A theoretical question regarding indexes...

Old_Joe

A theoretical question that I've long wondered about regarding indexes. Let's consider an equity index.

1) Perhaps my concept is faulty, but it's my understanding that such an index is essentially a representative array of stocks (companies), proportioned so as to mimic as closely as possible some specified index.

2) That specified index is going to be affected by the aggregate performance of all of the individual equity listings which constitute that index.

3) The performance of those individual equities, and thus the index itself, is determined by the buying and selling activities of active market participants.

For simplification, let's postulate that there is some fixed approximate number of market participants available, both active and passive (index) investors, for some particular market segment- let's say the Russell 2000 just as an example.

4) if there was no such thing as a Russell 2000 index fund, then all of these investors would by definition be "active".

5) OK, let's introduce a Russell 2000 index fund. Now, some percentage of those "active" investors become "passive" investors. These passive folks are content to let the remaining active investors determine the outcome of their index.

6) Presumably (and again theoretically) you could eventually get to the point where 50% of those investors interested in the Russell 2000 are active, and 50% passive.

7) Let's postulate that the passive index investors do so well and are so happy with life that more and more active investors join them. Let's exaggerate to the point of absurdity, just for the exercise:

Could we get to the point in any given index situation that, say 10% of the potential investors are determining the results for the remaining 90%?

There's a serious question underlying this obviously improbable scenario: Do passive indexers, by relinquishing activity, potentially change the performance of a particular index by deferring to the choices being made by the active investors? If we make the reasonable surmise that the active investors have a somewhat different personality spectrum from the passive group (maybe more informed, or financially more aggressive, or perhaps more inclined towards risk-taking, or financially very large vs small) doesn't the performance of any given index subtly change every time an active investor throws up their hands and becomes an "indexer"?

Ted

@Old_Joe & MFO Members: For almost all of 2015, just 5 stocks in the S&P 500 Index accounted for all of the gains. This is what you get with a cap weighted index. The same pattern is continuing in 2016 with the S&P 100 up 1.78% and the S&P 500 down -(0.04)%
Regards,
Ted
Bad S&P 500 Data Point Last Seen At Dot-com Bust:

fundalarm

Yes, OJ. It is active market participants who determine index values. the fewer active participants, the less price discovery. you're right to be concerned.

Old_Joe

@Ted- thanks for that observation, Ted. It actually adds to the seriousness of my question. Here we have a major index heavily influenced by 5!! stocks. Now lets take that index and make it more and more passive. Doesn't this inevitably work towards a concentration of power on the part of fewer active investors influencing the fates of just a few stocks, and thus an entire index?

I'm getting in over my head here, and it'd be nice if some of our intellectual heavyweights contributed their thoughts on this.

msf

I believe price discovery can take place even with only indexers. But what is "discovered" is only the value of the index (or market) as a whole. If investors feel the market is underpriced (even if individual components are overpriced), they will bid up the index, just as any individual security is bid up when there's demand for it.

What is "undiscovered" is the relative value of the components. Some companies may be overpriced and some underpriced, but if you're stuck buying all or none, you won't change their relative prices. Not dissimilar to buying a whole company even if you don't think all of their divisions are worth owning.

The net result is that capital is not deployed as effectively as possible, since the dogs continue to get fed at the expense of the stars. Or, a few active investors do all the allocating.

Old_Joe

Ask for some heavy firepower, and along comes msf, one of the folks that I was hoping for... thanks!

I hadn't considered the fact that merely buying or selling the index itself also is a type of input factor. How often does an index actually "reconstruct" itself to match the underlying individual assets? Is it once a day, after market close (like most funds), or is it an ongoing computerized adjustment?

msf

Well, I was also hypothesizing, so I hope my thinking made sense.

I think by "index" here you mean a fund tracking the index. Remember the standard disclaimer - "you cannot invest directly in an index."

I know people think that index funds can be run by computer, but there is real skill involved - because some funds use sampling, because funds need to manage cash flows (purchases, redemptions, portfolio dividends/interest received, dividend distributions). So trading is constant.

Here's an old "The Street" column entitled "Managing an Index Fund Isn't Monkey Business":

Gus Sauter's "team of traders is constantly investing throughout the day, looking for the best price and the most efficient trade"

Old_Joe

@msf- I really don't know what we'd do here if it wasn't for some of you pros. You are absolutely correct, I did mean "a fund tracking the index", but I was trying to keep my questions as simply worded as possible. Good article, and I'd never have guessed that anything like that was involved.

Thanks again, msf... you be a cool dude!