, that would be an answer to my question. – hmakholm left over Monica Mar 26 '18 at 22:32
  • Downvoted for not asking about an actual personal-finance problem which needs to be solved. – Beanluc Mar 26 '18 at 22:58
  • @Henning Makholm: OTOH, we could apply that same "ethical" test to any behavior. For instance, when the market nosedives and a lot of people start panic-selling, does that mean that we should follow them and sell too? Or should we look at it as a future buying opportunity? – jamesqf Mar 27 '18 at 04:10
  • @Beanluc Oh, is this an actual requirement? I find the question absolutely legit. – glglgl Mar 27 '18 at 12:01
  • A stock's price tends to rise in anticipation of being added to an index and fall in anticipation of being dropped from the index, so the effect is already there. Although one can imagine that trackers might magnify these effects – Mawg says reinstate Monica Mar 27 '18 at 12:27
  • @HenningMakholm Right, if you plot the S&P index against different ETFs or funds, you'll see they don't match exactly. This is one of a handful of reasons for that. – JimmyJames Mar 27 '18 at 14:56
  • If I rephrased the question to "What if everyone put their money into companies that industry analysts have decided are solid, benchmark companies?" does that help clarify things? – corsiKa Mar 27 '18 at 15:14
  • The question seems to have a fundamental misunderstanding in it. You say "it seems that the only place those people's profit could possibly come from, is the people who invest in index-tracking funds." I am not seeing how that follows at all. Profit arises from selling an asset for a price higher than the purchase price; it does not come from some completely other investor that is holding a similar asset. Can you explain your theory here, because I suspect this theory is the crux of your confusion. – Eric Lippert Mar 27 '18 at 15:24
  • Your question is also incoherent because it begins with the premise "what if everyone did that?" and then describes a scenario in which not everyone does that. This is a very confusing question. Can you clarify it? Is the question about what the world would be like in your counter-to-fact world where everyone invests in indexes? Or is it about whether index fund managers move the market? Or what? – Eric Lippert Mar 27 '18 at 15:25
  • @EricLippert: The profit would arise from buying a soon-to-enter stock at the ordinary market price in the weeks leading up to the index reconstitution, and then selling it to the index fund on reconstitution day where the price is unsually high due to the index funds' sudden and predictable need to acquire that stock. Similarly profit would be made by a fundamentals-oriented investor thinking, "hmm, this stock looks like a good deal now, but next Wendesday I'll be able to buy it even cheaper as tracker surplus" and then instead trading when the price has (momentarily) tanked. – hmakholm left over Monica Mar 27 '18 at 17:41
  • Thus, the savvy investor is not, as you describe it, a "completly other investor that is holding a similar asset" -- he is the counterparty that the index fund sells to and/or buys from. – hmakholm left over Monica Mar 27 '18 at 17:43
  • Now, @JimmyJames above has proposed an explanation for this that sounds convincing to me; unfortunately it is not posted as an answer. – hmakholm left over Monica Mar 27 '18 at 17:44
  • @corsiKa: No, that would make it a different question. This is specifically about index tracking funds, and why the mechanics of adapting their holdings to changes in the index on a well-known schedule is not being exploited by the market. – hmakholm left over Monica Mar 27 '18 at 17:49
  • @HenningMakholm The practice I describe is not, by itself the answer to your question. It's necessary due to the scale of these indexes. If all of them tried to sell all of a particular stock on the same day, it's highly unlikely there would be enough buyers (at a fair price) for the stock and you'd have a liquidity crisis for that stock. – JimmyJames Mar 27 '18 at 20:35
  • @JimmyJames: Drat. I thought you understood what my question is, since what you write looks exactly like an answer to it. But I'm running out of new ways to explain what I'm asking. – hmakholm left over Monica Mar 27 '18 at 22:22
  • Leaving aside the "what if everyone did this?" part of the question, I'm still not understanding your scenario. Suppose I announce "tomorrow I'm going to sell a large quantity of ABC and buy a large quantity of XYZ at market price". Your theory is that you can somehow make a profit on that information, but everyone has that information, so it will rapidly become priced in. If your scenario is that I don't make that announcement and you are making a guess about what I'll do tomorrow, then how is that any different from making any other guess about what ways prices will go? – Eric Lippert Mar 28 '18 at 18:55
  • As for the "what if everyone did this?" part, I don't understand that either. Suppose I have a strategy: sell everything I own but XYZ and buy XYZ at 10AM next Tuesday. But what if everyone did that? In that world, no one is buying anything but XYZ and no one is selling that, so worldwide commerce would come to a halt. The question "what if everyone did that?" is the same as every other version of "what if something impossible happened?" If something impossible happened then the world would be weird, but fortunately, it's impossible. – Eric Lippert Mar 28 '18 at 19:00
  • @EricLippert: You're getting entirely too hung up on "everyone". That's just a figure of speech, not a part of the question I'm asking. It suffices that a sizeable faction (say, about 10%) of the money invested in a particular market follows a strategy of following its leading index slavishly. Then, on the day when stock A drops out of the index and stock B instead enters it, around 10% of the total circulating amount of A stock will suddenly be for sale, and conversely the index-followers will suddenly all try to buy up around 10% of the circulating amount of B. ... – hmakholm left over Monica Mar 28 '18 at 19:05
  • ... The result of this ought to be (according to my perhaps flawed understanding of how markets work) that the price of A tanks, the price of B skyrockets, and those who follow the index are bound to lose big time. [Since investing is a zero-sum game, at least in the short term, someone else is going win, but that's not what my focus is. I appear to have misstated the question such that everyone thinks I'm talking primarily about who exactly wins. What I'm asking is why doesn't the market collectively fleece the index-followers, no matter where precisely the profit ends up eventually.] – hmakholm left over Monica Mar 28 '18 at 19:09
  • 1
    Where knowledge comes into the picture is exactly that the market does take the expected spike in demand for B into account; everyone has that information -- except that the index followers can't join the rest of the market in acting on that information, because the rule they're operating under says that the only information they're supposed to use is "which stocks make up the index today, and in which proportions". – hmakholm left over Monica Mar 28 '18 at 19:18
  • I once read a pop-finance magazine asserting a short term strategy of buying shares just booted out of big indexes: the theory is the reverse of the one in the OP, these stocks would have an artificial depression in price in a short period as institutional holders sold off; and would after some time return to their equilibrium P/E ratio. I tried it on a couple of positions with my gambling money, and it anecdotally worked pretty good until one stock that went bankrupt while I was holding it. – user662852 Mar 30 '18 at 16:43

7 Answers7

20

There are several things to keep in mind:

  • First, as Dilip's answer says, there are many indices, with different aims, so the number of "hot" stocks that someone would need to potentially target is much larger than your question suggests.

  • Even if you can whittle this down to a small number of (potential) "core" targets, however certain an "anticipatory trader" might be about the future direction of those stocks (i.e. which to short, which to go long with), the vagaries of the market mean that sometimes they'll get it right, often they'll get it wrong. (Remember, it's much easier after the event to "explain" why a stock rose or fell; accurately predicting this ahead of the event is much more error prone).

  • Even if someone were exceptionally good at picking the right stocks to target, the profit they would make is not directly (and certainly not exclusively) coming at the expense of the index funds. By successfully going long or short, the hypothetical trader is making money from correctly predicting the trend that the market is going to take: their profit comes mostly from those who think the reverse.

  • The action of an isolated "lucky guesser" is not going to materially affect the price of a stock. The extra expense (to an index fund) of stocks in demand, or the potential loss of stocks that have fallen out of favour is down to what the market as a whole thinks will happen.

  • The purpose of an index fund is not necessarily to provide the "best" investment (theoretically) possible, but to provide a "good enough" investment, at very low cost (because it's mostly automated).

Having said that, index funds aren't perfect: because they have to track "the whole market" (or a defined sub-section of it) they will track both its rises and its falls. They can get drawn into "bubbles" (be it "dotcom" or crypto-currencies) and have highly volatile gains/losses, when a more "savvy" trader might hold back from the initial hysteria until things have calmed down.

TripeHound
  • 9,173
  • 1
  • 26
  • 45
  • "Their profit money comes mostly from those who thinks the reverse" -- why wouldn't the profit money come from the index fund managers who are not supposed to think either way, but whom everybody knows will on such-and-such day need to dump such-and-such stocks and buy up such-and-such other stocks instead. There's no "isolated lucky guesses" involved in that, simply applying the (presumably openly available) formula that the sponsoring organization of the index uses to determine what its composition will be after its next periodic reformulation. – hmakholm left over Monica Mar 26 '18 at 18:06
  • @HenningMakholm Who is coordinating all of those people and forcing them to behave identically at identical times? Like this answer says, there are many indices with different aims. – Matthew Read Mar 26 '18 at 20:08
  • 1
    @MatthewRead: The people who publish indexes -- or perhaps the editors of financial newspapers who decide which indexes are important enough to be quoted in their pages, and warrant doom-class headlines when such-and-such index drops by X%. Just because Joe's Super-Spiffy Index of Nice Stocks With K In Their Name exists doesn't necessarily mean anyone cares. But there are a few ones that people do, for better or worse, care about. – hmakholm left over Monica Mar 26 '18 at 20:10
7

There are many different indexes (indices?) and many different mutual funds that track various indexes. Some indexes (such as Total Market Index) are untrackable in practice because there are too many different stocks and a fund cannot invest in all those stocks in the proper proportions for several reasons including that while a mutual fund will sell fractional shares of itself, its investments necessarily must be in integer numbers of shares in each company, perhaps even in round lots of multiples of 100 shares. A typical Total Market Index fund will run out of capital if it attempts to replicate the index; most such funds select a subset of all stocks such that the performance of the selected stocks will be a close approximation to what they believe the index will do.

So, too many different indexes for investors to choose from, many different index funds tracking each index with varying degrees of accuracy, and so even if everybody invests only in index funds, there is still enough flexibility for many companies to flourish even in they are not part of the S&P 500 or the Extended Market Index or the MidCap Index or the SmallCap Index or the Russell 2000 or ....

Dilip Sarwate
  • 31,562
  • 4
  • 55
  • 122
6

This is exactly the scenario which the market is exceptionally good at dealing with, by smoothing things out to an acceptable level.

The effect you describe exists. But being a market there are 2 effects:

  • If there was excessive profit to be made with the strategy, everyone would do it.
  • If everyone would follow that strategy, there would be no profit.

In this case the strategy is to sell before the stock is removed from the index, and buy back afterwards - if everyone did that, the price would collapse before being removed from the index, and skyrocket after. If this is a known and excessive pattern, other traders would exploit it by buying after the collapse but before stocks are being removed from the index.

In other words: You are absolutely right that there is an exploit, but the % effect on the individual stock price has to be small, otherwise the exploiters will get exploited themselves.

On top of that, index compositions are relatively stable, and the value of a single stock is commonly less than 5% of total index value, thus the predicted "losses" incurred by the individual investors are (negligibly) small compared to the predicted profits of the index fund.

If we assume a 1% loss on a stock that's 5% of the Index once per year, that's a loss of 0.05%, compared to 5+% expected growth in the same period. Some Indices change their composition more frequently than that, but with those the individual stocks make up a far lower percentage of the Index' value.

Peter
  • 2,113
  • 1
  • 10
  • 11
  • Just because the money lost by the index fund is shared out between many players who all speculate against the fund's predictable trading behavior, doesn't mean that the loss the fund itself takes is limited? – hmakholm left over Monica Mar 26 '18 at 17:50
  • @HenningMakholm Edited to avoid earlier misunderstanding. – Peter Mar 26 '18 at 18:17
5

Indexes are reconstituted from time time, meaning that index-tracking fund managers will suddenly need to dump a lot of the stocks that dropped out of the index into the market, and conversely, stocks that were just added to the index will suddenly be in high demand, and therefore expensive.

What stocks go in or come out although known is post facto. i.e. This is after a stock has done exceedingly well or a stock has not done well. To get a simplistic view, all other things being equal; the index is recomposed when;

  • A stock in Index loses its value over a period of time. At a designated date [generally quarterly], this will go out and next in line will get added.
  • A stock not in Index, gains heavily and is more that bottom most stock of Index. This will go into Index at appropriate place [can in theory become no 1 as well] and the last stock of index gives way / goes out.
  • There is a change in free float of a stock, can be both ways, a stock in Index, quite a few shares are purchased by promoter; reducing the free float or some stock not in Index has generated more free float stock [either by FPO or promoter off loading]

These events by themselves are unpredictable and take a while to materialize. Hence bets by fund manager may not always be accurate.

Although I don't have firm statistics for US markets, the total money invested by Index Funds would be less; around 10-20%. If we see the players in the market; there are

In the current situation Index funds buying or selling stock due to index movement does influence the price, albeit to a very small percentage. There is arbitrage opportunity, that is slightly unpredictable. The price may have already gone up in anticipation to getting added to index and may actually go down once it is added.

The price discovery is still largely by proprietary trades and to an extent by Institutional investors.

Brythan
  • 20,968
  • 6
  • 53
  • 67
Dheer
  • 57,070
  • 18
  • 88
  • 169
2

Lots of people think they can beat the market for lots of different reasons. You think you've thought of a reason. Here's how your hypothesis can go from spotting a market inefficiency to creating an investment strategy and your own active fund, to the result that a low cost index fund is still the most advantageous long term investment for most people.

Most importantly, I think you'd (or some savvy market participant, feel free to substitute all instances of "you" throughout this answer for "savvy market participant") have to do some backtesting to test your hypothesis to determine what if any effect there is on a company's stock as a result of being added to or dropping out of an index. I know there are answers related to the efficiency of the market and if an opportunity existed some number of people are exploiting it to the point that you can't see it. I doubt there is an inherent measurable effect here. Just because a market effect seems logical doesn't mean the phenomenon presents itself predictably in the market.

If you can predict what an index fund has to do can you profit from that movement? Maybe. If a company is falling out of a large-cap fund to a mid-cap fund because of bad news maybe by the time the funds have been forced to transact the company's stock has already been reduced to the point that other traders are interested in buying? There are literally incalculable numbers of forces and opinions being applied to securities transactions, I'm not sure any are anymore reliably predictable than any others.

Say you back test your hypothesis that you can regularly profit from constituent movements in and out of various index funds. You roll up a fund, and hire people, pay for your space, buy your access to the markets and have your own transaction fees. For all this work you charge a measly 1% to all your investors. Now you have yourself an actively managed fund, of which there are thousands. Your fund, and all other actively managed funds with similarly great ideas to profit in the market, and all other active market participants set the prices of securities, there was bad news, sell; you think the news wasn't that bad and the company is a buying opportunity, buy. This company has these new opportunities, buy; those new opportunities are overvalued, sell. And on and on. In a given year all of this active trading work roughly translates to the average result of the market. Some years a lot of these active funds will outperform the market, some years those same funds won't. The law of averages and regression to the mean, sort of dictate that on a long time line your Big Bad Active Fund will do about as well as the market did. But you're charging 1% for access to Big Bad Active Fund, and I can just go get VOO which will return the market average and only cost 0.03%. So even if your fund can manage to return the market average I'll lose 0.97% of my account every year. For me to stay flat, your fund has to beat the market return by at least 0.97% every single year. Sure, you'll have years where you earn double and even triple the market return, but regression to the mean shows that your extraordinary gains will come back to the average over time.

Jack Bogle, Warren Buffet, and a number of investing gurus (for lack of a better word) advise most people to simply allocate their stock market risk to a low cost broad market index and go about their life. Don't bother paying the active managers for their research and offices, just buy the market average return for 97% less cost to you. They say it's because the active managers can't beat the market, really it's simply that they have to charge too much more than the big broad market index funds that don't need research and analysis teams.

But we will never run out of individual investors because someone needs to control the boards of directors and we will never run out of active funds because there will always be optimists who think they can beat the market.

quid
  • 48,976
  • 11
  • 99
  • 161
  • 1
    You seem to be under the apprehension that I am proposing to make money for myself from the effect. I wonder how I can have written anything to provoke such a fundamental misunderstanding of what my question is. – hmakholm left over Monica Mar 26 '18 at 18:53
  • You said "How can tracking indexes then still be a good investment?" Maybe try reading my whole answer not just the first sentence. – quid Mar 26 '18 at 18:57
  • 2
    x @quid: Your whole answer does not contain an answer to my question. It just tries to beat me over the head with what I should do (not become a fund manager -- I get that, I have no intention of ever becoming one, and I'm very puzzled why you think from my question I have such an intention), without actually explaining why my reasoning about what I would expect to happen in the market doesn't work. – hmakholm left over Monica Mar 26 '18 at 18:58
  • We will never get to a place where everyone invested in index funds, because 1, people actually own these companies and stock generally comes with voting rights, 2, other people think they can beat the market and some of those people run investing funds. Index investing isn't niche free internet advice and the answer explains why. Again reread the answer (or don't). And good luck in life! – quid Mar 26 '18 at 19:06
2

They do

Yes, savvy funds managers do profit-take from index funds. But it also takes them a lot of work and research to do so. That's why they charge a 1.5% expense ratio on their index funds. That pays for the work, research staff and salary for the genius.

Whereas the index fund bears only about a 0.1% expense ratio (zero POINT one) because it is not picking stocks.

The difference being 1.4% -- that is, for the managed fund to win, it must beat the index fund by 1.4%, simply to pay its own expense ratio.

Even with the profit-taking it does from the relatively mindless index fund, it is still not able to beat it by 1.4%.

The managed fund is a better investment, if Santa Claus pays your expense ratio. If not, the index fund is a better bet.

This is a version of I don't have to outrun the bear. I only have to outrun you.

Harper - Reinstate Monica
  • 58,229
  • 10
  • 91
  • 195
  • Wouldn't it be (almost) just as cheap and mindless to run a pseudo-index fund that follows a composition made up by the same rules as actual indexes, except that it recomputes its stock selection on unpredictable (random and/or secret) days such that it would be less vulnerable to the savvies? – hmakholm left over Monica Mar 26 '18 at 20:02
  • 2
    John Bogle's thesis is that analysts can't outperform a random-number generator by enough to justify their expense ratio. I'd say the "manner of the random" is not critical, as long as it's on a large enough scale that the law of averages protects you from unlucky rolls. – Harper - Reinstate Monica Mar 26 '18 at 21:23
1

Index fund managers, despite managing index funds, are not dumb. Most index funds are used by very large, sophisticated institutions who track performance to the basis point (one-hundredth of one percent).

A calssic case is the "Russell Rebalance". The S&P indexes add and drop names as needed. Many of the Russell indexes have an annual rebalance.

There is a whole industry and process around that date, when index fund managers, hedge funds, and others try to play the game.

If an index fund manager is at all competent, they won't get caught on this. You need to remember they can use futures and swaps, as well (including on single stocks) to postpone or pre-trade the changes.

This is not worth worrying about as much as fees and taxes.

eSurfsnake
  • 217
  • 1
  • 3
  • 1
    How does it matter whether the manager is dumb or competent, if his job is simply to make the fund consist of exactly the stocks the index says it must consist of? Even if he would personally prefer a different strategy, how would he be allowed to do so (and still claim that the fund tracks that index)? Futures or swaps sounds like something that would belong in hedge funds or similar abandon-all-hope-ye-who-enter-here vehicles, rather than a supposedly simple and consumer-approved mutual fund. – hmakholm left over Monica Mar 30 '18 at 09:58
  • 1
    I worked in this area for decades, so I am very disappointed by being voted down. The index is an abstraction; it assumes you can reinvest dividends as they are paid (it takes a few days), there are no taxes (not true for many), etc. 20 days prior to the Russell rebalance we all know what is in the index fund; 20 days after all index funds will match the new index. Index fund managers work hard to make sure they don't get scalped doing stupid things (e.g., trade the whole change in a market-on-close order). Index funds are hard work to keep tracking - lots of math and trading skill. – eSurfsnake Apr 05 '18 at 04:29