Author Topic: Equal weight versus market cap weighted indexes -investment heretics thread  (Read 1264 times)

Buffaloski Boris

  • Handlebar Stache
  • *****
  • Posts: 2121
The majority of index funds are market capitalization weighted. This means that when you invest in an index fund youíre getting a larger share of the biggest companies (by market capitalization) in the index and a smaller share of the smaller companies in the index. So when you take a look at a popular market capitalization weighted index funds like VFINX (S+P 500), over 25% of the underlying holdings are concentrated in the top 10 largest stocks. Even the ever popular VTSAX is over 21% allocated to those top 10 stocks. In my view, thatís a problem: isnít the idea of owning a bunch of different stocks in an index to increase your overall diversification? It also means that youíre engaging in de facto momentum investing which may or may not be to your advantage.

There are index funds out there that do not rely on market capitalization weighting. There are equal weight indexes as well as those that allocate weights based on other algorithms. Hereís the interesting thing: the equal weight indexes can provide higher returns over time. A couple of articles:

https://www.morningstar.com/articles/990965/how-simple-rules-can-beat-the-market

Older but still interesting:
https://www.marketwatch.com/story/long-term-investors-can-beat-the-sp-500-by-favoring-equal-weighted-etfs-2018-09-26?mod=mw_quote_news

waltworks

  • Walrus Stache
  • *******
  • Posts: 5021
Is the Vanguard ETF really doing that, though? It's been horribly outperformed in the last decade/since inception by VTI (7% to 14%!), to the point that it would have an insanely hard time making up that ground in a crash.

Isn't that one just a cap-weighted index as well, but one that is buying value stocks?

-W

Buffaloski Boris

  • Handlebar Stache
  • *****
  • Posts: 2121
Is the Vanguard ETF really doing that, though? It's been horribly outperformed in the last decade/since inception by VTI (7% to 14%!), to the point that it would have an insanely hard time making up that ground in a crash.

Isn't that one just a cap-weighted index as well, but one that is buying value stocks?

-W
VTI is 21.72% in its top 10 holdings per Yahoo finance. Same percentage as VTSAX.

waltworks

  • Walrus Stache
  • *******
  • Posts: 5021
What?

I was referring to the performance of VTV (value ETF mentioned in the article) against VTI.

-W

Buffaloski Boris

  • Handlebar Stache
  • *****
  • Posts: 2121
What?

I was referring to the performance of VTV (value ETF mentioned in the article) against VTI.

-W
OK. I get it now. The author went down a bit of a rabbit hole in my view. Some background helps.  There is only one group of ETFs, mutual funds that do equal weighting that Iím aware of : invesco. The management fee is .40 for the SP 500 equal weight ETF. Significantly higher than what Vanguard charges for a comparable index fund ETF.

The author seems to be saying that since the advantage of an equal weight fund can be explained solely by the value and small cap tilt, that you can buy a value fund with a lower fee and mimic it. Fair enough. I disagree. How much of this value fund do you buy to mimic the value aspect? Value obviously hasnít done so well over the last 10 years. It may well do better in the future.  And weíre talking a grand total of 40 basis points in fees. I donít see much point in buying multiple ETFs to save 30 to 35 basis points.
« Last Edit: July 19, 2020, 11:02:58 AM by Buffaloski Boris »

ctuser1

  • Handlebar Stache
  • *****
  • Posts: 1741
SP500 is NOT capitalization weighted.

It is cap weighted and float adjusted.
Link from 2004 when this change was afoot: https://www.marketwatch.com/story/sp-move-to-float-adjusted-indexes-will-create-turnover

What does it mean? Capitalization that is closely held (by founders, e.g.) and not traded on the market, is not counted.

If you want to "own the market", then "cap weighted and float adjusted" is as close to it as it gets. Anything else is not representative of "the market".

e.g. equal weight, it is likely an ok idea at very small sizes. As soon as it becomes big, the tracking error will become horrendous and will mess with the markets. consider the scenario were 10% of the market is equal weight. The 500th component of SP500 does not likely have shares outstanding = (total market) X 10% X 1/500. Even if it did, that will result in horrible market distortion if the ETF held that big a chunk. In reality, market distortions will show up at MUCH smaller sizes due to how marginal transactions at any point of time work.

Any index that is to be held by the general public should be "float" weighted - else the cost to the market due to the index itself will outweigh it's positive economic value.

« Last Edit: July 19, 2020, 10:54:47 AM by ctuser1 »

Panly

  • Stubble
  • **
  • Posts: 132

Joel Greenblatt wrote a book about it, it's worth reading.




bacchi

  • Walrus Stache
  • *******
  • Posts: 5938
Over 17 years, RSP has done better than SPY/VTI (306% vs 273%). VTI has outpaced it in the last year however.

I expect the author is correct. It gets a lot of its out-performance from value stocks and, currently, growth stocks are the market leaders. Value stocks will come around again, as will emerging and international.

ctuser1

  • Handlebar Stache
  • *****
  • Posts: 1741
Joel Greenblatt wrote a book about it, it's worth reading.

A book is fine, but his fund did not do well using this touted strategy.

Boglehead thread: https://www.bogleheads.org/forum/viewtopic.php?t=193613
Forbes article: https://www.forbes.com/sites/antoinegara/2015/07/31/value-guru-joel-greenblatt-gotham-performance-market-magic-formula/#54294126142b

Quote from the Forbes article:
Quote
This year, all of Greenblatt's mutual funds have suffered losses, with his biggest funds down between 6%-to-10%, and none of his funds have outperformed the S&P 500 Index's total return since their inception.

Search for the ticker of his fund (FVVAX, found in the seeking alpha article from 2011: https://seekingalpha.com/article/263344-joel-greenblatts-big-secret-value-weighted-indexing), and it is nowhere to be found.

Let's just say his strategy looked great on paper but did not survive the first contact with the markets.

jim555

  • Magnum Stache
  • ******
  • Posts: 2726
Equal weight has seriously under performed.  This market is only about 10 stocks driving the whole thing.

Buffaloski Boris

  • Handlebar Stache
  • *****
  • Posts: 2121
Over 17 years, RSP has done better than SPY/VTI (306% vs 273%). VTI has outpaced it in the last year however.

I expect the author is correct. It gets a lot of its out-performance from value stocks and, currently, growth stocks are the market leaders. Value stocks will come around again, as will emerging and international.

Iím curious as to where you find your cumulative returns data on ETFs.

Iíve been listening to a lot of folks on value and it seems to be cyclical.

bacchi

  • Walrus Stache
  • *******
  • Posts: 5938
Over 17 years, RSP has done better than SPY/VTI (306% vs 273%). VTI has outpaced it in the last year however.

I expect the author is correct. It gets a lot of its out-performance from value stocks and, currently, growth stocks are the market leaders. Value stocks will come around again, as will emerging and international.

Iím curious as to where you find your cumulative returns data on ETFs.

Iíve been listening to a lot of folks on value and it seems to be cyclical.

I used Yahoo! finance. Google finance has a smaller return with RSP still ahead. Yahoo! does include reinvested dividends but I'm not sure that Google does.

Morningstar has even different numbers with the same results (RSP slightly ahead over 17 years).

https://www.morningstar.com/etfs/arcx/rsp/quote


Buffaloski Boris

  • Handlebar Stache
  • *****
  • Posts: 2121
Over 17 years, RSP has done better than SPY/VTI (306% vs 273%). VTI has outpaced it in the last year however.

I expect the author is correct. It gets a lot of its out-performance from value stocks and, currently, growth stocks are the market leaders. Value stocks will come around again, as will emerging and international.

Iím curious as to where you find your cumulative returns data on ETFs.

Iíve been listening to a lot of folks on value and it seems to be cyclical.

I used Yahoo! finance. Google finance has a smaller return with RSP still ahead. Yahoo! does include reinvested dividends but I'm not sure that Google does.

Morningstar has even different numbers with the same results (RSP slightly ahead over 17 years).

https://www.morningstar.com/etfs/arcx/rsp/quote
Thanks! Something I noticed was I got the expense ratio wrong. Itís .20% management fee. Compared to .03% for VTI.

Panly

  • Stubble
  • **
  • Posts: 132
Joel Greenblatt wrote a book about it, it's worth reading.

A book is fine, but his fund did not do well using this touted strategy.

Boglehead thread: https://www.bogleheads.org/forum/viewtopic.php?t=193613
Forbes article: https://www.forbes.com/sites/antoinegara/2015/07/31/value-guru-joel-greenblatt-gotham-performance-market-magic-formula/#54294126142b

Quote from the Forbes article:
Quote
This year, all of Greenblatt's mutual funds have suffered losses, with his biggest funds down between 6%-to-10%, and none of his funds have outperformed the S&P 500 Index's total return since their inception.

Search for the ticker of his fund (FVVAX, found in the seeking alpha article from 2011: https://seekingalpha.com/article/263344-joel-greenblatts-big-secret-value-weighted-indexing), and it is nowhere to be found.

Let's just say his strategy looked great on paper but did not survive the first contact with the markets.


just don't ever read the book,  keep exposing your ignorance.   Be proud of it.




   



ctuser1

  • Handlebar Stache
  • *****
  • Posts: 1741
Joel Greenblatt wrote a book about it, it's worth reading.

A book is fine, but his fund did not do well using this touted strategy.

Boglehead thread: https://www.bogleheads.org/forum/viewtopic.php?t=193613
Forbes article: https://www.forbes.com/sites/antoinegara/2015/07/31/value-guru-joel-greenblatt-gotham-performance-market-magic-formula/#54294126142b

Quote from the Forbes article:
Quote
This year, all of Greenblatt's mutual funds have suffered losses, with his biggest funds down between 6%-to-10%, and none of his funds have outperformed the S&P 500 Index's total return since their inception.

Search for the ticker of his fund (FVVAX, found in the seeking alpha article from 2011: https://seekingalpha.com/article/263344-joel-greenblatts-big-secret-value-weighted-indexing), and it is nowhere to be found.

Let's just say his strategy looked great on paper but did not survive the first contact with the markets.


just don't ever read the book,  keep exposing your ignorance.   Be proud of it.


I am keenly aware that I am ignorant of a lot of important stuff. Is "i-know-the-secret-sauce" Greenblatt? and his lemmings e.g. you?

Are you proud of being a walking-talking demonstration of the Dunning-Kruger effect?

MustacheAndaHalf

  • Magnum Stache
  • ******
  • Posts: 4425
Panly - Someone showed that the book's author didn't do well after publication, which is a key test of a theory.


Buffaloski Boris - You're saying the 3 factor model (market/small-cap/value) doesn't explain equal weight returns... does that mean you propose an additional factor?

On Yahoo Finance, 5 year performance gets worse as you go from the S&P 500 to a small-cap value ETF.  Both value and small-cap have trailed over 5 years, and year to date.  Now, that's not the whole story - I actually think some small cap value stocks are waiting for a recovery, at which point they'll do well.  But in the meantime, tech stocks are mostly large cap growth.

When I checked a month ago, most of the S&P 500 return came from the biggest 5 stocks.  If their weight goes from 4% of the S&P 500 to just 0.2%, that has a big impact - and that's what an equal weight S&P 500 fund would do.  That's my area of greatest concern for equal weight ETFs.

Buffaloski Boris

  • Handlebar Stache
  • *****
  • Posts: 2121
Panly - Someone showed that the book's author didn't do well after publication, which is a key test of a theory.


Buffaloski Boris - You're saying the 3 factor model (market/small-cap/value) doesn't explain equal weight returns... does that mean you propose an additional factor?

On Yahoo Finance, 5 year performance gets worse as you go from the S&P 500 to a small-cap value ETF.  Both value and small-cap have trailed over 5 years, and year to date.  Now, that's not the whole story - I actually think some small cap value stocks are waiting for a recovery, at which point they'll do well.  But in the meantime, tech stocks are mostly large cap growth.

When I checked a month ago, most of the S&P 500 return came from the biggest 5 stocks.  If their weight goes from 4% of the S&P 500 to just 0.2%, that has a big impact - and that's what an equal weight S&P 500 fund would do.  That's my area of greatest concern for equal weight ETFs.

I actually agree with the author in that the small(er) cap and value tilt explains a lot if not all of the difference so I wouldnít propose any added factors. What I didnít agree with was what appears to me to be a logical jump of saying ďsome value good, more value betterĒ and by the way you can get value cheaper using VTI. It seems to me that a more moderate approach of skewing a bit towards value and small(er) cap is something of a sweet spot. And the difference in fees is roughly 17 basis points. Small potatoes in the grand scheme.

As for reducing exposure to the largest constituents of the SP 500, to me thatís a feature, not a bug.

MustacheAndaHalf

  • Magnum Stache
  • ******
  • Posts: 4425
If expense ratio is a concern, Vanguard Value ETF has an 0.04% expense ratio (versus VTI with 0.03%).

Small cap value is more expensive - I own an ETF right now with a 0.35% expense ratio.  Since small cap and value have both been hit harder in Covid-19, I bought in order to wait for a recovery.  I'm counting on reversion to the mean at some point - that works for equal weight portfolios, too.  If they trail the market now, you might expect some catching up after Covid-19 is resolved (in months?  years?).

Equal weight avoids the big tech companies, which you identify as a feature.  Since I've been focused on deeply discounted stocks, I've been on the fence about tech for now.  What do you say to people like me who are on the fence about market weight - or even overweight of big tech stocks?  What convinced you to avoid them?

Buffaloski Boris

  • Handlebar Stache
  • *****
  • Posts: 2121
If expense ratio is a concern, Vanguard Value ETF has an 0.04% expense ratio (versus VTI with 0.03%).

Small cap value is more expensive - I own an ETF right now with a 0.35% expense ratio.  Since small cap and value have both been hit harder in Covid-19, I bought in order to wait for a recovery.  I'm counting on reversion to the mean at some point - that works for equal weight portfolios, too.  If they trail the market now, you might expect some catching up after Covid-19 is resolved (in months?  years?).

Equal weight avoids the big tech companies, which you identify as a feature.  Since I've been focused on deeply discounted stocks, I've been on the fence about tech for now.  What do you say to people like me who are on the fence about market weight - or even overweight of big tech stocks?  What convinced you to avoid them?
What Iíd say:
-you shouldnít take what I say as investment advice.
-since none of us have a working crystal ball that Iím aware of, the best thing to do is diversify. Putting disproportionate amounts into indices that are skewed toward a handful of companies is not optimal. Worse yet when lots and lots of other people are in those same handful of companies.
-reversion to the mean is a thing. By assuming that what has been so successful over the past few years is going to continue, arenít we really chasing returns and implying that This Time Will Be Different?
-people who want to own tech should own tech. Why invest indirectly in a bunch of other stuff to get tech exposure if tech is what you think will be a better strategy in the future? Iím definitely not convinced that tech is the way to go because of mean reversion, but I could be wrong. Interestingly, the second article indicated that equal weighting works within sectors as well.
-personal preference. I prefer owning a little of everything (within reason). I canít remember who said it, but I recall some credible investor  saying that about one out of 20 stocks deliver returns above treasuries. If thatís true, and we donít know which companies those are, wouldnít we want a more or less equal allocation to maximize that?
-A cap weighted weighted index is implicitly a momentum strategy. Great on the way up, not so great on the way down.

AdrianC

  • Handlebar Stache
  • *****
  • Posts: 1149
  • Location: Cincinnati
Joel Greenblatt wrote a book about it, it's worth reading.

A book is fine, but his fund did not do well using this touted strategy.

Boglehead thread: https://www.bogleheads.org/forum/viewtopic.php?t=193613
Forbes article: https://www.forbes.com/sites/antoinegara/2015/07/31/value-guru-joel-greenblatt-gotham-performance-market-magic-formula/#54294126142b

Quote from the Forbes article:
Quote
This year, all of Greenblatt's mutual funds have suffered losses, with his biggest funds down between 6%-to-10%, and none of his funds have outperformed the S&P 500 Index's total return since their inception.

Search for the ticker of his fund (FVVAX, found in the seeking alpha article from 2011: https://seekingalpha.com/article/263344-joel-greenblatts-big-secret-value-weighted-indexing), and it is nowhere to be found.

Let's just say his strategy looked great on paper but did not survive the first contact with the markets.

His later fund (GSPFX) is more along the lines of the book (Big Secret). Those earlier funds were 'Magic Formula'. That didn't work in mutual fund form.

GSPFX Gotham Enhanced S&P 500 Index - Value-weighted S&P 500

https://www.gothamfunds.com/Funds.aspx?FundID=136

3-year 12.78% compared to 10.98% for VTSAX/VTI

Greenblatt does not define value the way that Vanguard, et al, do.

https://funds.gotham.com/Download.aspx?ID=4745663f-08f5-43ec-ae0c-f1e5dba7a775&Inline=1

WSJ: Value investing means different things to different people. What does it mean to you?
Mr. Greenblatt: It does not mean low price-to-book-value, low price-to-sales ratio investing, which is how most people define it. Stocks aren't pieces of paper that bounce around, they're ownership shares of businesses that we value and try buy at a discount.

Makes sense to me.

MustacheAndaHalf

  • Magnum Stache
  • ******
  • Posts: 4425
AdrianC - I think a fund with "S&P 500" in it's name should be compared against the S&P 500.  In Google Finance when I compare GSPFX to VOO (Vanguard S&P 500 ETF), VOO is up +42% while GSPFX is up +22%.


MustacheAndaHalf

  • Magnum Stache
  • ******
  • Posts: 4425
Buffaloski Boris - I have about a thread's worth of comments to just that post of yours... so let me pick my top 2 and start there.  :)

I think momentum works.  In my passive portfolio, I hold core index funds and tilt to momentum.  The "3 factor" model of the market (market, small/big, value/growth) is improved by adding momentum as a 4th factor.  There's white papers and books that would do a better job of justifying it than I could, especially since it's been a few years since I studied it.

I agree with reversion to the mean - I bought stocks like DIN, M, DXPE counting on them at some point either recovering or going bankrupt.  That could be true of tech stocks... but "technology" does not mean revert.  People do not willingly give up their cell phones, so I believe technology has a permanent place in the stock market.  Twenty years ago, how many people checked their cell phones when waiting in line, or on public transport?  And now?  We search with Google, check on friends on Facebook, and buy products online with Amazon.  In my view, there's a good reason those 3 companies are at the top of the S&P 500, and they could continue being there.  A tech correction is possible, but their products are definitely ingrained in daily life now, suggesting they have a long-term future.

AdrianC

  • Handlebar Stache
  • *****
  • Posts: 1149
  • Location: Cincinnati
AdrianC - I think a fund with "S&P 500" in it's name should be compared against the S&P 500.  In Google Finance when I compare GSPFX to VOO (Vanguard S&P 500 ETF), VOO is up +42% while GSPFX is up +22%.

The performance difference between VTI and VOO is miniscule...but OK.

https://www.morningstar.com/etfs/arcx/voo/performance
3 year VOO 11.68%

https://www.morningstar.com/funds/xnas/gspfx/performance
3 year GSPFX 12.94%

https://www.gothamfunds.com/Funds.aspx?FundID=136#
Since Inception Cumulative
Gotham Enhanced S&P 500 Index Fund 51.64%
S&P 500 48.45%

Google finance shows price data only, and is not showing total return. GSPFX trades a lot and has distributions.