Market Capitalization

[eufo]

Ok, this is a thought that keeps rattling around in my head and it's making me rethink a few things.

When we invest in an indexed ETF (like SPY), our money is skewed towards the equities that have the most market capitalization. I'll use IVV (Large Cap S&P 500), IJH (Mid Cap S&P 400), and IJR (Small Cap S&P 600) for my examples, because they have no overlap.

If we invest $10,000 in IVV, we're going to have about $363 in Apple and about $1 in Urban Outfitters. Now we invest $10,000 in IJH, and have about $63 in AMD and $4 in Tootsie Rolls. Now we see the oddity that's happening. We're more invested in ALL of the various S&P 400 stocks than we are in the lower end of the S&P 500 stocks. When we invest $10,000 in IJR, it gets worse. Now we have $82 in Take Two!

There's nothing inherently wrong with the scenario, I suppose, but it makes the investments seem rather arbitrary. If a stock moves from the S&P 500 down to the S&P 400, we're suddenly likely to increase our exposure by many tens of times. Conversely, if a stock moves up between indexes we're going to suddenly own just a fraction of what we used to.

It does seem to pan out better to own all three of these than just the total stock market, but it makes me question the idea of using these indexes at all. Perhaps market capitalization isn't the right way to be invested.

Is anyone else disturbed by this stuff or am I just a crazy man?

Also, Happy Thanksgiving, everyone!

Edit: My fuzzy mind is hard to understand and put into words, but I've tried to clean it up a little.

[Tortoise]

I think the flaw in your example is that you’re investing equal dollar amounts in the three indexes, but the three indexes represent very different total market capitalizations.

Weight your investment in each index by its total market capitalization, and the “paradox” disappears and you’ve basically recreated a total stock market index fund.

[ochotona]

Eufo, your concerns about market cap weighting are real, they have been addressed, there was a thought-provoking podcast with Rob Arnott of Research Affiliates, here is the link.

http://mebfaber.2.cavendoclient.com/201 ... ectations/

Alternative weighting schemes sound like a great idea... but in their short lives, they have not been outperforming. I suspect the passive money is just slamming into market cap weighted products, and propelling them higher in a blind kind of way. Maybe this won't last forever.

As I am a Schwab customer, I have been watching the market-cap and Research Affiliates "Fundamental" weighted equivalent ETFs side-by-side, but I see no reason to change yet. Compare SCHX with FNDX, SCHB with FNDB, SCHF with FNDF, SCHA with FNDA, SCHC with FNDC, and SCHE with FNDE. You get some better "FN" performance with small cap stocks, not so much for large cap. They are more costly.

But the basic idea is important... market-cap is a decision. It is not arbitrary, nor natural. A committee decides who is in the S&P500. Do you think that might be a little bit political?

[eufo]

I think the flaw in your example is that you’re investing equal dollar amounts in the three indexes, but the three indexes represent very different total market capitalizations.

Weight your investment in each index by its total market capitalization, and the “paradox” disappears and you’ve basically recreated a total stock market index fund.

I wouldn't call it a flaw. To do what you're suggesting is pretty much out of the realm of what a person might/can do. To make it happen, you'd need to invest $10,000 in IVV, $158 in IJH, and $8 in IJR.

[Tortoise]

I wouldn't call it a flaw. To do what you're suggesting is pretty much out of the realm of what a person might/can do. To make it happen, you'd need to invest $10,000 in IVV, $158 in IJH, and $8 in IJR.

I didn’t mean to imply that it was a realistic strategy for an investor. To clarify: Your hypothetical example leads to strange results not because of an inherent problem or inconsistency in the concept of market cap weighting, but because your example is a hybrid between market cap weighting (each index is market cap weighted) and equal weighting (you invest an equal amount in each index).

By contrast, if you use pure market cap weighting by weighting your investment in each index by its total market cap, then there is no weird behavior when a stock moves out of one index and into another one. Your investment in that one company will change only slightly.

Hopefully I correctly understood your original question and this at least partly addresses it.

[eufo]

Eufo, your concerns about market cap weighting are real, they have been addressed, there was a thought-provoking podcast with Rob Arnott of Research Affiliates, here is the link.

http://mebfaber.2.cavendoclient.com/201 ... ectations/

Alternative weighting schemes sound like a great idea... but in their short lives, they have not been outperforming. I suspect the passive money is just slamming into market cap weighted products, and propelling them higher in a blind kind of way. Maybe this won't last forever.

As I am a Schwab customer, I have been watching the market-cap and Research Affiliates "Fundamental" weighted equivalent ETFs side-by-side, but I see no reason to change yet. Compare SCHX with FNDX, SCHB with FNDB, SCHF with FNDF, SCHA with FNDA, SCHC with FNDC, and SCHE with FNDE. You get some better "FN" performance with small cap stocks, not so much for large cap. They are more costly.

But the basic idea is important... market-cap is a decision. It is not arbitrary, nor natural. A committee decides who is in the S&P500. Do you think that might be a little bit political?

That's a great link, ochotona. It actually goes beyond what I expressed (poorly, I might add) and delves into another worry of mine... that market cap investments are essentially based on momentum... Bet more on the winners and less on the losers.

To clarify: Your hypothetical example leads to strange results not because of an inherent problem or inconsistency in the concept of market cap weighting, but because your example is a hybrid between market cap weighting (each index is market cap weighted) and equal weighting (you invest an equal amount in each index).

I understand what you mean, but I think it's a fairly common scenario for someone to be invested equally among different market cap weighted indexes. I know I am. I currently own equal parts IJH and IJR... decided to skip on IVV.

Also, based on the video linked by ochotona, I'm thinking I personally feel that market cap weighting IS inherently the wrong way to invest for me, but not based on my initial thoughts expressed on this thread.

[Tortoise]

Interesting podcast. If the fundamental-weighted indexes outperformed market cap-weighted indexes fairly consistently over, say, most 10-year periods I would seriously consider switching to fundamental weighting. But the proven ETF track record just isn’t there yet, so I’ll stick with market cap weighting for now.

[Kbg]

I have to call out Ochotona on a small error...RSP has done quite nicely compared to normal mkt cap SPX funds even with higher fees. Nothing complex factor wise that may not work going forward.

[ochotona]

I have to call out Ochotona on a small error...RSP has done quite nicely compared to normal mkt cap SPX funds even with higher fees. Nothing complex factor wise that may not work going forward.

YES, you're right. RSP has been very good. I was looking at FNDX, and don't see any uplift from that one yet.

[boglerdude]

IIRC, according to Asness, alternative weighting like equal weight gives you a small+value tilt. Which might continue to outperform...until too many people do it.

Vanguard's small value has a PE of 34
http://www.etf.com/VBR

[Kbg]

IIRC, according to Asness, alternative weighting like equal weight gives you a small+value tilt. Which might continue to outperform...until too many people do it. http://www.etf.com/VBR

For RSP, it is more like a mid-cap tilt. I read somewhere that one should just invest in something like IJH if they were inclined toward RSP. Since inception and at the 10 year lookback IJH has outperformed RSP while the past 5 years RSP has outperformed. Risk measures alternate lead horses depending on look back.

I think until late bull markets (like now), generally mid-caps are a better investment if one doesn't want to deal with the variability of small caps. Mid-caps aren't as flaky performance wise and tend to track the larger indexes' overall paths.

My view on equal weight indexes is that they exploit mean reversion vs. value (though value may be an aspect as well) while a market cap index exploits trend. Both are solid principles to invest/trade on. Market caps almost without fail tank harder than equal weights due to the trend effect when markets head south. That is because the most overbought stocks are also normally going to be the most oversold and by definition they are most of your market cap weight right before the fall. A simple technique to exploit the relative outperformance of one over the other (assuming always long the market) is to just use a ratio of the two with perhaps some type of band; go long the numerator when above the band, go long the denominator below the band and go 50/50 in the band.