Sentences to ponder please diversify (but not everyone)

…the entire gain in the U.S. stock market since 1926 is attributable to the best-performing four percent of listed stocks.

Here is the paper, via the excellent Kevin Lewis.


Yeah, my Studebaker shares haven't been doing too well lately.

Apparently the average lifespan of a company in the S&P 500 these days is only about 15 years.

Maybe not such a negative thing though....dynamism/creative destruction is not the worst thing....

Yes, this would seem consistent with the observation that about 90% of new businesses fail. Also, this serves as a good benchmark for government programs and public schools. Do our current systems keep (only) the top decile programs and schools and periodically replace the rest?

By the time we count up company demises, mergers and acquisitioins, is this surprising?

But how does that work given the massive turn over in the top 4%? Companies like Apple did not even exist a few short decades ago.

Do they mean that the top 4% are performing better and better all the time? Isn't it likely that productivity will be, on the whole, captured by the biggest and best companies?

Power laws, how do they work?

No surprise, it's the long end of the tail (average is over).

More freaky facts from the excellent Hebner coffeetable book on index fund investing*:

Figure 4-2 illustrates how an investor who hypothetically remained invested in the S& P 500 Index throughout the 20-year period from 1994 to 2013 (5,037 trading days) would have earned a sizable 9.22% annualized return, growing a $ 10,000 investment to $ 58,352. When the five best-performing days in that time period were missed, the annualized return shrank to 7.00%, with $ 10,000 growing to $, 38,710, and if an investor missed the 20 days with the largest gains, the returns were cut down to just 3.02%. If the 40 best-performing days were missed, an investment in the S& P 500 turned negative, with $ 10,000 eroding in value to just $ 8,149, a loss of $ 1,851.

Hebner, Mark T.. Index Funds: The 12-Step Recovery Program for Active Investors (Kindle Locations 939-941). IFA Publishing, Inc.. Kindle Edition.

*Bonus Trivia:
St. Louis chess club patron Rex Sinquefield is one of the founders of IFA; and this 1980s US chess champion, P. Wolff, ran a hedge fund funded by P. Thiel: ("Wolff was previously a managing director at San Francisco hedge fund Clarium, which is a $3B global macro hedge fund. He left Clarium to launch Grandmaster Capital Management, a hedge fund that received seed capital from Peter Thiel, the founder of Clarium and a strong chess player himself.[5] Hedge Fund Alert reported that Wolff has started the wind-down process of Grandmaster Capital in June 2015. Over the past several years, he has given a blindfolded simultaneous exhibition for all-comers at the annual Berkshire Hathaway shareholder meeting in Omaha, Nebraska headed by CEO Warren Buffett and Vice-Chairman Charlie Munger.")

TC's title: "Sentences to ponder please diversify (but not everyone)" - ah, master Cowen is obviously alluding to the well-known financial find that in order for any trade to occur, for any profit taking to happen, a market needs noise (uninformed traders). If not, then you can show mathematically that no trades will take place. Hence, the corollary is true that if everybody is smart, and everybody diversifies, then there's no alpha to capture. For this reason perhaps the stock market over the year has returned fewer and fewer bargains, as more and more people get smart and index, which cuts down on the alpha available for Bogleheads...

Ray, I don't understand. Bogleheads are not trying for alpha. They're just minimizing costs and getting what the market gives.

Thanks Shane M but I'm using α in a non-standard way (I'm Greek you know). In my mind, alpha is the excess return Bogleheads get from diversification that non-Bogleheads don't receive. If everybody diversified, there would be less return to Bogleheads. Why? Because if everybody diversified, the price of stocks out of favor (value stocks) would be bid up to the point where no Boglehead could make a profit. Remember, a growth stock is simply a very popular 'value stock'. Likewise, a value stock that everybody loves becomes a 'growth stock' that returns almost nothing (becomes part of that 75% that collectively has r=0%).

That's my thesis. Good night.

That's not how it works. That's not how any of this works.

Meh, that's oversimplifying.

For example, there's an ongoing debate in the Boglehead community regarding factor analysis, tilting, and whether factors should count as alpha. If they are really anomalies they probably should count; if they are compensation for risks they probably shouldn't.

This is to say nothing of the many less sophisticated folks arguing for market timing, dollar cost averaging, mathematically poor portfolio and debt choices, etc. It's a good community but it's not immune to human failing.

@Lord Action - thanks. A fascinating debate. Mind you it's the debate over how the average mutual fund investors makes significantly less than they should; it's the debate over concave risk/reward profiles (most of us) vs convex vs neutral (keep in mind mathematically it should not matter how risk adverse you are, since you'll just end up at different points on the Efficient Frontier); it's a debate about whether you can predict the future based on the past; it's a debate over why the equity premium is so high in the USA; and American exceptionalism, fat tails and the like.

Right now I'm slowly transitioning my folks out of their stock portfolio in favor of an index fund that captures the whole market. They had the majority of their 1M+ dollar stock portfolio in just three stocks, all media stocks. It BOGLES my mind that people are that ignorant (or risk adverse?, likely the former since three stocks will not get you on the Efficient frontier since they are not uncorrelated). I'm actually including TC's cite to this paper as one of the slides in my PowerPoint presentation I will make to them soon.

@Ray, do your parents need to liquidate these assets to generate income, or will these assets get passed on to some undeserving wastrel like yourself?

If the latter, you may want to understand inheritance tax rules around "stepped up basis" before liquidating those media stocks.

@BD: Those rules are about to be repealed along with the estate tax. Probably by July.

@msg, Yeah. Yay for the little guy! #populism

LA, correct, I'm probably not being fair. When I posted I was actually thinking of John Bogle's stated views on factor investing.

I personally am a factor investor, but probably unfairly viewed Bogleheads as more monolithic than I should have based on what John Bogle has said.

It's really not insightful to say "if you miss the best 5-40 days of the S&P, your returns go down a lot". What happens if you miss the worst 5-40 days? I bet it's a bigger effect. I bet your returns would go up!

Here's a relevant link, with some nice charts. It references The Ivy Portfolio as source. It's a study of stocks from 1983 forward:

39% of stocks have a negative lifetime return

64% of stocks underperformed the Russell 3000.

The bottom 75% of stocks cumulatively returned 0%

Not to mention the elimination of any failed company from being counted in various Dow Jones measures. Like removing Enron after bankruptcy. Or this type of reshuffling - 'Citigroup was removed from the Dow following the 2008 financial crisis, when the company's market cap shrank by over 90% and it teetered on the brink of bankruptcy. Travelers was spun off from Citigroup in 2002 and went on to replace Citigroup in the Dow in 2009.'

If you remove the true failures, the American stock market looks pretty good. Shame about the Worldcom stockholders, but really, who counts the past? 'On July 21, 2002, WorldCom filed for Chapter 11 bankruptcy protection in the largest such filing in United States history at the time (since overtaken by the bankruptcies of both Lehman Brothers and Washington Mutual in a span of eleven days during September 2008). The WorldCom bankruptcy proceedings were held before U.S. Federal Bankruptcy Judge Arthur J. Gonzalez, who simultaneously heard the Enron bankruptcy proceedings, which were the second largest bankruptcy case resulting from one of the largest corporate fraud scandals. None of the criminal proceedings against WorldCom and its officers and agents was originated by referral from Gonzalez or the Department of Justice lawyers.'

Oh right, those two financial institutions aren't included either, are they?

(And really, wouldn't it be possible to find another adjective than 'excellent' to use when describing Kevin Lewis? After all, this is not a place with a generally Homeric style of epic poetry.)

@prior_test2 - "Shame about the Worldcom stockholders, but really, who counts the past" - Free the Bern! Agree? ("The earliest date Ebbers can be released is in July 2028, at which time he will be aged 86"). I always felt because Ebbers was the 'first' he got a raw deal. Enron's J. Skilling got less time (he's out this year), and Andy Fastow of Enron, who some say was the real resident evil behind the crooked E, got a slap on the wrist (six years).

Yes, Enron and its fantastic footnotes that camouflaged an interesting business strategy. They along with Ebbers and Madoff can rot in Dante's Inferno

Here's my cure for all this stuff. The SEC, or a committee of disinterested wise men and women, appoint public audits, not the audit firms competing with each other. This way the public companies' executives cannot pressure auditors to cover up financial misdeeds.

I'm so old I remember my uncle coming home from the Korean War.

In addition to reading of news, analysis, spreadsheets, etc. (reminds of MacArthur's 1962 West Point Speech) the following words (not "duty, honor, country") run through the (alleged) mind,

"If it seems too good to be true, it is too good to be true." And, "It's not what you buy. It's what you pay."

Re: Enron, it was all in the (overlong) footnotes. You need to read the footnotes. After the "scandal," the FASB and Congress (Sarbanes-Oxley) forced the so-called audit industry to move covert ownership (beneficial interests, etc.) into (consolidate - one-line or whole statement) the financial statements from the footnotes. As usual, the "principles" are overly complicated and susceptible to multiple interpretations.

Do you speed on public highways? Then you too are an outlaw AlanG, in your own way.... first stone, glass house, etc.

Where is the post on the rioting, assault, and attempted murder by the left in Berkeley tonight? No words of wisdom from MarginalRevolutions's radical leftist apologists?

Is anyone under arrest for those charges?

"No arrests were made throughout the night."

"The violent protesters tore down metal barriers, set fires near the campus bookstore and damaged the construction site of a new dorm. One woman wearing a red Trump hat was pepper sprayed in the face while being interviewed by CNN affiliate KGO. She was able to respond that she was OK after the attack."

Did it disappear faster or slower than anti-Muslim terror in Canada?

My advice to the irrational (Cannot Understand Normal Thinking) vagina hats and violent, anti-Trump fascists; Resistance is Futile.

This is only useful if you know which companies will be in that 4% ahead of time!

That's why you diversify ;)

At the risk of making a couple of obvious observations:

1. you don't need to know with certainty the entire cohort of the 1000 "winning" 4% of stocks, you just have to pick a few of them.
2. you also do not have to be 100% correct to do better than the market of course, if you pick one winner and one loser, you are doing better than the 96%/4% split of the entire listed market.
3. accomplishing either (1) or even (2), appears to be immensely difficult and involve a lot of very hard work.

BUT it does appear that it is possible, and is in some sense the entire philosophy behind the value investing strategy, which eschews diversification. In a sense, index investing and skewness of returns are the friends of value investing. A very nice summary of the strategy written in plain English is Seth Klarman's book "Margin of Safety, Risk Averse Strategies for the Thoughtful Investor." It is out of print and sells for >$1k on ebay/amazon, but if you poke around online you may be able to find a digital copy.

The problem is that it is a lot of really, really hard work to find the nuggets, and it may require patience. These qualities are NOT in the skillset of most people, who tend to look for quick and easy gains and think about the stock market more like gambling than a really hard job of pouring through financial reports and legal filings looking for underpriced, cash-generating assets. Because of that, index investing is far better for most people who do not have the time, knowledge, and resources to devote to what is a really, really hard job.

One final point: as far as I can tell based on my very quick read of the article, it appears that the stats in the article quoted above are based on lifetime performance while each security was listed. Obviously most value investors don't invest in an IPO and hope they chose correctly. Rather, they look for companies that for one reason or another are facing what may be irrational downward pressure. A classic play is identifying a company that the market thinks is at risk of bankruptcy, but that the value investor knows from studying the facts is unlikely to actually be insolvent. That security may have a negative lifetime return over its listing period, but from the point where the market is perceives bankruptcy risk forward, it may have very nice returns. Obviously it could actually go into bankruptcy and be wiped out, which is why doing the really hard analysis is so important!

BTW, is the DBG name related to a view on volatility?

I have an excellent investment strategy based on this insight. Sign up for my Newsletter for only $19.95 and you too can Invest Like A Vulture.

Dive in and partake of the road kill to profit!

>Invest Like A Vulture.

Not a Turkey?

Is Turkey (stock market) a buy?

The Dow is setting records (winners) while right wing populists anticipate Armageddon any day now (losers). It seems that the former begets the latter. At least that's the conclusion of several essayists, including Walter Russell Mead and Will Wilkinson. According to Mead, prosperity begat secularization begat cultural division. According to Wilkinson, prosperity begat inequality begat economic division. I suppose both could be right, but the problem is that the prescriptions are very different. For Mead, all of America would join hands singing Kumbaya if everybody became a Presbyterian. For Wilkinson, all of America would join hands singing Kumbaya if taxes paid by the wealthy are increased along with spending on public goods and social welfare programs. No, Mead and Wilkinson don't offer those prescriptions, I just applied a little logic to derive a prescription.

Is this the death of actively managed 401(k) accounts?

This leads to an interesting question: all else equal, does an increase in the portion of total invested assets in index vs. actively-managed funds tend to increase, decrease, or have no effect on index fund returns?

In other words, are the prophets of indexing being altruistic or are they helping others while also helping themselves?

I see two possibilties and I am not sure which makes more sense:

(1) All else equal, asthe portion of invested assets in index vs. actively-managed increases (say it goes from 40% to 70% indexed), index fund holders will see an increase in returns compared to the but-for world without an increase.

This would occur because active management effectively acts as a tax on the market, as active managers take some portion of what would otherwise be a return in the market and consume it instead. Reducing that tax by increasing the indexed portion of invested assets should then increase returns to the indexers.


(2) The same thing I described in (1) happens, but this actually means that to finance the consumption, active mangers are selling cheaper than they should, which allows indexers to passively benefit from active mangers selling for consumption. In which case, a larger portion of invested funds in indexed funds lowers indexer returns.

Over what period? For example, are these the annual best performers. In which case this is a statement that variances are large. Perhaps yahoo is on the list for 1999. Not very informative given the losses in 2000.

Tl;Dr sounds like a sampling effect to me.

I assume they did something simple, but it is an interesting question how firms' lifecycles should really be mapped to market gains, and to hypothetical portfolios.

Index investing has an acquisition rule (as market cap increases relative to the pool) and a divestment rule (as market cap falls relative to the pool).

Accidental but near optimal? Accidental momentum investing?

Maybe not 'accidental' but definitely momentum. As a market cap weighted index, the S&P500 is dominated by the largest companies, so when mega caps are running the S&P is pretty hard to beat. When there's more breadth, or choppier markets, active managers can do pretty well. Take Apple out of the last 10 years and the S&P overall return is significantly lower

Just at a surface level, 4% is astonishingly low.

An indictment of "efficient" allocation of resources by public markets?

Is it really? Don't most businesses go out of business one way or another after a while?

I think it is two different questions. Even high earning companies have lifecycles. But I take this report to say that an astonishing 96% are not high earning, and just muddle along as public companies.

I was somewhat pulling legs about this as an indictment of capitalism, but at the same time we should probably admit a lot of capitalism is b.s.

For instance too many companies competing at too low a margin to deliver crap electronics to stores everywhere. Need a comically large 5 function calculator?

"Over what period?"

From the abstract:

" the entire gain in the U.S. stock market since 1926 is attributable to the best-performing four percent of listed stocks."

Also, what do the results look like if you start in 1946?

Like JWatts says, the period of time determines the "4%" or not. From the other presentation upstream cited by Shane M, 25% of companies make 100% of the profits (collectively) for the "Ivy Portfolio" which covers I would imagine the post-WWII years (when backdated). So as the title on the website "Zero Hedge" says, in the long run, everybody (even the 4% in the referenced sample) goes to zero. So the longer the T=time period, the smaller the percentage. 40 years might be 20%, 1926 to present might be 4%, and since the Buttonwood agreement on 68 Wall Street in 1792, it might be zero percent (no one company has returned all of the return in the US stock market since 1792).

Ok, what's the MR opinion on the best way to utilize (profit) from this information. Is the best advice still a low fee Whole Market fund?

Also, did the study take into account fees? And if so, how much? (I don't have access to read it.)

I will tell you, but first I need to raise the capital to take all of that profit for myself.

Well cancel my subscription to the Turkey Vulture Investing Newsletter then. You sir, are nothing but a scavenger.

The best performing four percent of stocks constantly changes over time, and of course so does the composition of the index, so this doesn't tell us anything at all.

" this doesn’t tell us anything at all."

It tells us that Treasury Notes may be a better investment than conventional wisdom indicates.

"Most common stocks do not outperform Treasury Bills. Fifty eight percent of common stocks have holding period returns less than those on one-month Treasuries over their full lifetimes on CRSP. "

And this:

"The results also help to explain why active strategies, which tend to be poorly diversified, most often underperform."

Which tentatively seems to affirm the more recent conventional wisdom of investing in a whole market fund with low fees.

Long term studies of stock and bond returns may be good, but they all suffer one problem as we apply them to our portfolios.

We were all born in just one year. We start working about 20 years later, we retire about 40 years after that.

Our birth year matters. Many older people bought 30 year trasuries north of 6%. I never did, and baring catastrophe, I never will.

So, watch out for historical studies that include conditions we are unlikely to see.

Agreed but that's why a lot of those studies do look at moving holding periods where the returns are premised on, say, a 40/50 year investment period, rather than 100 or more year periods. (Not this one it seems but....)

Even then, I think there can be a communication issue. Some tools plot all the different ways a portfolio "woulda gone" in history .. but then they might reduce it to a questionable "median return."

The average investor might not understand that not only are they not guaranteed median return, it is also possible that they set a new low.

Still, where there is life there is hope, and without equity exposure you are guaranteed to lose. (Or with less worry, you take the fixed annuity.)

Perhaps. However I've been told that the guys managing ETFs like SPY, XLF, XLI... do not put all the money equally into the underlying set of stocks/indices they are attempting to mirror. They over weight the best performers and underweight the poorer performing stocks to make their money.

I think if anything the moral here is about diversity and the value of entrepreneurial creative-destruction in financial assets. If we' take the view that the moral is that treasury bill might be a better option that thought you're still stuck with the equity premium puzzle where real returns at the aggregate levels (broad investment not single stock investments) seem to perform better than implied risk seems to suggest. (In other words, at the aggregate level the risk appears to be lower than the required returns -- or realized returns -- imply)

"It tells us that Treasury Notes may be a better investment than conventional wisdom indicates."

No. Don't act on such a fraught interpretation. It doesn't tell us much.

It doesn't tell us everything, but it does, I think, tell us the following (which is very valuable even if nothing new):

1. Your chances of capturing the returns of those 4 percent are much greater if one invests in a diversified index (*especially* if the 4 percent changes over time) than trying to pick those 4 percent by individual selection;

2. Holding that index rather than trying to time the market is also important. Not covered by the paper; but, this is also very true for tax reasons. By investing in an index, you can capture the 4 percent of today and tomorrow without selling and incurring tax and you therefore have more assets working for you.

But, I don't think it answers the question of which diversified index is likely to produce the greatest total holding return going forward (unfortunately, the paper and comments thereto vacillate between "stock market gains" and "returns", etc). E.g., is the greatest chance of superior total holding return to be found in 1) The S&P Index; 2) NASDAQ; 3) Total Market, 4) Some other index? Probably all companies in that 4 percent make it into the S&P 500 eventually, but does the greatest growth occur before they make it into that group? The study seems to capture the total market since 1926, but is some other index today better at capturing the 4 percent while eliminating some of the worst performers? It isn't simply a matter of investing in that 4 percent of the *total market* but also avoiding the dogs of the *total market*.

Case in point---S&P 500 versus Total Market since 1928:

Query for J; Watts: How does that compare with Treasury Bills (or even Notes)?

OMG a pop quiz. I am totally unprepared.

"over history from 1928 on, the annual return in the S&P 500 has been 10.4 percent while the return on the Total Stock Market has been 10.2 percent ...In 1928, the S&P index was comprised of only 90 companies."

So, those are both pretty darn good returns.

For comparison, I'm going to glance over at a smarter students desk and crib from their answer. Wow, that guy most have studied all weekend.

So, if I had invested a crisp Benjamin ($100) into 3 month T-Bills in 1928, I would have a quite respectable $1,988. Score!

Now, what about the S&P? If I had invested $100 in 1928, (assumably reinvesting dividends along the way) I would have accumulated $328,646.

Ding, ding, ding. We have a wonder.

Very good, J. Watts, but we are not finished yet.

The only thing I am interested in as an investor is the amount of real money I have after taxes and inflation. The point is to increase one's consumption or potential consumption or that of one's heirs. So, we need to adjust those returns for 1) annual taxes; and 2) accumulated inflation from 1928 until now.

Is that investment in US T Bills in the red after 86 years? Show your work!

It wasn't clear to me how the author handles the event of equities dropping from the database due to mergers. Even if the delisted equity was showing a poor or even negative return when delisted and merged with the other company the shareholders may well have enjoyed significant gains from their new ownership. On one level I can see this doesn't matter to the argument. However, to say that the returns are all attributed to the surviving company without making some adjustments for the new assets and fact that it is a new entity seems to perhaps bias the results towards the claims made.

Yes of course, see the upstream comment of mine on the Ivy Portfolio. Siegel's constant says you make 7%/yr in stocks, going back to 1800, but there's no surviving member of the original stock market of the Buttonwood Agreement, which took place on May 17, 1792, and started the New York Stock & Exchange Board now called the New York Stock Exchange (except maybe Dupont? But they weren't publically traded back then I don't think).

A little related Trump/Fed weirdness:

What type of findings from this author's study would support a concentrated investment approach rather than a diversified one?

I wonder if negative returns are more serially correlated than postive returns. In other words a good investment strategy is just to avoid the losers, not attempt to differentiate between the winners.

Fallacy of composition, like saying "Texas was responsible for 120% of US economic growth in 2008" if total national growth was less than that in Texas.

This is correctly stated as "the top four percent of stocks since 1926 had returns as large as the returns for the entire market over that period." This is really not terribly interesting given that everyone already knows most of the companies formed since 1926 don't exist anymore.

Comments for this post are closed