Small cap value index: S&P 600 vs Russell 2000

[sophie]

I'd never thought about this before, but it appears there is a BIG difference between small cap value funds depending on which of these indexes you use:

https://www.barrons.com/articles/a-tale ... 1516896321

Between 1994 and 2018, the S&P 600 beat the Russell 2000 by nearly two percentage points annually (10.2% vs 8.3%). That's pretty significant. It has to do with extra criteria that S&P uses to weed out poorly performing stocks. The Russell index doesn't do this. So there appears to be a systemic reason to expect that this difference in performance will continue, although it is rather surprising that the simple S&P algorithm should result in that big of a difference.

I've been using funds that - as it turns out - track the Russell 2000 index:

VISVX/VSIAX - Vanguard (ER 0.07)
VBR - Vanguard ETF (ER 0.07)
FISVX - Fidelity's new SCV mutual fund - ER 0.05

Whereas these funds track the S&P 600:
VSMVX/VSMAX - Vanguard, ER 0.05%
IJS - ETF, ER 0.25%
SLYV - ETF, ER 0.15%

OOPS!!!!!

Question for Tyler: which index did you use for backtesting the Golden Butterfly??

[Tyler]

Question for Tyler: which index did you use for backtesting the Golden Butterfly??

My data does not come from any specific index. I take source academic data published by Ken French and use it to reconstruct a real-world index using the same methodology as CRSP and MSCI. You can read more about the methodology and how various different indices work here: https://portfoliocharts.com/stock-index-calculator/

Here's a comparison of different SCV indices from that link. Every index is different! Rather than obsess about the "best" index, I think it's important to understand that there will always be some uncertainty involved.



BTW, your info on VSIAX and VBR is out of date. Vanguard switched indices a few years ago, and they now track a CRSP index. So the PC data should (theoretically) track the Vanguard data pretty well, although even then individual index funds also layer on additional trading considerations (bands, filters, etc) that are impossible for me to fully account for.

[sophie]

Thanks Tyler!

I should have known this was just another cherrypicking/hindsight article. I played around with some backtesting on ETFreplay, and over time all the indices are pretty consistent. Not too surprising that 2 percentage point differences can arise if you pick the start and end dates carefully.

Oddly, I started investigating after seeing a Bogleheads thread advising against using Fidelity's FISVX fund because of the inferiority of the Russell 2000 index. I really wanted to like that fund since mutual funds are infinitely easier to deal with than ETFs. At least for me.

[Kbg]

Question for Tyler: which index did you use for backtesting the Golden Butterfly??

My data does not come from any specific index. I take source academic data published by Ken French and use it to reconstruct a real-world index using the same methodology as CRSP and MSCI. You can read more about the methodology and how various different indices work here: https://portfoliocharts.com/stock-index-calculator/

Here's a comparison of different SCV indices from that link. Every index is different! Rather than obsess about the "best" index, I think it's important to understand that there will always be some uncertainty involved.



BTW, your info on VSIAX and VBR is out of date. Vanguard switched indices a few years ago, and they now track a CRSP index. So the PC data should (theoretically) track the Vanguard data pretty well, although even then individual index funds also layer on additional trading considerations (bands, filters, etc) that are impossible for me to fully account for.

Tyler...this is a huge point. I probably need to go dig out my reference but if IIRC correctly CRSP SCV is either smallest 10% or 20% of all stocks and as such is in fact not investible via an ETF or mutual fund because the mkt cap is just too small for commercial packaging.

In other words, none of the commercial SCV funds actually use the academic measurement.

Can you comment?

[Kbg]

Sophie...yep. This is known colloquially as the S&P 600 small cap crap filter...personally I won't touch the Russell.

[Tyler]

Tyler...this is a huge point. I probably need to go dig out my reference but if IIRC correctly CRSP SCV is either smallest 10% or 20% of all stocks and as such is in fact not investible via an ETF or mutual fund because the mkt cap is just too small for commercial packaging.

In other words, none of the commercial SCV funds actually use the academic measurement.

Can you comment?

CRSP and MSCI both define small cap as the bottom 15% of the market sorted by capitalization. That's actually quite normal and to my knowledge isn't a problem, although I do think they tend to cut off the very smallest microcaps that aren't doing much anyway. The differences I'm talking about are more along the lines of trading bands that prevent them from immediately selling a company the moment it grows out of the small cap designation (they hold onto it for a while to make sure the index doesn't pay double trading fees for no reason if it falls again). That's why most small cap funds usually have more mid-caps than you'd expect. They don't buy mid-caps, but they also don't immediately sell them.

[Kbg]

That's why most small cap funds usually have more mid-caps than you'd expect. They don't buy mid-caps, but they also don't immediately sell them.

...and for investing in the real world, this is really what I was driving at. You can't actually get the returns of academic small cap and if you use them then they are overstated from what is practically achievable. "you" is meant in a generic way, but for those following the thread I think it is not unfair nor inaccurate to say that the GB's performance is overstated. Personally, I don't pay any attention to small cap historical returns until after small cap index funds came into existence and as Tyler noted, most small cap funds aren't what academically is small cap.

Side note...the index fund comment also applies to pretty much anything. Without fail there is a decrease in stated historical performance once something is commercialized. The only real exception(s) are the huge markets...which were probably pretty efficient already due to their size.

[Tyler]

...and for investing in the real world, this is really what I was driving at. You can't actually get the returns of academic small cap and if you use them then they are overstated from what is practically achievable. "you" is meant in a generic way, but for those following the thread I think it is not unfair nor inaccurate to say that the GB's performance is overstated.

Fair enough, but I also have mid cap data that you can add in the same proportion as a real-world small cap fund to see how that affects the numbers.

[Kbg]

That would be interesting...

Also, I truly hope my comments are being received in a spirit of just peeling back the layers on a subject to illuminate subtleties that may not be readily apparent.

I am nothing but appreciative of both your site and your contributions to this forum.

[Tyler]

No worries! I appreciate your contributions and have zero problem with a healthy discussion about this kinda thing. I certainly don't claim to have all the answers.

[Smith1776]

No worries! I appreciate your contributions and have zero problem with a healthy discussion about this kinda thing. I certainly don't claim to have all the answers.

I love your community contributions, Tyler! You've always got this great balance between being authoritative but also understanding that different people have different views. Pretty rare these days -- especially on the Internet.

Forgive me if you've said this before, but what is your current portfolio allocation? Are you GB or PP? Or something else entirely?

[Tyler]

I love your community contributions, Tyler! You've always got this great balance between being authoritative but also understanding that different people have different views. Pretty rare these days -- especially on the Internet.

Forgive me if you've said this before, but what is your current portfolio allocation? Are you GB or PP? Or something else entirely?

Thanks! I just do my best to support good conversations, and I'm thankful that it's pretty easy here.

These days I use a slightly modified GB with 10% ST bonds and 10% REITs. Nothing crazy different, but my personal thought process tends to slowly evolve every time I get new data.

[vnatale]

I love your community contributions, Tyler! You've always got this great balance between being authoritative but also understanding that different people have different views. Pretty rare these days -- especially on the Internet.

Forgive me if you've said this before, but what is your current portfolio allocation? Are you GB or PP? Or something else entirely?

Thanks! I just do my best to support good conversations, and I'm thankful that it's pretty easy here.

These days I use a slightly modified GB with 10% ST bonds and 10% REITs. Nothing crazy different, but my personal thought process tends to slowly evolve every time I get new data.

Just catching up to this one....are the two 10%'s above what you are using for the "classic" 20% cash (for Golden Butterfly)? The remaining 80% being equally divided between - Total Stock Market, Small Cap Value, Long Term Treasuries, Gold?

Vinny

[Tyler]

Just catching up to this one....are the two 10%'s above what you are using for the "classic" 20% cash (for Golden Butterfly)? The remaining 80% being equally divided between - Total Stock Market, Small Cap Value, Long Term Treasuries, Gold?

Yep. I don't like to make a big deal about how I personally invest, as everyone is different and I also reserve the right to change my mind over time. But that's where I'm at right now.