Post New

This Stock Market May Not Be That Overpriced After All
https://www.barrons.com/articles/stock-market-not-...

source of new study -
https://www.researchaffiliates.com/publications/ar...

amend CAPE with “Current Constituents CAPE" to remove "constituent elimination" bias, this stock market, although still expansive, may not be that overpriced after all (relative to 2000).

... CAPE becomes flawed when a stock is deleted from the index and replaced by a new one. That’s because the deleted stock’s earnings will remain part of the traditional CAPE’s calculation for 10 years after its deletion, and only gradually will the replacement stock’s earnings be included.
... the researchers found that the CC-CAPE had a better track record over the 1964-2024 period when forecasting the stock market’s return over each of the subsequent one-, three-, five-, and 10-year periods.

The moral - There are always many biases in simple measures. Don't just blindly adopt.

I'm not sure I quite follow this as an optimistic view. As both articles are paywalled, I read only the summaries, but bear with me.

Starting from the basics underpinning the whole idea of CAPE analysis:

* The value of any set of equities ultimately derives solely from their aggregate net profit making ability, within rounding error.
* The S&P 500 is such a large fraction of the US economy that neither its profits nor value can grow any faster than the US economy over long time periods, which is a relatively steady long term trajectory once the business cycle is smoothed out. (profits from imports and exports matter, but they and their changes mostly cancel out well enough to ignore them)
* Taken together, this means that the likely aggregate profits of a broad enough set of equities can be extrapolated with sufficient accuracy to be useful, if not precisely, simply by doing the same sort of business cycle smoothing.
* If you know what future aggregate profits are likely to be (roughly), you have an idea of what would be a "normal" price to pay for that future stream by comparing the pricing to history. Future pricing may not be the same as the past norms, but you have to start somewhere.

If the set of stocks you're considering is still broad enough to be strongly representative of the corporate sector as a whole (a prerequisite for any CAPE discussion to have any meaning at all), and the new entrants to your sample have lower profitability than the ones leaving the index, it seems to me that this just means that the economy is less profitable than it used to be, not that a given dollar of upcoming profit is worth more than it was before.

Jim

" By keeping the deleted stocks earnings in the CAPE calculation instead of the full 10 years of the replacement stocks’ earnings, the CAPE will be biased. The extent of this bias is evident from the accompanying chart."

Isn't that grand? Anyway, for those of us who own brkb, the only opinion that matters is Buffett's. Buffett hasn't been a serious buyer of brkb since 15 % lower, and I can't recall the last 1/4, he was a net buyer of equities. I suspect we know which opinions Buffett respects.

The Research Affiliates article is not behind a paywall. One might have to 'register', I do so with a fake name and burner email. But these tend to be smart guys, so stuff they might send you might be interesting stuff.

Here's a quick synopsis:

A recent example illustrates the difference between CC CAPE and Shiller CAPE. In December 2020, the S&P 500 Index added Tesla and deleted Apartment Investment and Management (AIV). As we noted at the time (see here and also here), it was a classic example of a cap-weighted index buying high and selling low.

The calculation of Shiller CAPE would not have included Tesla's earnings losses in the decade before being added to the index, but it would have included AIV's modest earnings. The denominator would have been larger as a result of including AIV's positive earnings instead of Tesla's negative earnings. In contrast, the calculation of CC CAPE would have excluded AIV's earnings history as soon as the stock was dropped from the index and would have included Tesla's previous decade of negative earnings as soon as it was added. This approach would have made the denominator smaller. If Tesla's market cap at the time of inclusion was also proportionally smaller than that of AIV, the impact on CC CAPE versus Shiller CAPE would not have been so large. On the contrary, Tesla's market cap at the time of inclusion dwarfed that of AIV, exacerbating the difference and making CC CAPE higher than Shiller CAPE.


They plot the CC CAPE ('Current Constituent' CAPE) and Shiller CAPE over time, showing that they differ significantly, on same plot is the difference of CC - Shiller i.e. which they call the 'spread'.

They show that CC_CAPE predicts signifcantly better than Shiller_CAPE over 3 and 5 year future return horizons

Ultimately, the best explanatory power for future returns comes from combining Shiller CAPE and CAPE Spread because this combination contains information about both starting-point valuations and sentiment. As the table below shows, combining them (i.e., Shiller + Spread) more than doubles the correlation with future returns at the shorter horizons of 1, 3, and 5 years compared with using Shiller starting valuation alone.

The use a notation of "Shiller + Spread", which if taken literally is actually just their CC_CAPE.
Perhaps they mean they do a fit of Shiller and Spread to future returns, like this:
FutureReturns = a*ShillerCAPE + b*Spread + c
where a, b, c are adjustable constants.

If so, it would have been useful for them to fit a, c, b using data only up to a given date, i.e. walk the date of evaluation forward in time using only data available to that date to fit the coefficients.

Later they show a table that shows that CC_CAPE predicts a somewhat rosier return future than does Shiller CAPE.

I enjoyed the article and the additional analysis on some of the underlying drivers of Shiller CAPE multiples. The article doesn't spend a lot of time critiquing Shiller CAPE. Mostly, is supports the concept and the value of looking at starting valuations. By either measure, they argue that valuations are at the top decile of this historical range and the future returns are likely to be disappointing to most investors.

But they do show that the dot com bubble was a more extreme outlier than Shiller CAPE might indicate because index changes don't show up in the Shiller PE as traditionally calculated until the new components have time to get incorporated into the history. So today, the market valuations are high, and the CC CAPE is higher than Shiller CAPE, indicating some exuberance, but not nearly the extreme that the dot com period had.

In December 2020, the S&P 500 Index added Tesla and deleted Apartment Investment and Management (AIV). As we noted at the time (see here and also here), it was a classic example of a cap-weighted index buying high and selling low.

The thing that is amazing, even remarkable is that despite the "fund managers" of the S&P500 "classically" buying high and selling low, they STILL outperform 95+% of all other fund managers over the medium term, and STILL outperform 99+% of all other fund managers over the long term. Astounding, isn't it?

I think only Warren Buffett, Bill Miller, Peter Lynch, Stan Druckenmiller, and Ray Dalio have ever outperformed the S&P500 over any two or three decades.

The thing that is amazing, even remarkable is that despite the "fund managers" of the S&P500 "classically" buying high and selling low, they STILL outperform 95+% of all other fund managers over the medium term, and STILL outperform 99+% of all other fund managers over the long term. Astounding, isn't it?

Well, to be fair to the active fund managers, the biggest reason for that is fees. Rather surprisingly, the stock picks of most active managers do beat the cap weight index on average before fees and expenses and the drag from cash allocations. Not by much.

That said, it's not at all difficult to beat the S&P 500 if you're not letting taxes determine the outcome. A monkey with a dartboard has about a 99% chance of managing it. (number not picked from thin air, it was a real test)

Commissions and spreads and price slippage are pretty low these days for those of us with portfolios a few orders of magnitude smaller than that of Berkshire, and many portfolios are not taxed as they go. Test a strategy similar to this: pick a number of stocks to buy, N, maybe 40 or 75 or 150. Mainly it's a function of how much you dislike typing. Avoiding days anywhere near index reconstitution dates, put each 1/N of your portfolio value into a randomly selected stock from among the S&P 500 set (or Russell 1000 or S&P 1500 or whatever) that was not added to the index in the last few months. Hold each position for a year (anywhere 6-24 months) then replace it with another random pick. (Do NOT sell something early that is ejected from the index). The stock swap dates can be staggered: for example, you might swap 20 positions every few months. This is extremely likely to beat the S&P 500 over time, before tax, for a variety of empirical and theoretical reasons. Mainly by avoiding having outsized amounts of capital allocated at any given time to whatever is most overvalued at the time, but also avoiding the index-inclusion drag effect. (stocks newly added to an index are statistically extremely bad performers for the next year, causing a drag of at least 0.22%/year in the case of the S&P 500).

There are stretches that only outsized positions in the very largest firms will do well, as in the late 1990s or the last two years, and this strategy will not beat the index at those times. But it will most of the rest of the time, and overall, and will do so with a fraction of the single stock risk.

Interestingly, the "dart board" step is quite important. If you hand pick the stocks to buy, your chances of beating the index will plummet. You're suddenly playing the game everyone else is, with no better information, the exact trap that index investing was designed to avoid.

Jim

“ A monkey with a dartboard has about a 99% chance of managing it.”

Very Interesting. Entertaining visual as well- could dress it up & market it as Primate Asset Management 😉

Well, to be fair to the active fund managers, the biggest reason for that is fees. Rather surprisingly, the stock picks of most active managers do beat the cap weight index on average before fees and expenses and the drag from cash allocations.

Which active fund managers have beaten the S&P500 over any 2 or 3 decade period if you exclude management fees?

A monkey with a dartboard has about a 99% chance of managing it. (number not picked from thin air, it was a real test)

These studies showed that a "monkey with a dartboard" can outperform the S&P500 in the short term. The same studies have shown that it does NOT apply over the long term (let's say 2 or 3 decades).

These studies showed that a "monkey with a dartboard" can outperform the S&P500 in the short term. The same studies have shown that it does NOT apply over the long term (let's say 2 or 3 decades).

Sure they do. Just not by much.


Here is a super simple strategy to demonstrate, about four decades:
Buy everything in the Value Line database of 1700 stocks, equal dollar amounts. Hold for a year and repeat. If anything is bought out, just hold cash till the end of the year. If anything is delisted, sell it at the first pink sheets trade price and hold cash the rest of the year. Allowing 0.4% round trip trading costs, that beat the S&P 500 total return by 0.64%/year over the last 39 years. (0.4% is a bit low for the early years but reasonable to generous after that)

Of course, the cap weight index has really been on a tear just lately, so today is a particularly lucky endpoint for the S&P side of the test. The same scheme showed the equal weight annual approach with an advantage of 1.17%/year as recently as two years ago (i.e., 1986-2022 inclusive).

A monkey with a dartboard will have the same average return as the strategy mentioned, plus or minus random statistical noise each year depending on the number of stocks chosen. The figure will be a little different depending on the underlying index "universe" you choose, but not wildly so.

There are many possible weightings for a broad portfolio of medium-to-large cap equities. Only one seems to be a bit of an outlier: cap weight seems to be quite a lot worse than almost every other choice, even some choices that are nonsensical. Remember that cap weight index funds were designed to be low work (minimized trading events), not designed to be high return.

Jim

Philosophico-investment question for Mungofitch:

Of course, the cap weight index has really been on a tear just lately, so today is a particularly lucky endpoint for the S&P side of the test. The same scheme showed the equal weight annual approach with an advantage of 1.17%/year as recently as two years ago (i.e., 1986-2022 inclusive).

As I contemplate the difference between equal weight and cap weight, it seems to be equivalent to this:

1) Cap-weight means you own the same fraction of every company you own. If your universe of stocks has market cap $6trillion and your investment pool is $6million, you would simply buy 1/billionth of each stock you put in your port. If the company had 14 billion shares outstanding, you would buy 14 shares. Maintaining the port is easy since as stock prices go up and down, your fixed 14 shares stays correct. You would only have to "rebalance" if the company were materially changing its number of shares outstanding buy either a) share buybacks reducing its count or b) public offerings or dilution through exercise of options grants increasing its count.

2) Equal-weight means you own the same dollar value of every company you own. If you had a $6million portfolio and you had 5000 names in your universe, you would own $120 of each company in your port. An important part of an equal-weight port would be deciding your strategy for rebalancing. What are my results expected to be if I rebalance daily? weekly? monthly? quarterly? or annually? Ignoring trading costs my intuition is you would get around the same result for daily weekly and montly rebalancing, and might start seeing a difference for quarterly or annual rebalancing. As the length of time between rebalances increases, your results should look more and more like those for a cap-weighted portfolio.

So the philosophico-investment question is: is LTBH not a good idea? If we had owned BRK and SP500 in equal parts in a port since 1950, would not cap weighted strategy have outperformed equal-weighted strategy? I guess this is a question I should do the tests to back test, but I'm guessing you know the answer already.

The other intuitive gap from equal-weight I have is, are you sure you are better off selling the shares you own in companies that are seemingly rising more each year than most of your other stocks?

Philosophically, equal-weight and cap-weight each have arguments for seeming to be good choices. For equal-weight, the argument seems to be that most stock price motion is excursions from the mean so rebalancing regularly is equivalent to buy-low and sell-high. For equal-cap the argument might be, Large caps all started as small caps and proved that they could grow faster than their peers by becoming large caps, so you should hold your winners and let them keep doing their thing.

Thanks for your attention.
R:)

Philosophico-investment question for Mungofitch:

Hmm, is that my specialty? : )

This is the bit where I see some space between your thinking and mind. You mention:
1) Cap-weight means you own the same fraction of every company you own...

But this isn't right. You're owning the same fraction of the market cap of every company. But price (market cap) and value (the business reality) are usually different. There are indexes which attempt to hold the same fraction of the actual value of every company, often called "fundamentally weighted".
This is a nice summary, provided you don't place too much weight on the fourth section which is mainly claims about how fabulous their product is.
https://www.valueweightedindex.com/

Because of this distinction, the cap weight can be better thought of this way, divided into three types of holdings:
For things that are currently fairly valued, you have an allocation that is proportional to the firm's size. That should do fine.
For things that are currently undervalued, you have an allocation that is smaller than the amount that would be proportional to the firm's size. A missed opportunity to make a bit from mean reversion of valuation levels.
For things that are currently overvalued, you have an allocation that is larger than the amount that would be proportional to the firm's size. This will lose you money, sometimes a lot.

If some very large firms are currently overvalued (always slightly true, occasionally VERY true), effect #3 results in a very large dollar allocation to overvalued stuff. This is not something that is desirable.


Ignoring trading costs my intuition is you would get around the same result for daily weekly and montly rebalancing, and might start seeing a difference for quarterly or annual rebalancing.

Oddly enough, this can be measured, if you have access to a really good database and tester. No, it's not equal. There is random noise in market valuations at every time scale, and rebalancing more often gets you more because you can capture really tiny gains at a very high frequency...before trading costs. Shorter periods give higher returns, but at a shrinking incremental benefit as the times get shorter. Plus, there seems to be a rough price mean reversion period bell curve: things get overvalued for only so long on average. You want to capture a bit of the upswing, but get out before the down cycle, so there are hold periods that work better than others even without considering trading costs. In practical terms, trading costs matter a lot, and what matters even more is whether other folks are doing the same trade at the same time. Presumably a lot of people are taking advantage of the fact that RSP does its trades quarterly on precisely known dates, causing them a lot of drag. Plus of course the same drag effect that the cap-weight S&P has from pre-announced additions and deletions that get front-run. FWIW, based on a whole lot of tests (that may or may not be very meaningful), I use a 10 month rebalancing cycle in my own quant portfolio. I recalculate positions every 2 months, but positions that get held over several cycles in a row don't get trimmed back to the size of the average position until the 10 month mark.

The other intuitive gap from equal-weight I have is, are you sure you are better off selling the shares you own in companies that are seemingly rising more each year than most of your other stocks?

This ties into the previous comment. Obviously in an ideal world you would like to ride winners as they grow, but that is REALLY hard to do. Why is it hard? Because you want more and more money allocated to firms which are growing in true value, but only those not growing in price any faster than their value. Assessing true value is extremely difficult. The problem is that things getting overvalued among those rising in true value quickly is more common than not. So your first focus should be on avoiding having large bets on bubbly things, which (by contrast) is really easy to manage: you can't be overweight overvalued things if you're not overweight anything. Just don't have any big positions. Yes, it cuts out being overweight the best deals which is in general almost impossible to manage, but more importantly, it cuts out being overweight the WORST deals. Statistically, skipping big losses is more important than having a big position in winners.

The average stock in the average year-or-two goes up in value quite a bit. That observation alone is enough to make for a good investment strategy. There has been a lot of press over a study showing that over the long haul the great majority of stocks contributed nothing to the current value of the indexes, but that's a bit of a red herring. The explanation of the mismatch is only clear when you remember that when a firm ultimately dies, it is usually in a relatively short period relative to the length of the history of the firm. Imagine that the average public company has (say) 9 years of "nice" followed by 1-2 years of "oops", then out. Imagine thousands of such firms, out of sync with each other in terms of their life cycle and their cycle lengths. Imagine picking a big random sample from that group and holding for a year. On average you will get a lot more winners than losers, because stocks experiencing an "oops" year are rarer than those that aren't. So, even if it is true that the great majority of firms are worth nothing if you hold them forever, that observation doesn't matter unless you do in fact hold them right through senescence and death. A portfolio of randomly picked "plausible" firms (index inclusion is a simple way to start), held for a year or two, then (importantly) completely reconstituted, does amazingly well. Personally I use "loaded" darts, trying to do a tiny bit better than average quality picks, but that's just for bonus points. My quant portfolio is equally weighted, currently about 72 stocks.

Jim

If we had owned BRK and SP500 in equal parts in a port since 1950, would not cap weighted strategy have outperformed equal-weighted strategy? I guess this is a question I should do the tests to back test, but I'm guessing you know the answer already.


I have read a number of papers on rebalancing, including studies on the "best" rebalance period.
Which study claimed that the optimal time between rebalances is somewhere between 7 years and never.

Which intuitively makes sense. To rebalance you sell winners and buy losers & also-rans.

Anyway, it was easy enough to run your backtest. 50/50 S&P500 and BRK-A. Limited by BRK inception of 3/17/1980.
One portfolio never rebalanced, one annual rebalance.

Starting with $10,000, the B&H grows to $13.5 million, the annual rebalanced grows to $7 million.
CAGRs 17.4% vs. 15.7%


https://testfol.io/?s=iBPXwJqMie9

To rebalance you sell winners and buy losers & also-rans.

Anyway, it was easy enough to run your backtest. 50/50 S&P500 and BRK-A. Limited by BRK inception of 3/17/1980.
One portfolio never rebalanced, one annual rebalance.

Starting with $10,000, the B&H grows to $13.5 million, the annual rebalanced grows to $7 million.
CAGRs 17.4% vs. 15.7%

Clearly it is better not to reduce your exposure to a stock like Berkshire that ends up outperforming the S&P, from 1980 to 2025. The question is, is it a good idea to reduce your exposure to stocks that have gone up, overall, some of which will continue to outperform, and others which will fall back because they had become overvalued. Which of these 2 is the more important effect?

Although the maxim 'let your winners run' suggests that you should do that, and this will often work out well, over the universe of all S&P stocks, we know that the right answer is to rebalance. Why? Because the equal weight S&P has outperformed the cap weight S&P over all time periods that are sufficiently long. Not recently, mind you, but even now, if you go back X* years, the equal weight index outperforms. This means that, in general, it has been better to sell the winners and buy more of the losers, however much this might differ from our intuition.

DTB

*I will calculate what the exact number of years X is, but whatever it is the general point stands.

Which study claimed that the optimal time between rebalances is somewhere between 7 years and never.
Which intuitively makes sense...
Starting with $10,000, the B&H grows to $13.5 million, the annual rebalanced grows to $7 million.

Beware what intuitively makes sense : )
That result without rebalancing Berkshire is certainly a great demonstration, but only if you have a time machine and know the big winner in advance.

The whole point of equal weight is that (a) you almost certainly don't know in advance what will outperform, so (b) don't place any outsized bets on anything, which (c) ensures you never have an overallocation to something temporarily overvalued or doomed. Now, we here at the BRK board know in advance what will outperform, but the reasoning has merit for mere mortals elsewhere : )


Though it's not answering the same question, an equal weight portfolio of the S&P 500 members has beat the S&P 500 itself by quite a lot.
For a data set I have handy, 1926-01-02 through 2024-03-01, just under a century, the equal weight beat the S&P 500 and its cap-weighted predecessors by 2.191%/year. There are certainly stretches that cap weight wins, sometimes by a lot when there are some very large caps in fashion, but that is a minority of the time. Usually the very largest caps underperform the "S&P 495".

That uses RSP returns after fees since its inception in 2003, the historical construction of the S&P 500 Equal Weight index from S&P themselves 1990-2003, and a research database prior to that.
For earlier periods I estimated and applied the rebalancing costs using the overlap period that I have both the official series and the research database.

Jim

I have read a number of papers on rebalancing, including studies on the "best" rebalance period.
Which study claimed that the optimal time between rebalances is somewhere between 7 years and never.

Did a quick check using GTR1 for SP500 constituents.

The assumption is that when new stocks are added and old are removed, the funds from sales are equally distributed among new stocks, not all stocks. Then rebalancing takes place at some future date depending on the re-balance schedule.

No friction in calculation. From 1957 using sp500.a=1 in GTR1

10d - 12,5% CAGR
20d - 12.3%
63d - 12.2%
126d - 12.2%
253d - 12.2%
2yr - 12.1%
3yr -12.4%
7yr 12.2%

Looks to me, don't sweat the details. Do yearly for long term capital gains or if in a tax protected account do yearly and take out RMD. I think I would even wait for a year to add and remove stocks. No rush to add the new stocks as they have run up in price before the announcement.

Aussi

For a data set I have handy, 1926-01-02 through 2024-03-01, just under a century, the equal weight beat the S&P 500 and its cap-weighted predecessors by 2.191%/year.

The data that includes the Great Depression is a different world. What is the result for the "modern" world, say 1/1/1950 to now?

FWIW, the backtest I see for RSP vs. SPY (beginning 4/30/2003 RSP inception)
RSP beats SPY by 0.29%/yr.

SPY was a rocket in 2023 & 2024. Excluding those, RSP beat SPY 1.49%/yr 2003-2022.

What is the result for the "modern" world, say 1/1/1950 to now?

GTR1 allows this comparison to be made on a synthetic basis -- {!S5T} vs {!S5TE} -- back to the 1925. Since 1950, the equal weight S&P has out-performed the cap weight version by about 100 basis points.

Baltassar

Because the equal weight S&P has outperformed the cap weight S&P over all time periods that are sufficiently long. Not recently, mind you, but even now, if you go back X* years, the equal weight index outperforms. This means that, in general, it has been better to sell the winners and buy more of the losers, however much this might differ from our intuition.

Does the equal weight S&P500 also outperform if you don't sell the ones that are removed from the S&P500 each year? There's also some "survivors bias" here and I suspect that letting winners run while the failures fall off naturally might possibly overcome equal weight while dropping the failures from your list. I wonder if anyone has run those numbers?

Sometimes I'm not sure I understand everything you write, but I love reading it. It's going into my brain and I'm sure there will be a time when it becomes useful. If not directly as stated but as a principle to understand.

Sometimes I'm not sure I understand everything you write

Well, don't assume it all makes sense. Mainly I'm an excellent typist.

Some short points:

Empirically, there is a whole lot of historical observation evidence that an equally-weighted broad market portfolio will outperform a capitalization weighted one. In most 5-year intervals, and overall. It has led in the past for so long that it seems likely to continue. So if you believe otherwise, you should have a very good reason to hold that belief.

Even without historical back-up, there are sound theoretical reasons to think equal weight will do better. At any instant, a cap-weight index (and all other weightings with effectively constant share counts) is overweight whatever is most overvalued, and underweight whatever is most undervalued. An investing lifetime is made up of a string of moments, so that's the drag over the long haul.

Separately, the S&P 500 index since 1989 (and the Russell 1000 since forever I think) have a very serious problem: index additions and deletions are pre-announced, front run, and cause a scarily huge penalty. If you're going to follow the S&P 500, do it yourself; buy all the stocks in proportion to market cap (or any other weighting you like), but follow index changes only after at least 6, preferably 12 months.

An equally weighted strategy has a small fraction of the company-specific risk. If one of the biggest firms in the S&P 500 went "pop" tomorrow, it's the cap-weight folks who would be licking their wounds, as they had (say) 3-7% of their money in it rather than 0.2%.

An equally weighted strategy is never overweight an overvalued large cap, or a set of them. History shows that this is a situation which frequently presages very bad returns from the index for very long stretches. An equally weighted strategy will also do poorly in long bear markets, but not THAT poorly. In the ten years March 2000 to March 2010, the equal weight real total return beat the "standard" S&P 500 by a remarkably 6.55%/year. What is rather less appreciated is that it also beat the cap-weight S&P 500 in the ten years ending the tech bubble in March 2000, though only by 0.39%/year.

Cap weight has been the big winner lately, for sure. Things may indeed be different this time, to some degree. And there is no reason the advantage of equal weight has to be as big in future as it was at particular times in the past. Still, the evidence suggests that it seems more likely than not that the recent advantage of cap weight is likely to go into reverse to some degree at some point, so equal weight is the safer and smarter bet at this juncture.

It's all academic for me, as I consider index investing immoral. There are firms that I won't invest in for ethical reasons, and they are in most any index I might try to track. What was it Charlie said about raisins and turds?

Jim

Mungofitch:
This is the bit where I see some space between your thinking and mind. You mention:
1) Cap-weight means you own the same fraction of every company you own...

But this isn't right. You're owning the same fraction of the market cap of every company. But price (market cap) and value (the business reality) are usually different.


I chose my words carefully: you are owning the same fraction of every company in a cap-weighted index. If every company in a cap weighted fund has a billion shares, Then the fund would own, say 10,000 shares of each company in the index. That is it would own 1/100000th of each company. And this would be a market cap weighted fund. It doesn't matter that some of the stocks might be overpriced and others might be overpriced. You still own 1/100000th of every company in the index if it is a market cap weighted fund. N'est-ce pas?

By contrast in an equal-weight fund, you own a bigger fraction of the smaller companies so that all of your holdings have the same total price when in balance.

Not trying to be intentionally pedantic. Just asking because there may be a point you are making that I am missing.

Cheers,
R:

PS loved your answer! Equal weighted may the ticket I've been looking for as I am concderned that SP500 is overvalued, but it is indeed more so the large caps in there that seem likely overvalued.

It's all academic for me, as I consider index investing immoral. There are firms that I won't invest in for ethical reasons, and they are in most any index I might try to track (mungo, two posts thataway ^ )

Jim, thanks for the reassurance that I'm not the only one who thinks that way.

There's an anecdote in one of the Buffett biographies - probably the Lowenstein one, but I'm too lazy to look up - from thirty-plus years ago where Mr Buffett was at some fundraiser or some such.

I've forgotten the whole story, but the gist was that the emcee asked the audience something like, For the last fifty years, which security has continually increased its dividend AND has had the highest total return? (Again, I don't have the exact details close at hand but that was the general idea)

Anyhow, without a pause Mr B said "Philip Morris" (or maybe just "Mo")

The point is that then and now - cigar butts, value stocks, moats, whatever - MO had been right there as an obvious stellar buy, and for ethical reasons Berkshire had deliberately and steadily foregone investing in it.

To me, that was another nontrivial number of points in his favor which ultimately resulted in my directing most of my money for he and Charlie Munger to manage.

(My other substantial investment was Costco. Similar reasoning)

-- sutton
Yes, I'll stipulate that ESG investing is not much more than dumb and costly virtue signaling. But.

MO had been right there as an obvious stellar buy, and for ethical reasons Berkshire had deliberately and steadily foregone investing in it.

Quite some time ago, he did invest in tobacco companies. A few times in fact...though never anywhere near the "favourite holding period" of forever.
https://genehoots.substack.com/p/berkshire-hathawa...

I suppose that in some obscure ways it's even more impressive...not a lot of people can change their minds.

Jim

I understood everything you said. Thank you. :)

I chose my words carefully: you are owning the same fraction of every company in a cap-weighted index. If every company in a cap weighted fund has a billion shares, Then the fund would own, say 10,000 shares of each company in the index.

Yes, in that sense.

However, a minor point: there are no "cap weighted" indexes in that sense. (and QQQ isn't even close). They are almost all "float adjusted". So you're not getting the same fraction of each company, you're getting the same fraction of the listed and tradeable portion of each company. Some indexes will exclude various types of restricted or insider shares or shares held by other companies (often parent companies) or governments.

I guess a bigger question is simply questioning why you would want to own the same fraction of each company. Bigger isn't (statistically speaking) better, so why buy more of the big ones? The reason cap weight index exists is because it's easy for a fund to track, not because it's what's best for investors: until an index change or M&A, no trades to do, but riding confidently into every bubble.

Jim

The point is that then and now - cigar butts, value stocks, moats, whatever - MO had been right there as an obvious stellar buy, and for ethical reasons Berkshire had deliberately and steadily foregone investing in it.

I think it's not JUST ethical reasons. I think WEB and Munger prefer to invest in things that provide positive results, for the investment AND for the world. So, for example, they would never invest in gold or in cryptocurrencies, because there is no net positive from those, they are simply mechanisms to transfer money from one pocket to another, with nothing real earned. Remember a few decades ago in WEBs letter the parable about the big block of all the world's gold on one side, and a bunch of large companies on the other side that produced very useful stuff and produced earnings every year. Now, some people would consider railroads (or utilities) to be polluting "evils" in the world, and while that may be true to some extent, there is a substantial net positive to the world in that they move goods to where they are needed (or provide power where needed, etc). Now, in today's world we know a little more about health and nutrition, so maybe candy has no net positive effect on the world, but people do enjoy it (even though there is no positive nutritional value whatsoever). Luckily the candy business is so small that it isn't of great concern to the overall ethos of Berkshire. And I think WEB is too old to change his ways, he will eat junk food and drink full-sugared Coke until he dies. Maybe the next guy (Able) will discard some businesses that no longer provide any true positive to the world.

Bigger isn't (statistically speaking) better

Maybe bigger IS statistically better. For example, how often does the S&P500 remove the bigger stock when compared to the smaller stocks? Maybe the 100 smallest stocks in the S&P500 do better than the 100 largest ones, not because the 100 smallest ones are better, but rather because the 100 smallest ones are culled more often?

Maybe bigger IS statistically better.

Well, it's a fair thing to question, but history and theory are quite clear: probably not.

The biggest 5 stocks by market cap are likely to be quite bad bets on average through the decades based on experience. For example, in the 31 years 1986-2016 inclusive, a portfolio of the biggest 5 by market cap underperformed the S&P 500 (of which they form a large part!) by 3.53%/year in total return. That's a pretty big observation.

And also on theory: if all stocks vary in valuation level between over and undervalued over time, the order of market cap gets shuffled somewhat away from the order you'd get on pure absolute intrinsic value. Undervalued things slide down the list, overvalued things move up while it lasts. The ones at the top by market cap will be, statistically, ones that got shuffled upwards in the list order due to transient overvaluation.

So, other than recent results in the last decade, I personally wouldn't want to bet the other way because I'm more the type to go with theory and history. Of course, maybe corporate concentration will soar some more and the biggest will just keep getting bigger, even faster than the small get big. We'll all work for the surveillance capitalists in another few years--who knows? But barring that, which has a pinch of the "it's different this time" reasoning to it, there is no reason to think the biggest are a better bet than the others, and in fact some pretty good reasons to think they're worse.

Jim

And also on theory: if all stocks vary in valuation level between over and undervalued over time, the order of market cap gets shuffled somewhat away from the order you'd get on pure absolute intrinsic value. Undervalued things slide down the list, overvalued things move up while it lasts. The ones at the top by market cap will be, statistically, ones that got shuffled upwards in the list order due to transient overvaluation.

I don't think I made my point clear enough.

Let's start here. The order of market cap changes for a few reasons:
1. Some move up due to overvaluation.
2. Some move up due to higher earnings.
3. Some move down due to undervaluation.
4. Some move down due to lower earnings.
And likely all move around due to all 4 of the above.

But my point ISN'T about an index, let's say the S&P500, but more about the composition of the index. So, if you invest in the index, or perhaps a derivation of the index, maybe S&P500 minus the top 5 ("Next 495"), or maybe S&P500 minus the bottom 495 ("behemoth S&P"), or maybe equal weighted ("RSP"), or whatever. But when we say "small stocks outperform large stocks", we are referring to company size out of the entire universe of stocks, not just one of the indexes. If you look at "small stocks in an index", you have to account for the survivors bias. The ones that went bust, or close to bust, were removed from the index continuously so if you are looking only at index stocks then those are gone and you "lose" the data of counting them (as negative contributors to the overall return of "small stocks"). If you were to look at the TOTAL performance of all small stocks over the 31 years versus all large stocks over the 31 years, perhaps the large stock performance would look a little better than the small stocks ... mainly because a substantial percentage of the small stocks disappeared over the period (and essentially went to zero or were acquired at some level that stops its compounding at that point). I don't have sufficient data to say one way or another, and it remains a thought experiment for me. If anyone has data, and the wherewithal to massage it, it might be interesting to see the results and the methodology.

If you were to look at the TOTAL performance of all small stocks over the 31 years versus all large stocks over the 31 years, perhaps the large stock performance would look a little better than the small stocks ... mainly because a substantial percentage of the small stocks disappeared over the period (and essentially went to zero or were acquired at some level that stops its compounding at that point). I don't have sufficient data to say one way or another, and it remains a thought experiment for me. If anyone has data, and the wherewithal to massage it, it might be interesting to see the results and the methodology.

This surprisingly hard to test, even when you have the data : )

The reason is that you can't simply (say) divide stocks into deciles by market cap and see how each decile did on average over time. Each interval that you reconstitute, some of the stocks will have moved between bands, which causes them no longer to be counted in the same band...their move counts against a different band in the next interval. If those moves are caused by mean reverting random noise, you'll be pumping non-existent returns from large caps to small caps. (a small stock does well and that performance counted in its small cap band; it's promoted to the next band in the next period; it falls again, and that fall is counted against the higher band, repeat...)

If there is only one division, between small caps and large caps, that is much less of a problem, as there is only one boundary being crossed. You can add a "slush" band as well, I think some index providers do this: a small crossing of the dividing line isn't acted upon.

In the end, there isn't really much of a small cap effect to speak of, and to a large extent there never was. Mainly there is an illiquidity premium, compounded by the fact that illiquid stocks and very old prices have slightly more errors than liquid stocks and more recent periods. If some easily-investable small cap group clearly outperformed (say) some large cap group over very long periods, it would get bid up to prices that the effect disappeared. Instead, we see short and long cycles of one or the other being in the lead as fashions change.

A much more interesting study is to look at the differences in returns based on share turnover ratio (ratio of average daily volume to shares outstanding). But that's for another day.

Jim

But barring that, which has a pinch of the "it's different this time" reasoning to it, there is no reason to think the biggest are a better bet than the others, and in fact some pretty good reasons to think they're worse.

Could it be different this time? If you think of the modern economy as a sort of hybrid between humans and machines, as productivity has increased "obviously" (really?) the machine fraction of that hybrid has been rising. AI suggests we may get a step function increase in that effect as so many tech-bros are talking about a world in which we don't have to work, where the guaranteed minimum income will actually be pretty high. But even without the step function of AI, could the mix of machine and man have gotten so machine dependent that some large techie companies just own better tech to hybridize their humans with, and it has become an enduring advantage?

To paraphrase clichéed investing wisdom, everything is always the same as it used to be, until it is different. Even the moats around newspapers is now filled with old rusting printing presses. Even IBM is an also ran. Even the stocks of widows and orphans are also-rans.

I have been contemplating selling my SP Growth ETF which has done GREAT over the last few years, and buying Equal-Weight SP which has underperformed regular SP by 35% the last few years (last 20 years actually, but moreso the last few). On the other hand, what if regression to the mean is just a thing that happens except when it doesn't?

Has it never been different this time in the past?

R:

Could it be different this time?

Sure. The super-gigacaps might not be as bad a bet as they were in the past.
But I think the better question, is "how different"?

For example, I wouldn't be particularly surprised to see very largest firms have only half the drag in an average future year (i.e., compared to the average large cap) that they had in the past. That would definitely count as things being different this time, but it still wouldn't suggest them as the smart bet on average in future.

I wouldn't apply any such sanguine attitude to a starting point like today, which seems a bit of an outlier in a few different ways. Fine firms, but US tech stocks currently make up over 19% of the aggregate market cap of all global equities. Is that really the right number? Really?

... and buying Equal-Weight SP which has underperformed regular SP by 35% the last few years

I think the best way to think of it is this way: RSP tracks the set of big companies. SPY is that return, plus or minus in any period a huge random number depending on the current unpredictable stock price results of a very few super-gigacaps. When you appreciate it that way, you realize that extrapolating recent relative results isn't a good idea...the emphasis is on "random number". RSP can be extrapolated a whole lot more than SPY can be, since it doesn't depend so much on the fate of a tiny few firms. It's more of a macro thing.

Jim

Post New

Berkshire Hathaway FAQ

---

Contact Shrewd'm
Contact the developer of these message boards.