Author Topic: Are Traditional Index Funds Truly Optimal  (Read 18406 times)

mizzourah2006

  • Pencil Stache
  • ****
  • Posts: 535
  • Location: NWA
Are Traditional Index Funds Truly Optimal
« on: June 02, 2015, 08:14:29 AM »
There has been some discussion around here regarding index funds and I have mentioned the fact that index funds are market cap weighted may be a detriment to your portfolio. As a company's market cap grows your investment in them becomes even larger. Most S&P Index funds have 31-35% of the portfolio in the largest 20-25 companies in the index and as the companies become larger that % will continue to grow.

IMO it would be similar to owning a portfolio of 10 individual stocks where 1-2 of them drastically outperform the others for 3-4 years and you never bother to re-balance.

I found a recent academic paper on it, where they used monte carlo simulations of random buy and hold portfolios vs. a market cap weighted index fund over 5 year periods.

Here is a quote from the article.

Quote
These results imply that 74.8% of the completely random buy-and-hold portfolios beat the market. This is the central result and main take-away message of the paper. It is a result that calls for a fundamental revision in the way we perceive the optimality of the market portfolio.

Another good quote:

Quote
The main reason for the market’s inefficiency is that it is a portfolio with a very skewed distribution of weights. The weight of the largest market capitalization S&P500 stock (Apple, market cap of $571.5B, as of July 2014) is more than 100 times larger than the weight of the smallest cap S&P500 stock (U.S. Steel, market cap of $3.9B, as of July 2014). The 10 largest stocks in the S&P500 index account for 17.3% of the total S&P500 market capitalization, and the 20 largest stocks account for 27.6% of the total S&P500 market capitalization. 9 The largest firm is on average 68.1 times larger than the median S&P 500 firm. In contrast, in the random portfolios the largest portfolio weight is only twice the median portfolio weight. The very skewed distribution of weights in the market portfolio makes the diversification in this portfolio not very effective, because most of the portfolio is concentrated in a very small number of stocks. One could obtain a much higher diversification benefit with a more evenly weighted portfolio (such as the random portfolios considered here).

You can download the paper here if you are interested.

http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2488433

Levy, M. (2014). It's easy to beat the market. Social Science Research Network.

Just curious what others thoughts are. I post this in an effort to begin a discussion. Not an argument over passive vs. active investing.
« Last Edit: June 02, 2015, 08:29:39 AM by mizzourah2006 »

forummm

  • Walrus Stache
  • *******
  • Posts: 7389
  • Senior Mustachian
Re: Are Traditional Index Funds Truly Optimal
« Reply #1 on: June 02, 2015, 08:59:26 AM »
I'll have time to read the article later today. But I wanted to point out that I bought Apple in 2004 and 2003 and 2002 and 2005. And each year since then. Because of my index funds. I bought at those really low prices and have made a ton of money since then. If I'd bought some random stocks I would have most likely missed out on Apple.

I don't think a Monte Carlo simulation is a good way to have done this analysis. Instead, take stocks that were actually in existence at a particular time and then calculate the returns for randomly generated portfolios using real data. I've seen analyses doing this exact thing, using real data, and they show that the vast majority underperform the market because you miss out on the huge gainers like Dell or Home Depot back when they were tiny.

NorCal

  • Pencil Stache
  • ****
  • Posts: 805
Re: Are Traditional Index Funds Truly Optimal
« Reply #2 on: June 02, 2015, 09:15:36 AM »
Cap-weighting as absolutely sub-optimal from a total returns standpoint for two reasons:

1.  Cap weighting inherently gives more weight to over-valued stocks and less weight to under-valued stocks.  You're fighting value.
2.  You are over-weighting large-caps at the expense of small caps.  In the context of an S&P 500 fund, only the top ~50-100 stocks are actually having a measurable difference on your returns.  As the largest companies, these are the ones with relatively lower future potential returns.

You can offset problem #2 above by adding a small-cap fund, although these are usually cap-weighted as well.  You can offset #1 and #2 by using equal weight funds.  However, equal weight funds are tax-inefficient due to annual re-balancing, so these should only be used in tax advantaged accounts.  Fund fees are also typically higher, but should be offset over the long term due to the better index structure.

mizzourah2006

  • Pencil Stache
  • ****
  • Posts: 535
  • Location: NWA
Re: Are Traditional Index Funds Truly Optimal
« Reply #3 on: June 02, 2015, 09:30:33 AM »

I don't think a Monte Carlo simulation is a good way to have done this analysis. Instead, take stocks that were actually in existence at a particular time and then calculate the returns for randomly generated portfolios using real data. I've seen analyses doing this exact thing, using real data, and they show that the vast majority underperform the market because you miss out on the huge gainers like Dell or Home Depot back when they were tiny.

The simulations were based off of stocks that were in existence. I guess monte carlo simulation wasn't the most appropriate term. They randomly selected and randomly weighted stocks in random 5 year periods, but used the actual data from those stocks in those given years and compared them to a market cap weighted index fund from that same time period.

I'll have time to read the article later today. But I wanted to point out that I bought Apple in 2004 and 2003 and 2002 and 2005. And each year since then. Because of my index funds. I bought at those really low prices and have made a ton of money since then. If I'd bought some random stocks I would have most likely missed out on Apple.

Apple's Market Cap in 2002 was ~ 7 Billion. It wouldn't have even been a top 150 company in the S&P that year. So you would have owned virtually none of it had you held a market cap weighted index fund. It didn't even hit 100 billion until 2007, which would have put it in the mid-30s rank as far as market cap goes at that time.
« Last Edit: June 02, 2015, 09:34:21 AM by mizzourah2006 »

matchewed

  • Magnum Stache
  • ******
  • Posts: 4376
  • Location: CT
Re: Are Traditional Index Funds Truly Optimal
« Reply #4 on: June 02, 2015, 10:19:06 AM »
Well in order to execute any other style, value weighted or equal weighted...etc, you need more buying and selling within that index. That would require more fees being charged. So while you may be able to make an argument that value weighted or equal weighted is more optimal (gl w/ that) factor in fees.

The index itself and how the market performs is the rebalancing mechanism of an index fund. As I DCA I buy as the market fluctuates, as capitalization of various companies change so does the index.

Also the paper mentions five year intervals. I'm not investing for five years. I'm investing for 60.

The paper talks about random buy and hold versus the value weight (traditional index fund). You can view the random buy and hold as some form of equal weight fund. There has been previous discussion about equal weight funds here.

I've previously mentioned that in order to do an equal weight portfolio you'll probably need more buying and selling within your defined index. That means more fees. So I'd like to see a longer study with costs associated.

Scandium

  • Handlebar Stache
  • *****
  • Posts: 2317
  • Location: EastCoast
Re: Are Traditional Index Funds Truly Optimal
« Reply #5 on: June 02, 2015, 10:36:06 AM »
If I'm buying the total market to receive a share of the profits of the companies in it, why shouldn't I allocate more of my capital to the companies with more profits (i.e. market cap)?

You're saying I should allocate an equal amount of my investment capital to Apple as I would Jimbo's Gun Emporium in Des Moines (If it was listed)? How does that make sense?

I love when people come up with reasons for why index funds are money-loosing or suboptimal. But somehow the index funds just continue to beat market timers and fancy strategies over and over! To misquote Churchill: "index funds are the worst strategy, except for all the others"

mizzourah2006

  • Pencil Stache
  • ****
  • Posts: 535
  • Location: NWA
Re: Are Traditional Index Funds Truly Optimal
« Reply #6 on: June 02, 2015, 10:42:04 AM »
Well in order to execute any other style, value weighted or equal weighted...etc, you need more buying and selling within that index. That would require more fees being charged. So while you may be able to make an argument that value weighted or equal weighted is more optimal (gl w/ that) factor in fees.

The index itself and how the market performs is the rebalancing mechanism of an index fund. As I DCA I buy as the market fluctuates, as capitalization of various companies change so does the index.

Also the paper mentions five year intervals. I'm not investing for five years. I'm investing for 60.

The paper talks about random buy and hold versus the value weight (traditional index fund). You can view the random buy and hold as some form of equal weight fund. There has been previous discussion about equal weight funds here.

I've previously mentioned that in order to do an equal weight portfolio you'll probably need more buying and selling within your defined index. That means more fees. So I'd like to see a longer study with costs associated.

The authors didn't advocate equal weighting. They advocated buy and hold for the 5 year period. They just suggested the random allocation of the portfolio that was selected be more equally balanced than roughly 1/3rd of the weight being attributed to 25 of the 500 companies.

Quote
The largest firm is on average 68.1 times larger than the median S&P 500 firm. In contrast, in the random portfolios the largest portfolio weight is only twice the median portfolio weight.

I took a look at the RSP fund that you mentioned in the other thread vs. the SPY. Since June 1, 2005 (10 years ago) the RSP has returned 145%. The SPY has returned 115%.

ER of RSP is .4%, ER of SPY is .09%.

Very interesting. I hadn't heard of RSP before. Thanks for the heads up.

matchewed

  • Magnum Stache
  • ******
  • Posts: 4376
  • Location: CT
Re: Are Traditional Index Funds Truly Optimal
« Reply #7 on: June 02, 2015, 10:54:15 AM »
Well in order to execute any other style, value weighted or equal weighted...etc, you need more buying and selling within that index. That would require more fees being charged. So while you may be able to make an argument that value weighted or equal weighted is more optimal (gl w/ that) factor in fees.

The index itself and how the market performs is the rebalancing mechanism of an index fund. As I DCA I buy as the market fluctuates, as capitalization of various companies change so does the index.

Also the paper mentions five year intervals. I'm not investing for five years. I'm investing for 60.

The paper talks about random buy and hold versus the value weight (traditional index fund). You can view the random buy and hold as some form of equal weight fund. There has been previous discussion about equal weight funds here.

I've previously mentioned that in order to do an equal weight portfolio you'll probably need more buying and selling within your defined index. That means more fees. So I'd like to see a longer study with costs associated.

The authors didn't advocate equal weighting. They advocated buy and hold for the 5 year period. They just suggested the random allocation of the portfolio that was selected be more equally balanced than roughly 1/3rd of the weight being attributed to 25 of the 500 companies.


When the argument is stated that an index fund is over-weighted in certain companies compared to the random portfolios then inherently the random portfolios are going to be more equal weighted. Especially when the comparison is against 1000 different random buy and hold portfolios.

mizzourah2006

  • Pencil Stache
  • ****
  • Posts: 535
  • Location: NWA
Re: Are Traditional Index Funds Truly Optimal
« Reply #8 on: June 02, 2015, 10:58:34 AM »
If I'm buying the total market to receive a share of the profits of the companies in it, why shouldn't I allocate more of my capital to the companies with more profits (i.e. market cap)?

You're saying I should allocate an equal amount of my investment capital to Apple as I would Jimbo's Gun Emporium in Des Moines (If it was listed)? How does that make sense?

I love when people come up with reasons for why index funds are money-loosing or suboptimal. But somehow the index funds just continue to beat market timers and fancy strategies over and over! To misquote Churchill: "index funds are the worst strategy, except for all the others"

The market doesn't necessarily care about current profits as they are in the past. They care about future profits. If they just cared about current profits IBM, WMT, MSFT, Apple, etc. would have much higher P/Es and Facebook would have a significantly smaller market cap (and P/E) than 225 billion.

DCF does a decent job of explaining this.

http://en.wikipedia.org/wiki/Discounted_cash_flow
http://www.investopedia.com/terms/d/dcf.asp

Scandium

  • Handlebar Stache
  • *****
  • Posts: 2317
  • Location: EastCoast
Re: Are Traditional Index Funds Truly Optimal
« Reply #9 on: June 02, 2015, 11:05:00 AM »
If I'm buying the total market to receive a share of the profits of the companies in it, why shouldn't I allocate more of my capital to the companies with more profits (i.e. market cap)?

You're saying I should allocate an equal amount of my investment capital to Apple as I would Jimbo's Gun Emporium in Des Moines (If it was listed)? How does that make sense?

I love when people come up with reasons for why index funds are money-loosing or suboptimal. But somehow the index funds just continue to beat market timers and fancy strategies over and over! To misquote Churchill: "index funds are the worst strategy, except for all the others"

The market doesn't necessarily care about current profits as they are in the past. They care about future profits. If they just cared about current profits IBM, WMT, MSFT, Apple, etc. would have much higher P/Es and Facebook would have a significantly smaller market cap (and P/E) than 225 billion.

Exactly. The expected future profits are priced in, and reflected in the market cap. So by buying a cap weighted index I'm buying in proportion to future earnings. That seems logical to me. As earnings rise I get more of that company, and as they fall I own less.

thd7t

  • Handlebar Stache
  • *****
  • Posts: 1334
Re: Are Traditional Index Funds Truly Optimal
« Reply #10 on: June 02, 2015, 11:05:31 AM »
I have only briefly looked over the article and can't yet determine by how much the average "random portfolio" has beaten the market, but it appears that one is taking a 25% chance risk of under-performing in order to secure a slightly larger return.  A 75% chance of success is good, but I'm not sure how successful the median portfolio is, when compared to the market at large.

I'm a red panda

  • Walrus Stache
  • *******
  • Posts: 8035
  • Location: United States
Re: Are Traditional Index Funds Truly Optimal
« Reply #11 on: June 02, 2015, 11:11:15 AM »
I have only briefly looked over the article and can't yet determine by how much the average "random portfolio" has beaten the market, but it appears that one is taking a 25% chance risk of under-performing in order to secure a slightly larger return.  A 75% chance of success is good, but I'm not sure how successful the median portfolio is, when compared to the market at large.

If a portfolio were truly random, wouldn't it be representative of the market?

Or do you mean a random selection, of a typical investors selected portfolio?

thd7t

  • Handlebar Stache
  • *****
  • Posts: 1334
Re: Are Traditional Index Funds Truly Optimal
« Reply #12 on: June 02, 2015, 11:12:59 AM »
I have only briefly looked over the article and can't yet determine by how much the average "random portfolio" has beaten the market, but it appears that one is taking a 25% chance risk of under-performing in order to secure a slightly larger return.  A 75% chance of success is good, but I'm not sure how successful the median portfolio is, when compared to the market at large.

If a portfolio were truly random, wouldn't it be representative of the market?

Or do you mean a random selection, of a typical investors selected portfolio?
It's a random selection.  The term was taken from the paper.  I believe that if the weighting is different, then it isn't representative of the market.

matchewed

  • Magnum Stache
  • ******
  • Posts: 4376
  • Location: CT
Re: Are Traditional Index Funds Truly Optimal
« Reply #13 on: June 02, 2015, 11:14:54 AM »
I have only briefly looked over the article and can't yet determine by how much the average "random portfolio" has beaten the market, but it appears that one is taking a 25% chance risk of under-performing in order to secure a slightly larger return.  A 75% chance of success is good, but I'm not sure how successful the median portfolio is, when compared to the market at large.

How they defined it was that they took random 5 year periods between 1926 and 2014. In those 5 year periods they selected 10 stocks. They did this 100 times. They compared this against an index fund. This did better than the index fund 75% of the time according to the paper.

beltim

  • Magnum Stache
  • ******
  • Posts: 2964
Re: Are Traditional Index Funds Truly Optimal
« Reply #14 on: June 02, 2015, 11:16:53 AM »
If I'm buying the total market to receive a share of the profits of the companies in it, why shouldn't I allocate more of my capital to the companies with more profits (i.e. market cap)?
 

An efficient market would have every publicly traded company priced to give the same future risk-adjusted return. So why should you overweight large companies just because they're larger?  You are starting from that conclusion without giving a good justification.

One reason is fees - market cap weighted fuss don't have to rebalance, so they have lower fees.

Another reason could be risk - small caps are generally considered higher risk.

Of course, the other side of that coin is the small caps traditionally have higher returns as well to compensate for that risk. 

thd7t

  • Handlebar Stache
  • *****
  • Posts: 1334
Re: Are Traditional Index Funds Truly Optimal
« Reply #15 on: June 02, 2015, 11:18:29 AM »
I have only briefly looked over the article and can't yet determine by how much the average "random portfolio" has beaten the market, but it appears that one is taking a 25% chance risk of under-performing in order to secure a slightly larger return.  A 75% chance of success is good, but I'm not sure how successful the median portfolio is, when compared to the market at large.

How they defined it was that they took random 5 year periods between 1926 and 2014. In those 5 year periods they selected 10 stocks. They did this 100 times. They compared this against an index fund. This did better than the index fund 75% of the time according to the paper.
I had a hard time figuring out how much better, though.  A median return vs market return would be helpful (and is probably in there).  I don't know if I'd take the risk for a marginal improvement on returns.

matchewed

  • Magnum Stache
  • ******
  • Posts: 4376
  • Location: CT
Re: Are Traditional Index Funds Truly Optimal
« Reply #16 on: June 02, 2015, 11:24:02 AM »
I have only briefly looked over the article and can't yet determine by how much the average "random portfolio" has beaten the market, but it appears that one is taking a 25% chance risk of under-performing in order to secure a slightly larger return.  A 75% chance of success is good, but I'm not sure how successful the median portfolio is, when compared to the market at large.

How they defined it was that they took random 5 year periods between 1926 and 2014. In those 5 year periods they selected 10 stocks. They did this 100 times. They compared this against an index fund. This did better than the index fund 75% of the time according to the paper.
I had a hard time figuring out how much better, though.  A median return vs market return would be helpful (and is probably in there).  I don't know if I'd take the risk for a marginal improvement on returns.

Well the paper does discuss the Sharpe ratio.

thd7t

  • Handlebar Stache
  • *****
  • Posts: 1334
Re: Are Traditional Index Funds Truly Optimal
« Reply #17 on: June 02, 2015, 11:32:00 AM »
I have only briefly looked over the article and can't yet determine by how much the average "random portfolio" has beaten the market, but it appears that one is taking a 25% chance risk of under-performing in order to secure a slightly larger return.  A 75% chance of success is good, but I'm not sure how successful the median portfolio is, when compared to the market at large.

How they defined it was that they took random 5 year periods between 1926 and 2014. In those 5 year periods they selected 10 stocks. They did this 100 times. They compared this against an index fund. This did better than the index fund 75% of the time according to the paper.
I had a hard time figuring out how much better, though.  A median return vs market return would be helpful (and is probably in there).  I don't know if I'd take the risk for a marginal improvement on returns.

Well the paper does discuss the Sharpe ratio.
I've looked back at it and it appears that they're showing about a 1% (annual) average improvement over the market return and a slightly higher average standard deviation.  I think that they use a mean average, so I'll have to look again to see if there were any outliers in that calculation that make things look better or worse than they appear (or if I misread and it was a median).

NorCal

  • Pencil Stache
  • ****
  • Posts: 805
Re: Are Traditional Index Funds Truly Optimal
« Reply #18 on: June 02, 2015, 03:12:32 PM »
If I'm buying the total market to receive a share of the profits of the companies in it, why shouldn't I allocate more of my capital to the companies with more profits (i.e. market cap)?

This statement is fundamentally incorrect.  You are buying shares with the highest market capitalization, not the highest earnings.

As a hypothetical example (because it's so extreme), let's create a three company portfolio of Uber (currently valued at $50B on the private market), Lockheed Martin ($60B) Market Cap, and Intuit ($29B Market Cap).  In a cap-weighted index, you would allocate 36% to Uber, 43% to LMT, and 21% to Intuit.  In an equal-weighted index, you would allocate a third to each.  Does a company that is losing money (Uber) deserve premium placement in your portfolio just because someone else thinks it's worth $50B?  Or does it make sense to allocate an equal amount to each company?

Index's were fundamentally designed for different purposes than they're being used for.  They were designed to track economic activity, not to create an effective investing vehicle.

While passive investment as a strategy is a great investment approach, the common underlying index's are not designed to optimize returns over time.  They aren't bad, but equal weight is better.

forummm

  • Walrus Stache
  • *******
  • Posts: 7389
  • Senior Mustachian
Re: Are Traditional Index Funds Truly Optimal
« Reply #19 on: June 02, 2015, 03:36:47 PM »
Cap-weighting as absolutely sub-optimal from a total returns standpoint for two reasons:

1.  Cap weighting inherently gives more weight to over-valued stocks and less weight to under-valued stocks.  You're fighting value.
2.  You are over-weighting large-caps at the expense of small caps.  In the context of an S&P 500 fund, only the top ~50-100 stocks are actually having a measurable difference on your returns.  As the largest companies, these are the ones with relatively lower future potential returns.

You can offset problem #2 above by adding a small-cap fund, although these are usually cap-weighted as well.  You can offset #1 and #2 by using equal weight funds.  However, equal weight funds are tax-inefficient due to annual re-balancing, so these should only be used in tax advantaged accounts.  Fund fees are also typically higher, but should be offset over the long term due to the better index structure.

But from a risk-adjusted total returns standpoint, cap weighting should be effective, if you believe the EMH. The market is saying those are the most valuable companies to own and in that proportion.

Quote
The largest firm is on average 68.1 times larger than the median S&P 500 firm. In contrast, in the random portfolios the largest portfolio weight is only twice the median portfolio weight.

I took a look at the RSP fund that you mentioned in the other thread vs. the SPY. Since June 1, 2005 (10 years ago) the RSP has returned 145%. The SPY has returned 115%.

ER of RSP is .4%, ER of SPY is .09%.

Very interesting. I hadn't heard of RSP before. Thanks for the heads up.

Equal weighting essentially just overweights mid-cap stocks relative to a market cap weighting, and at somewhat higher cost. So in periods where mid-caps outperform large caps (like the last 10 years), an equal weight portfolio will provide higher returns. But mid caps also have more volatility. It's not a free lunch.

Scandium

  • Handlebar Stache
  • *****
  • Posts: 2317
  • Location: EastCoast
Re: Are Traditional Index Funds Truly Optimal
« Reply #20 on: June 02, 2015, 03:44:54 PM »
Yes that allocation totally makes sense. The market cap is a proxy for earnings. Future earnings! If the market has decided that the future earnings (and earnings growth) of Uber is worth that much I have no reason to believe otherwise. Do you? Why?

The more I think about it the more it seems equal weigh is just a form of market timing, you belive you know better than the collective market.
« Last Edit: June 02, 2015, 03:47:25 PM by Scandium »

Rubic

  • Handlebar Stache
  • *****
  • Posts: 1083
Re: Are Traditional Index Funds Truly Optimal
« Reply #21 on: June 02, 2015, 03:47:08 PM »
Investors favoring equal-weight funds believe that mid-cap stocks will outperform the market. While there have been times in the past that has held true, you could choose different time frames to show sub-par performance.  Also, you'll pay higher fees for equal-weight funds.

References:

mizzourah2006

  • Pencil Stache
  • ****
  • Posts: 535
  • Location: NWA
Re: Are Traditional Index Funds Truly Optimal
« Reply #22 on: June 02, 2015, 03:49:04 PM »
Cap-weighting as absolutely sub-optimal from a total returns standpoint for two reasons:

1.  Cap weighting inherently gives more weight to over-valued stocks and less weight to under-valued stocks.  You're fighting value.
2.  You are over-weighting large-caps at the expense of small caps.  In the context of an S&P 500 fund, only the top ~50-100 stocks are actually having a measurable difference on your returns.  As the largest companies, these are the ones with relatively lower future potential returns.

You can offset problem #2 above by adding a small-cap fund, although these are usually cap-weighted as well.  You can offset #1 and #2 by using equal weight funds.  However, equal weight funds are tax-inefficient due to annual re-balancing, so these should only be used in tax advantaged accounts.  Fund fees are also typically higher, but should be offset over the long term due to the better index structure.

But from a risk-adjusted total returns standpoint, cap weighting should be effective, if you believe the EMH. The market is saying those are the most valuable companies to own and in that proportion.

Quote
The largest firm is on average 68.1 times larger than the median S&P 500 firm. In contrast, in the random portfolios the largest portfolio weight is only twice the median portfolio weight.

I took a look at the RSP fund that you mentioned in the other thread vs. the SPY. Since June 1, 2005 (10 years ago) the RSP has returned 145%. The SPY has returned 115%.

ER of RSP is .4%, ER of SPY is .09%.

Very interesting. I hadn't heard of RSP before. Thanks for the heads up.

Equal weighting essentially just overweights mid-cap stocks relative to a market cap weighting, and at somewhat higher cost. So in periods where mid-caps outperform large caps (like the last 10 years), an equal weight portfolio will provide higher returns. But mid caps also have more volatility. It's not a free lunch.

It only overweights mid caps if you believe large caps should have higher weights. IMO it treats all stocks in the S&P the same regardless of market cap.

beltim

  • Magnum Stache
  • ******
  • Posts: 2964
Re: Are Traditional Index Funds Truly Optimal
« Reply #23 on: June 02, 2015, 03:56:50 PM »
The more I think about it the more it seems equal weigh is just a form of market timing, you belive you know better than the collective market.

Huh?  Where do you get timing from?

NorCal

  • Pencil Stache
  • ****
  • Posts: 805
Re: Are Traditional Index Funds Truly Optimal
« Reply #24 on: June 02, 2015, 04:22:59 PM »
Yes that allocation totally makes sense. The market cap is a proxy for earnings. Future earnings! If the market has decided that the future earnings (and earnings growth) of Uber is worth that much I have no reason to believe otherwise. Do you? Why?

The more I think about it the more it seems equal weigh is just a form of market timing, you believe you know better than the collective market.

You're actually thinking about this backwards.  Using cap-weighting means that you are implicitly following active investors by buying overbought markets and selling oversold markets.  You're accepting the valuations active managers are using, which is what the EMH cautions against.  My example is extreme because I compared Uber (a company with no earnings) to a company with a comparable valuation (Lockheed) and $3.6B in annual earnings.  Should these companies have a comparable valuation?  Probably not.

Equal weight treats all companies the same, regardless of "hot" companies valuation concerns.

For another way to think about it, cap-weighting will overweight hot-momentum stocks and underweight undervalued stocks.  Equal-weighting will do the inverse on a relative basis.

Also remember, per the EMH, small caps will outperform large-caps over time.  That is, higher risk equities will outperform lower risk equities when averaged over time.

forummm

  • Walrus Stache
  • *******
  • Posts: 7389
  • Senior Mustachian
Re: Are Traditional Index Funds Truly Optimal
« Reply #25 on: June 02, 2015, 04:47:48 PM »
Yes that allocation totally makes sense. The market cap is a proxy for earnings. Future earnings! If the market has decided that the future earnings (and earnings growth) of Uber is worth that much I have no reason to believe otherwise. Do you? Why?

The more I think about it the more it seems equal weigh is just a form of market timing, you believe you know better than the collective market.

You're actually thinking about this backwards.  Using cap-weighting means that you are implicitly following active investors by buying overbought markets and selling oversold markets.  You're accepting the valuations active managers are using, which is what the EMH cautions against.  My example is extreme because I compared Uber (a company with no earnings) to a company with a comparable valuation (Lockheed) and $3.6B in annual earnings.  Should these companies have a comparable valuation?  Probably not.

Equal weight treats all companies the same, regardless of "hot" companies valuation concerns.

For another way to think about it, cap-weighting will overweight hot-momentum stocks and underweight undervalued stocks.  Equal-weighting will do the inverse on a relative basis.

Also remember, per the EMH, small caps will outperform large-caps over time.  That is, higher risk equities will outperform lower risk equities when averaged over time.

Huh? What does "the EMH caution against"?

MrMoneyPinch

  • 5 O'Clock Shadow
  • *
  • Posts: 82
  • Location: Soviet Canuckistan
Re: Are Traditional Index Funds Truly Optimal
« Reply #26 on: June 02, 2015, 04:50:36 PM »
The index strategy has never been *optimal*.  It's just superior to the vast majority of real-world, professional investors' returns.

Yes, there are some exceptions like Warren Buffett, which has fantastic returns in his past (and present).  But if you were an investor when he opened his first private fund, you could not have differentiated him from any average manager.

Anyway, to create an optimal strategy, you would have to get some advance knowledge of future events, by crook or by cunning modelling.  All financial models have broken down in time, some catastrophically.  Market-timing works well, except when your timing is off.  Hedging works well, if your hedge really counter-balances the rest of the portfolio at all times.  You can't count on that.   Even "strategic" allocation (meaning timing the market with portfolio composition changes) mostly relies on soothsaying.

IF you have the chops to jump in and do the financial equivalent of getting an olympic gold medal, go for it;  I do not have the drive, talent or deep pockets needed so buying the market and getting "average" returns is perfect for me. 

neil

  • Stubble
  • **
  • Posts: 216
Re: Are Traditional Index Funds Truly Optimal
« Reply #27 on: June 02, 2015, 05:08:25 PM »
I don't understand the point of the simulations.  If you plot Sharpe ratios and show that the median stock selection does not have a Sharpe ratio of zero, your conclusion is already set.  If your goal is to beat the cap-weighted index, it doesn't make sense to select 10 stocks, it makes sense to take them all.  At that point we are back at a RSP vs VIIIX discussion.  (No matter how many stocks you select, 10, 50, 499, the N+1 stock still has the same percentage chance to beat the cap-weighted index.  With that as a stated goal, it makes sense to keep selecting because you will otherwise introduce variance and decrease return.)  Reducing portfolio size has the expected impact (same median as RSP, more variance than RSP) and I don't think anyone would make the reasonable recommendation that a dartboard approach is superior to selecting RSP.

(edit: there would be deviation if you are only "rebalancing" every five years versus quarterly, which is an argument in itself.)
« Last Edit: June 02, 2015, 05:10:36 PM by neil »

Scandium

  • Handlebar Stache
  • *****
  • Posts: 2317
  • Location: EastCoast
Re: Are Traditional Index Funds Truly Optimal
« Reply #28 on: June 02, 2015, 05:14:42 PM »
No, I believe the market is efficient. It has efficiently priced in the differences in future earnings. I don't believe in over or under- sold. Or at least not in my ability to tell.

You talk as if there is no value in the market price. You ignore all the aggregate information in the market. If company A's earnings double while company B's are cut in half you would not want to own more of A and less of B?

mizzourah2006

  • Pencil Stache
  • ****
  • Posts: 535
  • Location: NWA
Re: Are Traditional Index Funds Truly Optimal
« Reply #29 on: June 02, 2015, 05:17:49 PM »
No, I believe the market is efficient. It has efficiently priced in the differences in future earnings. I don't believe in over or under- sold. Or at least not in my ability to tell.

You talk as if there is no value in the market price. You ignore all the aggregate information in the market. If company A's earnings double while company B's are cut in half you would not want to own more of A and less of B?

i see market inefficiencies almost every week. I guess it just depends on how closely you are looking.

Scandium

  • Handlebar Stache
  • *****
  • Posts: 2317
  • Location: EastCoast
Re: Are Traditional Index Funds Truly Optimal
« Reply #30 on: June 02, 2015, 05:19:26 PM »
No, I believe the market is efficient. It has efficiently priced in the differences in future earnings. I don't believe in over or under- sold. Or at least not in my ability to tell.

You talk as if there is no value in the market price. You ignore all the aggregate information in the market. If company A's earnings double while company B's are cut in half you would not want to own more of A and less of B?

i see market inefficiencies almost every week. I guess it just depends on how closely you are looking.
I assume that means you're posting this from your own tropical island then?

mizzourah2006

  • Pencil Stache
  • ****
  • Posts: 535
  • Location: NWA
Re: Are Traditional Index Funds Truly Optimal
« Reply #31 on: June 02, 2015, 05:25:23 PM »
No, I believe the market is efficient. It has efficiently priced in the differences in future earnings. I don't believe in over or under- sold. Or at least not in my ability to tell.

You talk as if there is no value in the market price. You ignore all the aggregate information in the market. If company A's earnings double while company B's are cut in half you would not want to own more of A and less of B?

i see market inefficiencies almost every week. I guess it just depends on how closely you are looking.
I assume that means you're posting this from your own tropical island then?

Yes because I have clearly been able to accumulate enough assets in 3 years in the workforce to buy a tropical island, I always poor every dime into the position, and it only takes a few hours for me to make my money. Market inefficiencies often take 6-12 months to correct. 

skyrefuge

  • Handlebar Stache
  • *****
  • Posts: 1007
  • Location: Suburban Chicago, IL
Re: Are Traditional Index Funds Truly Optimal
« Reply #32 on: June 02, 2015, 05:27:30 PM »
Interesting. I didn't see any obvious holes in the research, though I also don't feel like I'm a very good judge. If it holds up, what's the takeaway? Just invest in 10 randomly-selected S&P 500 companies, hold them forever, and you're likely to outperform the index? It doesn't seem like this was difficult research to perform, so why is 2014 the first time anyone discovered this?

In terms of equal-weight indexes, the thing I always bring up (and never get a response on) is the inelegance that come from splits and mergers. If a company who sells 2000 cars a year splits into two companies who each sell 1000 cars a year, why would I suddenly want to own twice as much of those companies as I did before (and less of every other company?) And conversely, if a company that sells oats merges with one that sells wheat, I suddenly get only half the share of the profits I was getting from oats and wheat before?

Using cap-weighting means that you are implicitly following active investors by buying overbought markets and selling oversold markets.  You're accepting the valuations active managers are using

It seems that you aren't familiar with the beliefs of that support the theory of passive investing? Those of us who buy into it believe that "the market" (which includes the votes of all the active investors) knows better than we do what the true value of a company should be. Thus, there is no concept of "overbought" or "oversold" markets, and accepting the valuations of active managers is exactly what we intend to do.

So yeah, your hypothetical market makes total sense to me. If everyone else thinks Uber is worth $50B, then who am I to tell them they're wrong? (though in this case I might be a bit more skeptical since the "everyone else" valuing Uber is smaller than the "everyone else" valuing the publicly-traded companies).

beltim

  • Magnum Stache
  • ******
  • Posts: 2964
Re: Are Traditional Index Funds Truly Optimal
« Reply #33 on: June 02, 2015, 06:00:16 PM »
Something can be inelegant and still be a better investment strategy.

brooklynguy

  • Handlebar Stache
  • *****
  • Posts: 2204
  • Age: 39
Re: Are Traditional Index Funds Truly Optimal
« Reply #34 on: June 02, 2015, 07:19:18 PM »
Interesting indeed.  I'm having trouble following the methodology described in the paper.  Is what matchewed described above ("they took random 5 year periods between 1926 and 2014. In those 5 year periods they selected 10 stocks. They did this 100 times.") what the researchers actually did?  The paper says "[f]or each 5-year sub-period we draw 10 random passive portfolios" (not 10 individual stocks).  It then talks about randomly assigning initial portfolio weights for the stocks in the portfolios -- it sounds to me like they constructed portfolios containing all 500 of the stocks in their selected universe, but just randomized the weighting.  But it's not clear to me -- how are others interpreting their methodology?

In any event, like matchewed, I'm confused about why they used 5-year time windows.  Why not examine longer time horizons that are actually representative of a typical stock investor's time horizon?  Still, over any time period, you wouldn't expect more than half of a large number of randomly selected portfolios to beat the market index, so these results are interesting.

I'm also confused about this portion of the paper:

Quote
More formally, consider a portfolio that has the exact same portfolio weight distribution as the market portfolio (i.e. it has the diversification drawback due to a skewed weight distribution, as discussed above), but it is not tilted toward large stocks.

I don't understand what this means.  How can a portfolio have the same weightings as the market cap weighted index and not be tilted towards large stocks?  Doesn't the former by definition imply the latter?

tyir

  • 5 O'Clock Shadow
  • *
  • Posts: 60
Re: Are Traditional Index Funds Truly Optimal
« Reply #35 on: June 02, 2015, 07:49:45 PM »
Two thoughts:

1 - If one truly believed in equal-cap indexing, then shouldn't it naturally follow to apply that to country allocation as well? Assuming every country was equally tax efficient, then a equal-cap indexer should buy each country in equal amounts, with US the same allocation in % as Japan, Brazil, and all other countries. Another viewpoint would have all companies equal everywhere, and you would allocate your country % simply based on the number of companies each country has.

2 - There was the point earlier of buying Apple (and companies like it) while it was cheap, and not overvaluing them when they are overly large. However if you are equally cap weighted, as Apple became its huge rise, you would be constantly selling it to keep the allocation to Apple the same as every other company. So while you would have owned virtually none of it if it was market-cap weighted while it was small, you would still own a very small amount of it when it was large, and I don't think (however I didn't do the math) you would have made any more profit off it.

beltim

  • Magnum Stache
  • ******
  • Posts: 2964
Re: Are Traditional Index Funds Truly Optimal
« Reply #36 on: June 02, 2015, 08:01:39 PM »
Two thoughts:

1 - If one truly believed in equal-cap indexing, then shouldn't it naturally follow to apply that to country allocation as well? Assuming every country was equally tax efficient, then a equal-cap indexer should buy each country in equal amounts, with US the same allocation in % as Japan, Brazil, and all other countries. Another viewpoint would have all companies equal everywhere, and you would allocate your country % simply based on the number of companies each country has.

No. Each country has it's own risks and most of them do not have the historical returns the US does.  In principal one would expect a higher reward because of traditionally higher risk , but historically that hasn't been the case.

skyrefuge

  • Handlebar Stache
  • *****
  • Posts: 1007
  • Location: Suburban Chicago, IL
Re: Are Traditional Index Funds Truly Optimal
« Reply #37 on: June 02, 2015, 08:03:37 PM »
it sounds to me like they constructed portfolios containing all 500 of the stocks in their selected universe, but just randomized the weighting.  But it's not clear to me -- how are others interpreting their methodology?

Ah, yes, I was just influenced by matchewed's interpretation. It's really 500 stocks selected in 10 portfolios for each 5 year period. Because in the "Robustness" section he re-runs the simulation with 1000 stocks instead of 500 stocks. Interestingly, while the randoms still beat the market, the success rate is reduced from 75% to 66%. Makes me wonder if the outperformance is somehow an artifact of using the antiquated S&P500 to represent "the market" rather than an actual all-cap index.

I'm also confused about this portion of the paper:

Quote
More formally, consider a portfolio that has the exact same portfolio weight distribution as the market portfolio (i.e. it has the diversification drawback due to a skewed weight distribution, as discussed above), but it is not tilted toward large stocks.

I don't understand what this means.  How can a portfolio have the same weightings as the market cap weighted index and not be tilted towards large stocks?  Doesn't the former by definition imply the latter?

The set of weightings is the same as the actual index, and the set of stocks is the same, but the two are mixed up. For example, in the real index, Apple's weight is 3.5% or whatever. In his example, that 3.5% is instead assigned to Chipotle, and Chipotle's actual weight is assigned to CarMax, and so on...

In that example, he actually demonstrates the opposite of mizzou's belief: "it is not the tilt toward large stocks that hurts the market portfolio", since the portfolio with random companies getting the big weightings performs even worse than the real one where the big companies get the big weightings.
« Last Edit: June 02, 2015, 08:06:08 PM by skyrefuge »

YoungInvestor

  • Bristles
  • ***
  • Posts: 411
Re: Are Traditional Index Funds Truly Optimal
« Reply #38 on: June 02, 2015, 08:18:01 PM »
If I'm buying the total market to receive a share of the profits of the companies in it, why shouldn't I allocate more of my capital to the companies with more profits (i.e. market cap)?

You're saying I should allocate an equal amount of my investment capital to Apple as I would Jimbo's Gun Emporium in Des Moines (If it was listed)? How does that make sense?

I love when people come up with reasons for why index funds are money-loosing or suboptimal. But somehow the index funds just continue to beat market timers and fancy strategies over and over! To misquote Churchill: "index funds are the worst strategy, except for all the others"

The market doesn't necessarily care about current profits as they are in the past. They care about future profits. If they just cared about current profits IBM, WMT, MSFT, Apple, etc. would have much higher P/Es and Facebook would have a significantly smaller market cap (and P/E) than 225 billion.

Exactly. The expected future profits are priced in, and reflected in the market cap. So by buying a cap weighted index I'm buying in proportion to future earnings. That seems logical to me. As earnings rise I get more of that company, and as they fall I own less.

That's not very logical at all. What matters is how much you pay for that profit, not the total quantity the company is making.

If company A makes 15 billion dollars a year in profit but has a market cap of 450 billion dollars (P/E = 30) and company B makes 1 Billion and has a market cap of 10 Billion (P/E = 10), I want company B's shares, all else being equal.


forummm

  • Walrus Stache
  • *******
  • Posts: 7389
  • Senior Mustachian
Re: Are Traditional Index Funds Truly Optimal
« Reply #39 on: June 02, 2015, 08:36:49 PM »
A different study with imperfect methodology

http://www.efficientfrontier.com/ef/900/15st.htm

Quote
In order to investigate this problem, I looked at the stocks constituting the S&P 500 as of 11/30/99, and formed 98 random equally-weighted 15-stock portfolios for the 12/89-11/99 10-year holding period. Below is a histogram of the annualized portfolio returns:

.....
the TWD of these 15-stock portfolios is staggering—three-quarters of them failed to beat "the market."

Scandium

  • Handlebar Stache
  • *****
  • Posts: 2317
  • Location: EastCoast
Re: Are Traditional Index Funds Truly Optimal
« Reply #40 on: June 02, 2015, 08:40:09 PM »
If I'm buying the total market to receive a share of the profits of the companies in it, why shouldn't I allocate more of my capital to the companies with more profits (i.e. market cap)?

You're saying I should allocate an equal amount of my investment capital to Apple as I would Jimbo's Gun Emporium in Des Moines (If it was listed)? How does that make sense?

I love when people come up with reasons for why index funds are money-loosing or suboptimal. But somehow the index funds just continue to beat market timers and fancy strategies over and over! To misquote Churchill: "index funds are the worst strategy, except for all the others"

The market doesn't necessarily care about current profits as they are in the past. They care about future profits. If they just cared about current profits IBM, WMT, MSFT, Apple, etc. would have much higher P/Es and Facebook would have a significantly smaller market cap (and P/E) than 225 billion.

Exactly. The expected future profits are priced in, and reflected in the market cap. So by buying a cap weighted index I'm buying in proportion to future earnings. That seems logical to me. As earnings rise I get more of that company, and as they fall I own less.

That's not very logical at all. What matters is how much you pay for that profit, not the total quantity the company is making.

If company A makes 15 billion dollars a year in profit but has a market cap of 450 billion dollars (P/E = 30) and company B makes 1 Billion and has a market cap of 10 Billion (P/E = 10), I want company B's shares, all else being equal.

We are talking past each other because you, and others here, believe you can tell what is a "good price" and a "bad price" for a stock (i.e. future earnings). I'm simply throwing up my hands and saying I can't. If the P/E =30 there's a reason for that (higher future growth for example). If it's 10 there's also a reason. If the P/E was 1 would you jump all over it? Do people really believe there is no reasoning in the market, and prices are set totally haphazardly?

I may agree that a P/E for 300 (or whatever it is) for netflix is a bit crazy, but I have no better guess so I trust the market, for good or bad.
« Last Edit: June 02, 2015, 08:42:02 PM by Scandium »

brooklynguy

  • Handlebar Stache
  • *****
  • Posts: 2204
  • Age: 39
Re: Are Traditional Index Funds Truly Optimal
« Reply #41 on: June 02, 2015, 08:54:47 PM »
The set of weightings is the same as the actual index, and the set of stocks is the same, but the two are mixed up. For example, in the real index, Apple's weight is 3.5% or whatever. In his example, that 3.5% is instead assigned to Chipotle, and Chipotle's actual weight is assigned to CarMax, and so on...

Ah, got it.  Thanks.

Quote
Because in the "Robustness" section he re-runs the simulation with 1000 stocks instead of 500 stocks. Interestingly, while the randoms still beat the market, the success rate is reduced from 75% to 66%. Makes me wonder if the outperformance is somehow an artifact of using the antiquated S&P500 to represent "the market" rather than an actual all-cap index.

Shouldn't it be the other way around?  That is, the broader-market simulation (not using the antiquated S&P 500 stock universe) should have had greater outperformance, no?  (In each case, the S&P 500 market index was still serving as the benchmark for comparison.)

hodedofome

  • Handlebar Stache
  • *****
  • Posts: 1314
  • Age: 40
  • Location: Texas
Are Traditional Index Funds Truly Optimal
« Reply #42 on: June 02, 2015, 09:31:52 PM »
If you got a few million+ to invest, you could make your own equal weight S&P 500 index. I'd put money on that outperforming the standard index over the long term, not to mention your fees could be lower.

http://blog.alphaarchitect.com/2015/05/18/what-drives-the-sp-500-equal-weight-return-premium/

Sent from my iPhone using Tapatalk
« Last Edit: June 02, 2015, 09:39:28 PM by hodedofome »

mizzourah2006

  • Pencil Stache
  • ****
  • Posts: 535
  • Location: NWA
Re: Are Traditional Index Funds Truly Optimal
« Reply #43 on: June 03, 2015, 05:55:59 AM »
it sounds to me like they constructed portfolios containing all 500 of the stocks in their selected universe, but just randomized the weighting.  But it's not clear to me -- how are others interpreting their methodology?

Ah, yes, I was just influenced by matchewed's interpretation. It's really 500 stocks selected in 10 portfolios for each 5 year period. Because in the "Robustness" section he re-runs the simulation with 1000 stocks instead of 500 stocks. Interestingly, while the randoms still beat the market, the success rate is reduced from 75% to 66%. Makes me wonder if the outperformance is somehow an artifact of using the antiquated S&P500 to represent "the market" rather than an actual all-cap index.

I'm also confused about this portion of the paper:

Quote
More formally, consider a portfolio that has the exact same portfolio weight distribution as the market portfolio (i.e. it has the diversification drawback due to a skewed weight distribution, as discussed above), but it is not tilted toward large stocks.

I don't understand what this means.  How can a portfolio have the same weightings as the market cap weighted index and not be tilted towards large stocks?  Doesn't the former by definition imply the latter?

The set of weightings is the same as the actual index, and the set of stocks is the same, but the two are mixed up. For example, in the real index, Apple's weight is 3.5% or whatever. In his example, that 3.5% is instead assigned to Chipotle, and Chipotle's actual weight is assigned to CarMax, and so on...

In that example, he actually demonstrates the opposite of mizzou's belief: "it is not the tilt toward large stocks that hurts the market portfolio", since the portfolio with random companies getting the big weightings performs even worse than the real one where the big companies get the big weightings.
I'm not sure where you got the idea that the only reason I believe a market cap weighted fund is sub-optimal is because it weights the highest the best. I think that is one reason, it is rewarding past performance for expected future performance. But I would also expect randomly assigning an S&P company a weighting of 300x another S&P company to be equally as poor with more volatility. I actually am not totally against a market cap weighting strategy, but I think a more balanced strategy makes sense.
Perhaps something like this:

Stocks 1-25: 3x median
Stocks 26-100: 2x median
Stocks 101-250: median
Stocks 251-400: 2x smaller than median
Stocks 401-500: 3x smaller than median

I just find it odd that people who are devout index fund investors that are essentially saying "we don't know, so the best thing to do is own it all" are comfortable with an allocation that is essentially taking strong guesses that Apple will outperform any other stock every year and that "guess" continues to grow.

As someone else already mentioned the S&P 500 and Dow Jones Industrial average were designed to be indicators of economic movement. Our goal should be capital appreciation, not to track economic trends in the US.
« Last Edit: June 03, 2015, 06:08:23 AM by mizzourah2006 »

matchewed

  • Magnum Stache
  • ******
  • Posts: 4376
  • Location: CT
Re: Are Traditional Index Funds Truly Optimal
« Reply #44 on: June 03, 2015, 06:11:28 AM »
it sounds to me like they constructed portfolios containing all 500 of the stocks in their selected universe, but just randomized the weighting.  But it's not clear to me -- how are others interpreting their methodology?

Ah, yes, I was just influenced by matchewed's interpretation. It's really 500 stocks selected in 10 portfolios for each 5 year period. Because in the "Robustness" section he re-runs the simulation with 1000 stocks instead of 500 stocks. Interestingly, while the randoms still beat the market, the success rate is reduced from 75% to 66%. Makes me wonder if the outperformance is somehow an artifact of using the antiquated S&P500 to represent "the market" rather than an actual all-cap index.

I'm also confused about this portion of the paper:

Quote
More formally, consider a portfolio that has the exact same portfolio weight distribution as the market portfolio (i.e. it has the diversification drawback due to a skewed weight distribution, as discussed above), but it is not tilted toward large stocks.

I don't understand what this means.  How can a portfolio have the same weightings as the market cap weighted index and not be tilted towards large stocks?  Doesn't the former by definition imply the latter?

The set of weightings is the same as the actual index, and the set of stocks is the same, but the two are mixed up. For example, in the real index, Apple's weight is 3.5% or whatever. In his example, that 3.5% is instead assigned to Chipotle, and Chipotle's actual weight is assigned to CarMax, and so on...

In that example, he actually demonstrates the opposite of mizzou's belief: "it is not the tilt toward large stocks that hurts the market portfolio", since the portfolio with random companies getting the big weightings performs even worse than the real one where the big companies get the big weightings.
I'm not sure where you got the idea that the only reason I believe a market cap weighted fund is sub-optimal is because it weights the highest the best. I think that is one reason, it is rewarding past performance for expected future performance. But I would also expect randomly assigning an S&P company a weighting of 300x another S&P company to be equally as poor with more volatility. I actually am not totally against a market cap weighting strategy, but I think a more balanced strategy makes sense.
Perhaps something like this:

Stocks 1-25: 3x median
Stocks 26-100: 2x median
Stocks 101-250: median
Stocks 251-400: 2x smaller than median
Stocks 401-500: 3x smaller than median

I just find it odd that people who are devout index fund investors that are essentially saying "we don't know, so the best thing to do is own it all" are comfortable with an allocation that is essentially taking strong guesses that Apple will outperform any other stock every year and that "guess" continues to grow.

That's the key mis-perception of index fund investing. We're not saying we know or don't know. We're saying the market is a representative force of everyone's knowledge of the market. And that this aggregate knowledge of the market provides the best returns considering cost and risk.

If the market (meaning active traders and the like) tomorrow thinks that Apple is a poor investment regardless of the reason and start selling which brings about a drop in the price of the stock then so be it.

We're more saying that the idea of "knowing" is already there present in the price. If you think you're detecting some inefficiency in the market because of capitalization and can act on it then so be it. We're saying that that inefficiency probably isn't there or cannot be acted upon in a cost efficient manner without taking on some risk we're unwilling to take on.

Consider that (more) equal weighted portfolio will be weighted (more) heavily towards small and mid cap companies. That portfolio will have more turnover, forcing more trades, thereby increased costs.

forummm

  • Walrus Stache
  • *******
  • Posts: 7389
  • Senior Mustachian
Re: Are Traditional Index Funds Truly Optimal
« Reply #45 on: June 03, 2015, 07:27:52 AM »
The market is saying that people believe Apple will provided twice the risk-adjusted gains going forward as Microsoft. So it's going to cost twice as much to buy 1% of Apple as to get 1% of Microsoft. Cap-weighted indexing is just getting a certain share of each company based on it's expected future profits. If we believe, but aren't certain, Apple is going to provide twice the risk-adjusted gains going forward, we'd want twice as much of our money going into it. This is using the market to allocate our dollars.

hodedofome

  • Handlebar Stache
  • *****
  • Posts: 1314
  • Age: 40
  • Location: Texas
Re: Are Traditional Index Funds Truly Optimal
« Reply #46 on: June 03, 2015, 08:07:39 AM »
If you truly believed the market knows best, then you should have a global allocation to roughly 43% stocks and 57% bonds. Any deviation from that and you are making an active bet against 'the market.' - Cullen Roche



I'd almost put money that nobody on this thread advocating passive investing has this allocation. All of you are active whether you admit it or not. You're just using index funds as your vehicle of choice.

To argue against non-cap weighted indexes while not going with the market's asset allocation is irrational, IMO.

matchewed

  • Magnum Stache
  • ******
  • Posts: 4376
  • Location: CT
Re: Are Traditional Index Funds Truly Optimal
« Reply #47 on: June 03, 2015, 08:23:25 AM »
If you truly believed the market knows best, then you should have a global allocation to roughly 43% stocks and 57% bonds. Any deviation from that and you are making an active bet against 'the market.' - Cullen Roche

*snip*

I'd almost put money that nobody on this thread advocating passive investing has this allocation. All of you are active whether you admit it or not. You're just using index funds as your vehicle of choice.

To argue against non-cap weighted indexes while not going with the market's asset allocation is irrational, IMO.

Eh that's been discussed before, and is closer to a no true scotsman argument than anything.

I think the difference between active and passive investing is just intrinsically linked to how much you as an individual investor believe in the efficiency of the market you're investing in. You could be passive in one market and allow yourself to have market level returns or you could be actively pursuing what you think is exploitable inefficiencies and well be active...

The market itself is rather irrelevant to whether it is active or passive. It is the level which you are trying to take advantage of perceived inefficiencies.

Now in the article the claim is being made that the global index is the "only true passive benchmark" which is further reinforced with the conclusion -
Quote
In summary, investors are starting to acknowledge the overwhelming evidence that active security selection is a loser’s game. This realization has caused a massive exodus from traditional mutual funds and Separately Managed Accounts and into passive Exchange Traded Funds. Investors who choose to follow this trend face a new set of challenges related to the expression of a passive view in their asset allocation. The Global Market Portfolio represents the most coherent expression of this view, and any deviation from this portfolio represents an active bet. Thus most investors who think they are passive are actually active; worse, they are making large concentrated bets unintentionally.

Now I think the difference is that the conclusion is assuming I don't know that using VTSAX tilts in a particular asset class and ignores other asset classes. And this somehow means I'm "active". The thing is I don't think there is some mysterious inefficiency I'm exploiting. I'm just comfortable with the level of risk and anticipated (conservatively) return. That doesn't make me an active investor and I'm no more active than someone who would be buying into that global market ETF strategy that the author is proposing.

Maybe the company behind the post have an interest in global ETF's?  :) Not that that is a bad thing. It just sounds a little silly to say "I'm more passive than you because I choose X asset class."

Note that this is also from a similar discussion of "you're not truly a market investor unless you invest in X market".

hodedofome

  • Handlebar Stache
  • *****
  • Posts: 1314
  • Age: 40
  • Location: Texas
Re: Are Traditional Index Funds Truly Optimal
« Reply #48 on: June 03, 2015, 08:36:26 AM »
It does make you an active investor whether you'd like to admit to it or not. But IMO people need to get away from the dogma of being a passive investor. We're all active investors here apart from those who invest in the global financial portfolio, the only ones arguing against that are those who are committed to the Bogle belief system and have ingrained into their heads that anything with the world 'active' in it is bad, and anything with the word 'passive' in it is good. He's just a man, not an investing god, and we can take the good things he's said and apply them, and throw out what is silly.

Let's all admit we're active investors here making giant macro bets as to what asset classes and what countries are going to do the best going forward. Then let's move on to admitting that cap weighted index funds' main advantage is costs. And then go from there.

beltim

  • Magnum Stache
  • ******
  • Posts: 2964
Re: Are Traditional Index Funds Truly Optimal
« Reply #49 on: June 03, 2015, 08:40:51 AM »
The market is saying that people believe Apple will provided twice the risk-adjusted gains going forward as Microsoft. So it's going to cost twice as much to buy 1% of Apple as to get 1% of Microsoft. Cap-weighted indexing is just getting a certain share of each company based on it's expected future profits. If we believe, but aren't certain, Apple is going to provide twice the risk-adjusted gains going forward, we'd want twice as much of our money going into it. This is using the market to allocate our dollars.

In nominal terms, yes.  But in percentage terms Apple and Microsoft should return the same risk-adjusted returns. 

Furthermore, your comment doesn't make sense - why do you stop at market cap weighting if you think that market cap is proportional to future return?  Why wouldn't you just buy the largest stock and be done with it?  Or the largest 30 for sufficient diversification?