It thus takes roughly 40 large stocks today to equal the diversification benefit investors got from just 20 stocks in the early 1990s, warns Mr. Sullivan of the Financial Analysts Journal. Anyone who doesn't use index funds should be aware that portfolios of only a handful of stocks are more prone to sharp fluctuations than ever.
It appears that the conclusion above is tied to the proliferation of index funds buying and selling many different stocks at a time but the work behind coming up with 40 being the new 20 was not disclosed and so I cannot vouch for it--based on the article it might simply be an assumption of some sort.
Regardless of the accuracy of the numbers cited the idea is fascinating; does it really take many more holdings today to create diversification than it did 20 years ago and if so how many more stocks does it really require?
This concept is more about the evolution of portfolio construction and markets more than anything else IMO. Top down analysis says the most important decision is whether to be in or out of the market. More practically this means whether to be defensive or not as getting completely out has some logistical issues (taxes, commission drag) along with strategic issues too (what if you're wrong?).
A point about top down analysis I have made several times in the past is that if some industry, for example semiconductors, collectively goes down 30% for some reason it is very unlikely that the typical investor will pick the one that somehow goes up. It is possible to pick the one that goes up but it is not probable.
So in a decade where US markets went down, an overly US-centric portfolio appears to not have much diversification as most of the holdings probably went down but there were no such worries in the 1990s, when everything goes up diversification, in hindsight, becomes less important (obviously this is a psychological issue).
Not isolated in the article is what type of diversification is being sought. A portfolio of 20 holdings each targeted at 5% is not taking unreasonable single stock risk--US financial companies in 2008 notwithstanding the road to $0 is rare and when it happens there is usually plenty of time to see trouble (even if there is no expectation of zero) and get out before $0.
Considering the above, then the issue is correlation and this is mentioned in the article however I would go back to the idea of evolution and point out for the umpteenth time that while the US was going down 24% on a price basis there were plenty of markets that had normal returns and other markets where returns were better than normal. Anyone who has been reading this site for even a little while knows the countries I mean in this context, and anyone reading this site from back in 2004 when I started blogging knows how simple most of the macro analysis was for these places.
I'd love to portray this has having been very complicated and sophisticated but it wasn't. Chances are that if the benchmark index in some country goes up 100% in ten years (that is enough to get the job done) then many of the larger constituents in that index are going to be up about the same amount. Picking a few countries correctly and correctly avoiding a few and pretty soon you've got a return that is close to normal and could be done with 20 holdings.
As it happens, I don't think 20 is ideal for accounts above a certain dollar amount. I've targeted in the mid 30s for a while now but that is just me; 20 may be great for you or you might want 50 or 60. I would submit that what has changed is not the number of stocks needed for diversification but where you go for diversification and when you take protective action in the portfolio.
We had two more lightning caused tree fires over the weekend. We had much shorter hikes this time ( a couple of weeks ago we hiked for two and a half hours to get to a fire). Nine of us responded on Saturday and five on Sunday which are both very good turnouts. I started out as incident commander for both and ceded to the Forest Service when they arrived, both were on Forest Service land.