Equal Weighted Funds - Better?

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply
Topic Author
Momus
Posts: 1023
Joined: Tue Feb 21, 2012 8:23 pm

Equal Weighted Funds - Better?

Post by Momus »

Is this really better than a regular cap weighted index fund? Should I switch over and dump my Vanguards?

http://www.marketwatch.com/story/ignore ... -10?page=2
User avatar
nisiprius
Advisory Board
Posts: 52211
Joined: Thu Jul 26, 2007 9:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: Equal Weighted Funds - Better?

Post by nisiprius »

Notice what word is missing in this article?

"Risk."

Also "Risk-adjusted reward." Also "Sharpe ratio.""

All the article talks about is return. There's no trick to getting higher return if you are willing to take more risk.

If you look at Morningstar's analysis you will see that RSP--the equal-weighted S&P 500 fund he talks about--Morningstar says its risk is "above average" for the last 3 and 5 years, "high" for 10 years and overall. For the Vanguard 500 index fund, the risk rating is "below average" for 3 years, "average" for 5, 10, and overall.

Image

RSP is a riskier fund than Vanguard 500 Index Fund (VFINX).

Now if we look at a measure of risk-adjusted reward, the Sharpe ratio, we see that the 10-year Sharpe ratio for RSP is 0.51, for the Vanguard 500 Index fund, 0.49. That is virtually identical, and it means that the superior performance of RSP is almost completely explained by its having additional risk. It no better and no worse than the Vanguard 500 Index Fund, it just has more smaller and riskier stocks and concomitantly higher return.

More risk (see the "standard deviation" numbers), more reward, same risk-adjusted reward. No magic here.

Image
Image
Annual income twenty pounds, annual expenditure nineteen nineteen and six, result happiness; Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery.
Random Walker
Posts: 5561
Joined: Fri Feb 23, 2007 7:21 pm

Re: Equal Weighted Funds - Better?

Post by Random Walker »

I think equal weighting just creates a value tilt compared to cap weighting.

Dave
lack_ey
Posts: 6701
Joined: Wed Nov 19, 2014 10:55 pm

Re: Equal Weighted Funds - Better?

Post by lack_ey »

Once you're done looking at the risk-adjusted return, you have to wonder how much of what happened during one 10-year period might happen again in the future. The greater risk is probably persistent at least to some degree. The return may be, just maybe or probably not to the same extent.

To me, the higher expense ratio is a turnoff and the clear deciding factor. If unsure about an investment, I'm not all that enthusiastic about paying extra for it.
pascalwager
Posts: 2327
Joined: Mon Oct 31, 2011 8:36 pm

Re: Equal Weighted Funds - Better?

Post by pascalwager »

I think equal weighting just creates a value tilt compared to cap weighting.
I looked at the M* style boxes for RSP (S&P 500 equal-weighting) and SPY (S&P 500).

RSP has 37% value and 30% growth. SPY has 34% value and 33% growth.
RSP has 54% large and 46% medium. SPY has 88% large and 12% medium.

So I see a small value tilt and a larger size tilt. You can get these same tilts with less cost by using VG index funds.
VT 60% / VFSUX 20% / TIPS 20%
User avatar
asset_chaos
Posts: 2629
Joined: Tue Feb 27, 2007 5:13 pm
Location: Melbourne

Re: Equal Weighted Funds - Better?

Post by asset_chaos »

Momus wrote:Is [equal weighting] really better than a regular cap weighted index fund? Should I switch over and dump my Vanguards?
Equal-weighted is different, not better. It may have a higher expected return, but if it does, it would come with the assumption of higher risk, as Nisiprius has already shown. If you understand the extra risk, and wish to bear it in order to possibly accumulate a bit more absolute return, and know in your heart of hearts that you'll stick with this investing strategy through thick and thin for the next 40 years, then, by all means, learn more about it. Note that equal weight is typically more expensive than cap-weighted; it's also typically more expensive than other, more direct, ways to bear the same risks and possibly earn the same long-term risk premium. Since you ask, I suggest you not dump your cap-weighted index funds just yet: cogitate a bit more on this idea.
Regards, | | Guy
lack_ey
Posts: 6701
Joined: Wed Nov 19, 2014 10:55 pm

Re: Equal Weighted Funds - Better?

Post by lack_ey »

pascalwager wrote:
I think equal weighting just creates a value tilt compared to cap weighting.
I looked at the M* style boxes for RSP (S&P 500 equal-weighting) and SPY (S&P 500).

RSP has 37% value and 30% growth. SPY has 34% value and 33% growth.
RSP has 54% large and 46% medium. SPY has 88% large and 12% medium.

So I see a small value tilt and a larger size tilt. You can get these same tilts with less cost by using VG index funds.
I seem to also recall there being a rebalancing bonus observed in the last ten years, so it's not just the factor tilts, and you wouldn't quite get the same rebalancing on your own. Again, none of this is necessarily expected to increase returns in the future.
User avatar
nisiprius
Advisory Board
Posts: 52211
Joined: Thu Jul 26, 2007 9:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: Equal Weighted Funds - Better?

Post by nisiprius »

Two random bits of "spin." Warning: propaganda ahead.

1) The first attempt to create an index portfolio was equal-weighted--and failed.

The first attempt to create an index portfolio was an equal weighted portfolio. It was created in 1971 as an internal retirement account for the Samsonite. It was to hold "an equal dollar amount of each of the 1,500 or so stocks listed on the NYSE." It was a disaster due to trading costs and the company moved to cap-weighting in 1976. (This fund antedated the Vanguard 500 index fund (launched in 1976) is the origin of a squabble about what was "the first index fund.")

OK, thanks to automation and much lower trading cost, fund companies can do things in 2014 that weren't practical in 1971, but still.

2) Equal-weighting is illogical.

If it makes sense to invest in an equal-weighted S&P 500, why doesn't it make seven times as much sense to invest in an equal-weighted Total Stock Market index? There are 3,804 stocks in the total market index used by VTSMX. If you are equal-weighting, then the equal-weighted S&P 500 only includes less than 13% of the equal-weighted total market, and you are missing 87% of what you'd be getting in an equal-weighted total market portfolio. On the theory that you want equal weighting, how does it make sense to miss out on 87% of the market?

How do you decide when to stop?

With cap-weighting, 500 is an arbitrary round number but it is not totally arbitrary because it represents most of the dollars in the market. With equal-weighting, 500 is bizarrely arbitrary.

If you are cap-weighting, then when you invest in the S&P 500 you are only missing out on about 20% of the dollars in the stock market. Maybe missing out on 20% is a good reason not to stop with the S&P 500, but even if you do, you are still getting 80% of the market and the difference between it and the total market, if you check a growth chart, is surprisingly small. More important, if you choose to cap-weight, it is completely practical to cover 98% (Russell 3000 and other "broad" market indexes) or very very close to 100% (CRSP and other "total" market indexes) and it is being done by many good index funds and ETFs from many providers.

If it is logical to equal-weight, then I can't think of any logical reason not to equal-weight the whole market, and I am sure the reason why RSP only covers the S&P 500 is that it is utterly impractical to do much more than that (because there aren't enough small-cap stocks to go around). It might even be impossible under mutual fund diversification restrictions. [Added] And I surely was wrong. See posting below by Angst.
Last edited by nisiprius on Sun Dec 28, 2014 9:09 am, edited 1 time in total.
Annual income twenty pounds, annual expenditure nineteen nineteen and six, result happiness; Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery.
User avatar
nisiprius
Advisory Board
Posts: 52211
Joined: Thu Jul 26, 2007 9:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: Equal Weighted Funds - Better?

Post by nisiprius »

RSP?

Insert sophomoric joke here. :P
(Yes, stupider than the Vardy article. But only a little stupider).

Image
Annual income twenty pounds, annual expenditure nineteen nineteen and six, result happiness; Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery.
Angst
Posts: 2968
Joined: Sat Jun 09, 2007 11:31 am

Re: Equal Weighted Funds - Better?

Post by Angst »

nisiprius wrote:If it is logical to equal-weight, then I can't think of any logical reason not to equal-weight the whole market, and I am sure the reason why RSP only covers the S&P 500 is that it is utterly impractical to do much more than that (because there aren't enough small-cap stocks to go around). It might even be impossible under mutual fund diversification restrictions.
No worries, Nisi!
Just combine 33.3% of EWRI with 66.7% of EWRS and you're covered:

EWRI Guggenheim Russell 1000® Equal Weight ETF
EWRS Guggenheim Russell 2000® Equal Weight ETF
User avatar
nisiprius
Advisory Board
Posts: 52211
Joined: Thu Jul 26, 2007 9:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: Equal Weighted Funds - Better?

Post by nisiprius »

Angst wrote:
nisiprius wrote:If it is logical to equal-weight, then I can't think of any logical reason not to equal-weight the whole market, and I am sure the reason why RSP only covers the S&P 500 is that it is utterly impractical to do much more than that (because there aren't enough small-cap stocks to go around). It might even be impossible under mutual fund diversification restrictions.
No worries, Nisi!
Just combine 33.3% of EWRI with 66.7% of EWRS and you're covered:

EWRI Guggenheim Russell 1000® Equal Weight ETF
EWRS Guggenheim Russell 2000® Equal Weight ETF
:oops: You learn something new every day.

So why aren't the equal-weighting fans recommending these?

(Cue Dana Carvey "Church Lady" voice and echo chamber effect, and roll the "R") Could it be.... RRRRRRRECENCY? :twisted:

Charting the money market fund first, blue line, in order to make Morningstar charting total return growth charts for ETFs...

Since inception, the equal-weighted Russell 2000 fund, EWRS, has, since inception, underperformed Vanguard's cap-weighted Russell 2000 fund, VTWO.
And the equal-weighted Russell 1000 fund, EWRI, has, since inception, basically tied (tiny underperformance) Vanguard's cap-weighted Russell 1000 fund, VONE.

So, the equal-weighting magic wins once (S&P 500), ties once (Russell 1000), and loses once (Russell 2000).

What is the supposed explanation of why it wouldn't work all the time? Could it be a case of a strategy that sometimes outperforms cap-weighting and sometimes doesn't, and people only mention the times that it does?

Source: Morningstar

Image

Image
Annual income twenty pounds, annual expenditure nineteen nineteen and six, result happiness; Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery.
TradingPlaces
Posts: 1245
Joined: Sun Nov 09, 2014 12:19 pm
Location: 30.286029, -97.530011

Re: Equal Weighted Funds - Better?

Post by TradingPlaces »

I am a very STRONG proponent of equal-ish weighted index. However, in practice, this is not very easy to implement:

- cap-weighting has the beautiful feature that no rebalancing is required stay true to the index,
- equal-weighting requires frequent rebalancing, and one of the most important issues is to figure out the correct rebalance frequency.

A lot of mid-frequency statistical hedge funds trade daily, or intraday, and harvest the intraday to 1-day mean reversion. They have a lot of technology, and use shorter-term signals to reduce their transaction costs.

Thus, doing frequent rebalancing will simply put you against these types of hedge funds.

Rebalancing on something like weekly or monthly creates additional issues: do you have reversion or momentum at those frequencies?

However, assuming someone has solved this problems, it then brings up the following question:

- is RSP better, on a risk-adjusted basis, than whatever benchmark you choose?

I would argue that if RSP has 0.02 higher Sharpe than your benchmark, and that number is statistically significant, then it is a huge win for RSP. Why are bogle heads unwilling to admit small, but statistically improvements over the established index?

However, understand, that unlike SPY, RSP does not have the same capacity. As more people pile into RSP, performance will drag, because the rebalancing required to keep the equal-weighting will eventually increase the costs. On the other hand, cap-weighed index requires no rebalances.

Thus, as a large scale implementation idea, RSP is dead on arrival.
TradingPlaces
Posts: 1245
Joined: Sun Nov 09, 2014 12:19 pm
Location: 30.286029, -97.530011

Re: Equal Weighted Funds - Better?

Post by TradingPlaces »

nisiprius wrote:
Angst wrote:
nisiprius wrote:If it is logical to equal-weight, then I can't think of any logical reason not to equal-weight the whole market, and I am sure the reason why RSP only covers the S&P 500 is that it is utterly impractical to do much more than that (because there aren't enough small-cap stocks to go around). It might even be impossible under mutual fund diversification restrictions.
No worries, Nisi!
Just combine 33.3% of EWRI with 66.7% of EWRS and you're covered:

EWRI Guggenheim Russell 1000® Equal Weight ETF
EWRS Guggenheim Russell 2000® Equal Weight ETF
:oops: You learn something new every day.

So why aren't the equal-weighting fans recommending these?

(Cue Dana Carvey "Church Lady" voice and echo chamber effect, and roll the "R") Could it be.... RRRRRRRECENCY? :twisted:

Charting the money market fund first, blue line, in order to make Morningstar charting total return growth charts for ETFs...

Since inception, the equal-weighted Russell 2000 fund, EWRS, has, since inception, underperformed Vanguard's cap-weighted Russell 2000 fund, VTWO.
And the equal-weighted Russell 1000 fund, EWRI, has, since inception, basically tied (tiny underperformance) Vanguard's cap-weighted Russell 1000 fund, VONE.

So, the equal-weighting magic wins once (S&P 500), ties once (Russell 1000), and loses once (Russell 2000).

What is the supposed explanation of why it wouldn't work all the time? Could it be a case of a strategy that sometimes outperforms cap-weighting and sometimes doesn't, and people only mention the times that it does?

Source: Morningstar

Image

Image

I think it is pretty much useless to equal weight R2K. The reason is that the weights in R2K are much more balanced. Any equal-weighting simply creates unnecessary trading costs, and using stocks that are far less liquid, have higher trading costs, and more in quantity. At that point, unless you have close to zero trading costs, you are not winning the game.

A somewhat in-between idea (in-between an EW index and a cap-weight index) would be to have an SP500-like index, but with log-cap weights, and rebalance less frequently.
TradingPlaces
Posts: 1245
Joined: Sun Nov 09, 2014 12:19 pm
Location: 30.286029, -97.530011

Re: Equal Weighted Funds - Better?

Post by TradingPlaces »

nisiprius wrote:Two random bits of "spin." Warning: propaganda ahead.

1) The first attempt to create an index portfolio was equal-weighted--and failed.

The first attempt to create an index portfolio was an equal weighted portfolio. It was created in 1971 as an internal retirement account for the Samsonite. It was to hold "an equal dollar amount of each of the 1,500 or so stocks listed on the NYSE." It was a disaster due to trading costs and the company moved to cap-weighting in 1976. (This fund antedated the Vanguard 500 index fund (launched in 1976) is the origin of a squabble about what was "the first index fund.")

OK, thanks to automation and much lower trading cost, fund companies can do things in 2014 that weren't practical in 1971, but still.

2) Equal-weighting is illogical.
+1000

I fully agree with this argument. But every type of long-short strategy: from stock-picking, to statistical arbitrage to large-scale portfolio management, uses close-to equal weights, or log weights.

The ideas that come from long-short portfolio management translate poorly into long-only portfolio management, simply because they don't make sense. If you have $400B, like VTI does, it is virtually impossible to EW-invest in the top 6000 stocks of the US stock market? Why? Because the bottom 3000 stocks would get half the weights, and those 3000 stocks would get $200B if your asset allocation. But is the total market cap of the bottom 3000 even equal to that value? I doubt.

So there are rationally developed answers to these questions, when to stop:

- you ensure that your notional in a given stock does not exceed x% of the market cap or float of that stock, where x is a small number like 1-3,
- you ensure that your daily trading volume requirement for a given stock does not exceed y% of the stock's daily trading volume, where y is again s a small number like 0.5 to 3,
- you scale up your strategy in capitalization deciles.

Once you put these constraints, it becomes clear what the bounds to your strategy are.
Bfwolf
Posts: 2108
Joined: Thu Oct 14, 2010 11:19 am

Re: Equal Weighted Funds - Better?

Post by Bfwolf »

Equal weighting just doesn't make any sense.

So if you own a company and it splits in 2, you're now going to double your investment in the 2 companies it became?

Or if 2 companies merge, you're going to cut your investment in those companies by 50% for the newly formed company?

There's an old saying that if a newspaper headline asks a question, the answer is "no"....if the answer was yes, it would just state it as such instead of asking the question. That applies equally well with this thread. :D
TradingPlaces
Posts: 1245
Joined: Sun Nov 09, 2014 12:19 pm
Location: 30.286029, -97.530011

Re: Equal Weighted Funds - Better?

Post by TradingPlaces »

Random Walker wrote:I think equal weighting just creates a value tilt compared to cap weighting.

Dave
I don't see how this is the case. 250 of the stocks in SP500 are "value" and 250 are "not-value". In general, these are somewhat random, so the weight of value in SP500 is close to 50%.

But the weight of the 250 "value" stocks in the EW index is also close to 250.

But EW does create e small mkt-cap bias.

If you really wanted to remove that bias, you can simply do the EW strategy in the 5 100-stock quantiles:

- SP500 quantile 1, stocks 1 to 100, with total weight W1,
- SP500 quantile 2, stocks 101 to 200, with total weight W2,
- SP500 quantile 3, stocks 201 to 300, with total weight W3,
- SP500 quantile 4, stocks 301 to 400, with total weight W4,
- SP500 quantile 5, stocks 401 to 500, with total weight W5.

Each of this groups gets the same group weight as in SP500 cap-weight. So group 1, stocks 1 to 100, get the same weight W1 as in Sp500. However, within group, the weights are equal. In this manner, you have less bias to smaller caps.

I think at this point you are really stretching it, and the performance should be very similar to SP500.
User avatar
JoMoney
Posts: 16260
Joined: Tue Jul 23, 2013 5:31 am

Re: Equal Weighted Funds - Better?

Post by JoMoney »

Random Walker wrote:I think equal weighting just creates a value tilt compared to cap weighting.

Dave
You might be right, the average market-cap size of the S&P 500 "Growth" index is slightly larger than the S&P 500 Value index... but only slightly..
Since the equal weighting raises up the weighting of the smaller cap size this may also be adding slightly more "Value Style" stocks.
The bigger impact is that it's a trading strategy that emphasizes mean-reversion versus momentum. It will be constantly selling the stocks that are growing the fastest to buy more of the stocks falling the fastest. So in a way, equal weighting is anti-momentum and since momentum is generally a "growth" characteristic it probably is more aligned with value style trading.
"To achieve satisfactory investment results is easier than most people realize; to achieve superior results is harder than it looks." - Benjamin Graham
User avatar
LAlearning
Posts: 1365
Joined: Wed May 09, 2012 12:26 pm
Location: Los Angeles

Re: Equal Weighted Funds - Better?

Post by LAlearning »

Momus wrote:Is this really better than a regular cap weighted index fund? Should I switch over and dump my Vanguards?

http://www.marketwatch.com/story/ignore ... -10?page=2
Yes.
Let us know what happens in 10 years.
I know nothing!
Beliavsky
Posts: 1233
Joined: Sun Jun 29, 2014 10:21 am

Re: Equal Weighted Funds - Better?

Post by Beliavsky »

Have you considered fundamental indexing? I think it makes sense to invest more in companies that are large based on metrics such as earnings, dividends, sales, and book value. Equal weighting does not do this. Cap weighting does, on average, but a potential weakness of cap-weighted indexing is that it owns too much of the most expensive stocks. I own both cap-weighted and fundamentally weighted funds.
User avatar
Rick Ferri
Posts: 9707
Joined: Mon Feb 26, 2007 10:40 am
Location: Georgetown, TX. Twitter: @Rick_Ferri
Contact:

Re: Equal Weighted Funds - Better?

Post by Rick Ferri »

There is more risk in equal weight, therefore a higher expected return:

No Free Lunch From Equal Weight S&P 500

Rick Ferri
The Education of an Index Investor: born in darkness, finds indexing enlightenment, overcomplicates everything, embraces simplicity.
User avatar
nisiprius
Advisory Board
Posts: 52211
Joined: Thu Jul 26, 2007 9:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: Equal Weighted Funds - Better?

Post by nisiprius »

Beliavsky wrote:...a potential weakness of cap-weighted indexing is that it owns too much of the most expensive stocks...
That's not at all obvious to me. Why do you think that is true? For everyone screaming that Facevook is overvalued

Image


there is someone else screaming

Image
"Exxon Mobil Corporation, Currently Undervalued And A Strong Buy."

I can't bridge the chasm between those who are so darned sure that the stock of all great big famous blue chip companies must automatically be overvalued (hence the refrain "anything, anything but cap-weighted") and those who are so darned sure that you just can't go wrong with the great dividend stocks of the world's greatest companies. There is a big crowd on each side of the chasm.

P.S. And just to complete the tour:

This
Image
and this
Image
Annual income twenty pounds, annual expenditure nineteen nineteen and six, result happiness; Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery.
Beliavsky
Posts: 1233
Joined: Sun Jun 29, 2014 10:21 am

Re: Equal Weighted Funds - Better?

Post by Beliavsky »

nisiprius wrote:
Beliavsky wrote:...a potential weakness of cap-weighted indexing is that it owns too much of the most expensive stocks...
That's not at all obvious to me. Why do you think that is true?
IIRC, the paper "Fundamental Indexation" and book "The Fundamental Index: A Better Way to Invest" co-authored by Arnott and Hsu argue that stock prices equal fair value plus a noise term (the deviation of market value and fair value), and the noise term causes cap-weighted indices to overweight expensive stocks.
rkhusky
Posts: 17763
Joined: Thu Aug 18, 2011 8:09 pm

Re: Equal Weighted Funds - Better?

Post by rkhusky »

Bfwolf wrote:Equal weighting just doesn't make any sense.

So if you own a company and it splits in 2, you're now going to double your investment in the 2 companies it became?

Or if 2 companies merge, you're going to cut your investment in those companies by 50% for the newly formed company?
Exactly. Suppose further that there was a large corporation that owns a 100 different brands. If they split into a 100 different companies, with their own management teams, distribution networks, etc, would you really want to increase your stake in those brands a hundred-fold?
User avatar
nisiprius
Advisory Board
Posts: 52211
Joined: Thu Jul 26, 2007 9:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: Equal Weighted Funds - Better?

Post by nisiprius »

rkhusky wrote:
Bfwolf wrote:Equal weighting just doesn't make any sense.

So if you own a company and it splits in 2, you're now going to double your investment in the 2 companies it became?

Or if 2 companies merge, you're going to cut your investment in those companies by 50% for the newly formed company?
Exactly. Suppose further that there was a large corporation that owns a 100 different brands. If they split into a 100 different companies, with their own management teams, distribution networks, etc, would you really want to increase your stake in those brands a hundred-fold?
Now, if it were a cap-weighted index and that happened, you would find that immediately after the split, the total of the cap-weighted values of their stocks would be about the same as that of the company before. No problem, either logically or practically. (And I actually know about that personally, because I actually held AT&T stock when the Bell System broke up).
Annual income twenty pounds, annual expenditure nineteen nineteen and six, result happiness; Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery.
User avatar
JoMoney
Posts: 16260
Joined: Tue Jul 23, 2013 5:31 am

Re: Equal Weighted Funds - Better?

Post by JoMoney »

rkhusky wrote:
Bfwolf wrote:Equal weighting just doesn't make any sense.

So if you own a company and it splits in 2, you're now going to double your investment in the 2 companies it became?

Or if 2 companies merge, you're going to cut your investment in those companies by 50% for the newly formed company?
Exactly. Suppose further that there was a large corporation that owns a 100 different brands. If they split into a 100 different companies, with their own management teams, distribution networks, etc, would you really want to increase your stake in those brands a hundred-fold?
People who believe small-cap stocks outperform should love the idea of spin-offs and carve-outs. The relative market-cap size of the portfolio would keep getting smaller therefore the expected returns must be going up right? :P
As bizarre as it sounds, in a maybe-not-so-efficient market things get a bit wonky :
http://www.nber.org/papers/w8302
"To achieve satisfactory investment results is easier than most people realize; to achieve superior results is harder than it looks." - Benjamin Graham
Beliavsky
Posts: 1233
Joined: Sun Jun 29, 2014 10:21 am

Re: Equal Weighted Funds - Better?

Post by Beliavsky »

nisiprius wrote:
rkhusky wrote:
Bfwolf wrote:Equal weighting just doesn't make any sense.

So if you own a company and it splits in 2, you're now going to double your investment in the 2 companies it became?

Or if 2 companies merge, you're going to cut your investment in those companies by 50% for the newly formed company?
Exactly. Suppose further that there was a large corporation that owns a 100 different brands. If they split into a 100 different companies, with their own management teams, distribution networks, etc, would you really want to increase your stake in those brands a hundred-fold?
Now, if it were a cap-weighted index and that happened, you would find that immediately after the split, the total of the cap-weighted values of their stocks would be about the same as that of the company before. No problem, either logically or practically. (And I actually know about that personally, because I actually held AT&T stock when the Bell System broke up).
The same is true for a fundamental index -- the total earnings, dividends, book value and sales of the stocks will be about the same after a split.
User avatar
noyopacific
Posts: 359
Joined: Wed Jan 28, 2009 5:06 pm
Location: Mendocino

Now Dividend Aristocrats are Better !

Post by noyopacific »

Nick Vardi is at it again:
http://www.marketwatch.com/story/heres- ... 2015-01-09
Vardi says: "I recently had the temerity to recommend an equally weighted S&P 500 index fund as an alternative to Vanguard’s market-cap-weighted S&P 500 Index fund.
Bogleheads across the land rushed to the defense of Vanguard's low-cost, bread-and-butter S&P 500 Index funds with an almost religious fervor."
Apparently Nick reads this forum so I will chastise him for mischaracterizing the nature of the debate.
He seems to believe that mining relatively recent data to find an approach that beats the S&P 500 is a remarkable talent.
The information contained herein, while not guaranteed by us, has been obtained from from sources which have not in the past proved particularly reliable.
User avatar
steve r
Posts: 1300
Joined: Mon Feb 13, 2012 7:34 pm
Location: Connecticut

Re: Equal Weighted Funds - Better?

Post by steve r »

I am NOT an equal weighter ... but ... 50+ years of data does make one pause. I would not refer to it as illogical. So I will make some counter arguments.

1) Risk adjusted returns--- This has already been discussed.

2) Casino -- Investors are the casino in that the odds are slightly in our favor. A casino that does $10 M in bets in a night would much rather have that spread out over a million bets of ten dollar than 2 bets of $5 million. With the S&P500, some of the largest holding means taking a disproportionately large bet on them. Why bet on Apple or Exxon when you can make hundreds of other bets (with an equal expected return or better if smaller means higher returns). This is not a problem for indexes that do not hold the very largest caps. In the Extended Market Index, for example, if the largest holding gets disproportionately large it joins the S&P500. It is no accident that when the mega caps do well, the EW approach under-performs.

3) Overweight/underweight--- Lets lay in a cap weighted index there are 20 stocks overvalued and 20 stocks undervalued. No one knows which one these are. The 20 overvalued stocks should be 4 percent of the portfolio, but because they are overvalued they are 4.4 percent of the portfolio. Coveresely, The 20 undervalued stocks should be 4 percent of the portfolio, but because they are undervalued they are 3.6 percent of the portfolio.

My problem with the last two hypotheses is that you would expect a casino that has a bunch of equal bets to have a tighter range of outcomes than one that makes a small number of large bets. Likewise, an index that overweights overvalued stocks to be more volatile. But as has been pointed out, the S&P500 index is less volatile than the EW alternative.

This brings us back to point 1. Smaller stocks have a greater return and greater risk. EW puts more weight into smaller stocks in any given index.
"Owning the stock market over the long term is a winner's game. Attempting to beat the market is a loser's game. ..Don't look for the needle in the haystack. Just buy the haystack." Jack Bogle
User avatar
nisiprius
Advisory Board
Posts: 52211
Joined: Thu Jul 26, 2007 9:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: Now Dividend Aristocrats are Better !

Post by nisiprius »

noyopacific wrote:Nick Vardi is at it again:
http://www.marketwatch.com/story/heres- ... 2015-01-09
Vardi says: "I recently had the temerity to recommend an equally weighted S&P 500 index fund as an alternative to Vanguard’s market-cap-weighted S&P 500 Index fund.
Bogleheads across the land rushed to the defense of Vanguard's low-cost, bread-and-butter S&P 500 Index funds with an almost religious fervor."
Apparently Nick reads this forum so I will chastise him for mischaracterizing the nature of the debate.
He seems to believe that mining relatively recent data to find an approach that beats the S&P 500 is a remarkable talent.
What debate? He wrote an article, he says unidentified Bogleheads criticized what he said. "Debate" would imply some attempt to address and rebut the criticism. He does no such thing. He engages in some mild name-calling ("religious fervor," etc.) says "no amount of academic research" will convince them... and proceeds to cite no academic research.

And then moves on to a totally 'nother fund.

As best I can follow it is, his reasoning is, "Since I am making a fresh and unrelated claim that the Dividend Aristocrats Index outperforms the S&P, therefore it must be true that the S&P 500 Equal-Weighted Index also outperforms the S&P."
Annual income twenty pounds, annual expenditure nineteen nineteen and six, result happiness; Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery.
mrboast
Posts: 159
Joined: Sat Aug 01, 2009 4:29 pm

Re: Equal Weighted Funds - Better?

Post by mrboast »

Angst wrote:
nisiprius wrote:If it is logical to equal-weight, then I can't think of any logical reason not to equal-weight the whole market, and I am sure the reason why RSP only covers the S&P 500 is that it is utterly impractical to do much more than that (because there aren't enough small-cap stocks to go around). It might even be impossible under mutual fund diversification restrictions.
No worries, Nisi!
Just combine 33.3% of EWRI with 66.7% of EWRS and you're covered:

EWRI Guggenheim Russell 1000® Equal Weight ETF
EWRS Guggenheim Russell 2000® Equal Weight ETF
OK, I'm relatively new to the EW debate but I certainly understand the concept. One thing I don't understand is that a quick trip to the "Holdings" tab of the Gugg Rusell 2000 link above shows me that the holdings within the fund are nowhere close to an equal weighting; individually they seem to account for 0.3% to 0.01% of net assets - when one would expect each to be fairly close to (1/2000), no? Similarly, the R1000 one has holdings weight that reach a high of 0.29% and low of 0.01%. How do they justify the embarrassing fact that some holdings are weighted >10x the amount of others?

Am I wrong?
User avatar
nisiprius
Advisory Board
Posts: 52211
Joined: Thu Jul 26, 2007 9:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: Equal Weighted Funds - Better?

Post by nisiprius »

mrboast wrote:...OK, I'm relatively new to the EW debate but I certainly understand the concept. One thing I don't understand is that a quick trip to the "Holdings" tab of the Gugg Rusell 2000 link above shows me that the holdings within the fund are nowhere close to an equal weighting; individually they seem to account for 0.3% to 0.01% of net assets - when one would expect each to be fairly close to (1/2000), no? Similarly, the R1000 one has holdings weight that reach a high of 0.29% and low of 0.01%. How do they justify the embarrassing fact that some holdings are weighted >10x the amount of others?

Am I wrong?...
Yeah, I noticed that too. It turns out that the equal weight index isn't what I thought it was: "Rather than simply assigning an equal weight to each constituent of the index, Russell's sector equal weight index methodology equally weights each sector within the index and then equally weights the companies within each sector." They say this is a good thing. Since I don't understand the rationale for equal weight, I can't judge whether their rationale for unequally weighted equal weights makes sense. These stocks are volatile, and if I understand correctly, the index is only reconstituted annually, on the last trading day in May, so if they are equal in weight at the start of June they could be quite far from equal by end of the next May. I think.
Annual income twenty pounds, annual expenditure nineteen nineteen and six, result happiness; Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery.
Angst
Posts: 2968
Joined: Sat Jun 09, 2007 11:31 am

Re: Equal Weighted Funds - Better?

Post by Angst »

nisiprius wrote:
mrboast wrote:...OK, I'm relatively new to the EW debate but I certainly understand the concept. One thing I don't understand is that a quick trip to the "Holdings" tab of the Gugg Rusell 2000 link above shows me that the holdings within the fund are nowhere close to an equal weighting; individually they seem to account for 0.3% to 0.01% of net assets - when one would expect each to be fairly close to (1/2000), no? Similarly, the R1000 one has holdings weight that reach a high of 0.29% and low of 0.01%. How do they justify the embarrassing fact that some holdings are weighted >10x the amount of others?

Am I wrong?...
Yeah, I noticed that too. It turns out that the equal weight index isn't what I thought it was: "Rather than simply assigning an equal weight to each constituent of the index, Russell's sector equal weight index methodology equally weights each sector within the index and then equally weights the companies within each sector." They say this is a good thing. Since I don't understand the rationale for equal weight, I can't judge whether their rationale for unequally weighted equal weights makes sense. These stocks are volatile, and if I understand correctly, the index is only reconstituted annually, on the last trading day in May, so if they are equal in weight at the start of June they could be quite far from equal by end of the next May. I think.
I agree that the sector weighting within the index needs to be looked at. I suspect that the percentages in the ETF are generally not so far off the actual index, but I haven't confirmed that. I did note however on pg. 48 of the Russell Methodology document that Nisi referenced above that Russell re-weights the index quarterly. I don't believe the annual reconstitution determines weightings, per se, so much as just simply "membership". Nonetheless, in fairness to Russell, how well Guggenheim actually tracks the index matters too. According to Guggenheim:
EWRS Prospectus pg.17 wrote:The Advisor rebalances the Fund’s portfolio at the same rebalance interval(s) utilized by the Underlying Index.
I guess that means they do it every three months? That would mean March, June, September and December, according to Russell. If there just were an easy way to see how well EWRS actually tracks its index...
Post Reply