Why Market Cap Weighting of Index Funds?

[TripleB]

One investing question that I never figured out the answer is why do we use market cap weighting on index funds? I've asked on Bogleheads a few years back and wasn't impressed by the answer. The PP uses the same type of indexing, so it's worth discussing here.

The basic few answers I've gotten to this question are:

1) Using market-cap weighting index style lets you remove emotion from investing because you are just buying the market

2) Using this lets you capture the return of the market, and 80% of active trading managers fail to beat the market so at least settle for what the market returns

3) It's cheaper on an operational level because you aren't buying and selling stocks everyday depending on their price. You are buying and holding, so the expense ratios are lower.

and the answer that seems the ridiculous yet the most pragmatic at the same time:

4) There's no better options

There's an equal-share SP500 index that puts 0.20% of money into each of the SP500 stocks. One downside is the expense ratio is pretty high because they don't have as much assets as a regular SP500 index fund.

Another downside is that now you are "overweighting" smaller companies relative to bigger companies, in comparison to the SP500 market weighted index. For example, you have equal amounts invested in Netflix as you do Apple. Whether this is good or bad is up to you to decide.

The market cap weighting system seems highly anti-PP-like (even though I know HB recommended it). In the market cap weighted system, if shares of a company appreciate relative to the overall index, then you hold more of it. If they lose value relative to the index, then you hold less of it.

The PP, on the other day, functions to sell off winners to buy losers. If Apple doubles in share price relative to the index, the market cap system would then have you holding twice as much Apple at that point, until Apple drops 50% and then you hold half as much. The PP system is designed so that if Apple were to double, you would sell off Apple and buy a "loser" stock.

I suppose one difference between looking at this from a stock index perspective as opposed to a total PP is that the PP is partially designed to have reversion to the mean. Such that if Gold drops, you buy more gold because gold eventually has to go back up, or at the very least stay the same. Gold can't (or is highly unlikely to) drop to zero. The same can't be said of individual stocks.

Thus, in theory, if Apple rises relative to Netflix, and we sell Apple to buy Netflix, we could eventually rebalance everything into the abyss and wind up with zero if Netflix drops to zero. However, if using the SP500 Equal-Weighed Index, in theory, the stock will be removed from the index before it falls to zero. That does offer some protection against abyss-making, but the stock could lose a lot of money before getting removed from the index.

Then again, at 0.2% per stock in the SP500, it seems unlikely that any one or two stocks falling to zero would really hurt the overall stock position too much.

Over the last few years, the Equal Weight SP500 index has beaten the market weighted one, but that doesn't mean anything going forward. Since the entire SP500 is highly correlated to the overall economy, there won't be too much deviation in either direction. My big concern is that while I feel the equal weight index will win, the expense ratios of 0.5% plus reduced liquidity in the ETFs will wipe out any marginal gains.

I don't know if equal-weighted indexing is better than market cap, but I don't understand why market cap is used, and fail to believe the answer of "there's nothing better" is good enough.

[stone]

For all the reasons you mention I stear clear of index funds/etfs. IMO most (and an increasing proportion) of potential stock market returns come in the form of stock prices moving up and down relative to other stocks just as stocks as a whole move up and down relative to LTT, gold and cash. PP harvests such volatility from stocks as an asset class so why the slavish bogglehead-like belief in index funds that let all such potential return pass by unharvested? If rebalancing is accepted as a neccessity between asset classes, then why ignore it between stocks? 

[TripleB]

For all the reasons you mention I stear clear of index funds/etfs. IMO most (and an increasing proportion) of potential stock market returns come in the form of stock prices moving up and down relative to other stocks just as stocks as a whole move up and down relative to LTT, gold and cash. PP harvests such volatility from stocks as an asset class so why the slavish bogglehead-like belief in index funds that let all such potential return pass by unharvested? If rebalancing is accepted as a neccessity between asset classes, then why ignore it between stocks? 

Please discuss your proposed alternative.

[stone]

TripleB, in practice what I do is hold three stock investment trusts (closed end managed funds) as in:
http://gyroscopicinvesting.com/forum/ht ... ic.php?t=6
If my permenant portfolio was larger I might be tempted to hold ten individual stocks at 2.5% of the total portfolio each as I wittered on about in:
http://gyroscopicinvesting.com/forum/ht ... 727#p14727

I might be being naive about this. It just seems to me that with companies now using buybacks and share issues rather than paying much dividends, most of  the possible profits just wriggle past an index fund. A company that has a lifetime of thirty years, and never pays a dividend, would return nothing to an index fund that held it for the whole thirty year period even if it formed an Apple or Exxon sized chunk at its peak.

[Indices]

The key is cost. If what the OP does can capture market returns reasonably at a lower cost than an index fund, it will work. If it can't, then in all likelihood it won't. Market cap weighting is generally the lowest cost and simplest way of catching an asset class' returns.

[TripleB]

I might be being naive about this. It just seems to me that with companies now using buybacks and share issues rather than paying much dividends, most of  the possible profits just wriggle past an index fund. A company that has a lifetime of thirty years, and never pays a dividend, would return nothing to an index fund that held it for the whole thirty year period even if it formed an Apple or Exxon sized chunk at its peak.

Wouldn't you have capital appreciation, which is superior to dividends because you control when you take the tax hit?

[TripleB]

The key is cost. If what the OP does can capture market returns reasonably at a lower cost than an index fund, it will work. If it can't, then in all likelihood it won't. Market cap weighting is generally the lowest cost and simplest way of catching an asset class' returns.

I can come up with a bunch of ways to invest in stocks that have an even lower cost than market-cap indexing. Low cost doesnt = good investment strategy.

I agree that taking two comparable strategies then using cost as a secondary differentiator is a good idea.

[Indices]

The key is cost. If what the OP does can capture market returns reasonably at a lower cost than an index fund, it will work. If it can't, then in all likelihood it won't. Market cap weighting is generally the lowest cost and simplest way of catching an asset class' returns.

I can come up with a bunch of ways to invest in stocks that have an even lower cost than market-cap indexing. Low cost doesnt = good investment strategy.

I agree that taking two comparable strategies then using cost as a secondary differentiator is a good idea.

I'm curious as to what those are. If you can find a way that is cheaper than indexing, I would use it in my portfolio.

I disagree with you that low cost is not a good investment strategy. The lower the costs the greater the chance of success against a comparable fund investing in the same things with a higher cost.

[stone]

I might be being naive about this. It just seems to me that with companies now using buybacks and share issues rather than paying much dividends, most of  the possible profits just wriggle past an index fund. A company that has a lifetime of thirty years, and never pays a dividend, would return nothing to an index fund that held it for the whole thirty year period even if it formed an Apple or Exxon sized chunk at its peak.

Wouldn't you have capital appreciation, which is superior to dividends because you control when you take the tax hit?

How do you get benefit from capital appreciation that subsequently reverses whilst you still hold the fund? Imagine you have an SP500 index fund and Apple enters the index back in 199?; Apple grows to be 10% of the fund or whatever; it then gradually withers away to being 1% then 0.1% and then leaves the index. You haven't gained a single cent from the entire history of Apple. Stocks are like 30year amortizing bonds. If they don't pay a dividend, then they are 30year zero coupon amortizing bonds. Holding a ladder of 30year zero coupon amortizing bonds to maturity will not provide you with anything. Only by actively rebalancing or trading in and out of individual stocks will you get anything.

[pershing83]

I think it worth noting in the WSJ this AM the larges pure stock and asset allocation funds were listed as to performance. It was a page full. If you left out gold and the QQQ's it appeared to me and you had to search this out, PRPFX had the best return for past 1,3 and 5 years. this included indexes, S&P all of them. The funds were not listed as to best or worse so I had to look at all,  so I could be wrong.

[Indices]

I might be being naive about this. It just seems to me that with companies now using buybacks and share issues rather than paying much dividends, most of  the possible profits just wriggle past an index fund. A company that has a lifetime of thirty years, and never pays a dividend, would return nothing to an index fund that held it for the whole thirty year period even if it formed an Apple or Exxon sized chunk at its peak.

Wouldn't you have capital appreciation, which is superior to dividends because you control when you take the tax hit?

How do you get benefit from capital appreciation that subsequently reverses whilst you still hold the fund? Imagine you have an SP500 index fund and Apple enters the index back in 199?; Apple grows to be 10% of the fund or whatever; it then gradually withers away to being 1% then 0.1% and then leaves the index. You haven't gained a single cent from the entire history of Apple. Stocks are like 30year amortizing bonds. If they don't pay a dividend, then they are 30year zero coupon amortizing bonds. Holding a ladder of 30year zero coupon amortizing bonds to maturity will not provide you with anything. Only by actively rebalancing or trading in and out of individual stocks will you get anything.

You could make the same argument with gold. No dividends, no interest. Obviously worthless. Except it's not.

[stone]

Indices "You could make the same argument with gold. No dividends, no interest. Obviously worthless. Except it's not."

For me, a major point of holding gold is the rebalancing against stocks and LTT. Lets take silver as another example. If you had bought silver last winter at $30/ounce and kept it until now (when it is back down to the same price), then that would have given you no benefit. If you had maintained a 50%silver:50%cash portfolio, then the rebalancing benefit would have given you a return. Imagine something that halves and then doubles in price every year. You could hold that for ever and never gain anything. If you rebalanced against cash then you would gain 12.5% per year. An index fund might make sense if all individual stocks moved in perfect unison. In reality they do not and so an index fund lets potential gains slip by. The companies in the SP500 have made trillions of dollars in profits over the past dozen years and yet someone holding an index fund for that period has seen none of that. Those profits have been returned to shareholders but not in a way that buying and holding an index fund is able to capture.

[stone]

Clive "I suspect part of the reason why equal weighted portfolio's relatively outperform is that you're more exposed to low correlation motions. Ten different holdings of 1% weighting each is likely to have some low/inverse correlations between individual holdings over time. In contrast an index that contains one stock that amounts to 10% of the total portfolio in effect has 10 x 1% holdings that all have 1.0 (perfect) correlation and as such miss out of potential low correlation/rebalance trading benefits."

I think it is more dramatic than that. A market cap index does not trade at all except when a stock enters or leaves the index, or to mirror share buybacks, issues of new stock etc. I guess that it is the quarterly rebalancing of the equal cap funds that gives them the out performance.