On the one hand Jim linked to research that cap weighted indicies lead to a risk of the largest holdings being overvalued.
On the other hand just a tiny number of stocks deliver the bulk of the gains in the stock market. See:
That’s quite a well known study. And probably one of the most misleading studies ever done.
It’s particularly misleading (even to its authors) because the study is well constructed and presumably mathematically correct.
The problem is that the construction method contains a quirk that makes the obvious conclusion the wrong one.
To start with, bear in mind that the average company in the average year rises in value.
A “monkey with a dartboard” strategy that picks (say) 100 stocks, holds them for a year, and repeats, will do extremely well over time.
It is also a near certainty that it will beat a cap weight index over time. (roughly 99% chance, in one study)
Yet, having only 100 positions, it’s also pretty unlikely that portfolio will contain one of the tiny number of “long term winners” identified by the study.
So how to reconcile it:
Think about the lifetime of a public company.
A simple model would be something like this: the business lasts for a pretty unpredictable number of years.
Ignore the ones that are bought out, since those aren’t on average a problem for long run returns.
The remainder are generally profitable and have a positive real total return (on average) from the market for most of that life, however long it may last.
Then they fade and fail in a comparatively short time span, which is by comparison pretty constant.
The key bit here is “comparatively short”.
Let’s say the averagely prosperous lifespan is a random number 2-100 years, followed by 1-4 years of fade and death, as a mental model.
So what’s the subtle quirk in that study that leads people astray?
It held all positions for their entire lifetime…all the way through the fade and death step.
In our toy universe, it’s clear that most stocks will be a total loss during that end stage.
Yet we know that on any given day, most firms are not in that category at the moment.
A random selection of firms in a random finite time interval will have a strongly positive expected return,
because most picks, being chosen at random points in their lifetimes, will not be in their fade-and-die stage.
There is no need at all to be an owner of one of the few long term big winners.
It’s peach season here. Perhaps you have noticed that individual peaches ripen and rot at very different rates.
Imagine buying a basket of peaches, most good and a few rotten.
One person buys those for consumption within three days. Most will be fine, and very enjoyable.
Another person holds the original peaches forever, and observes that they all eventually go rotten.
He concludes that all peaches are a waste of time and can never offer enjoyment.
A separate observation:
That study is also an interesting philosophical mirror.
Most industry participants who chose to comment on it used it to bolster their own pre-existing biases.
The Bogleheads used it as proof that it’s necessary to own everything because otherwise it’s very unlikely you’ll be an owner of one of the few firms that will make you money.
The stock pickers used it as proof that it’s necessary to show judgment, because the only way to succeed is to own the few obvious winners.
I just concluded it was a study constructed in a way that can’t give a useful conclusion.