Houston, we have a problem... [Small cap historical performance]
- privatefarmer
- Posts: 779
- Joined: Mon Sep 08, 2014 2:45 pm
Houston, we have a problem... [Small cap historical performance]
I am a small/value "tilter". I have believed the evidence presented by Fama/French.
However, can someone answer me this: when you compare Vanguard's small cap blend index (NAESX) to vanguards s/p 500 index (VFINX), going back nearly 30 years, on portfoliovisualizer.com the s/p 500 DESTROYS the small cap index over that 28-year period.
When you look at what the index's "should" have done, according to porftoliovisualizer, there basically should've been a tie over that period.
They both are nearly identical in expenses. NAESX is not pure small, it is a mix of mid-cap and small-cap, but when you account for that in portfoliovisualizer the mid/small mix should've actually beaten the large blend by a solid 1%/yr over the '86-'14 period.
I am starting to suspect that perhaps the small/value "factors" work in theory only. When you actually invest in index funds/ETFs that track these they don't do what you would expect them to do. Any thoughts?
However, can someone answer me this: when you compare Vanguard's small cap blend index (NAESX) to vanguards s/p 500 index (VFINX), going back nearly 30 years, on portfoliovisualizer.com the s/p 500 DESTROYS the small cap index over that 28-year period.
When you look at what the index's "should" have done, according to porftoliovisualizer, there basically should've been a tie over that period.
They both are nearly identical in expenses. NAESX is not pure small, it is a mix of mid-cap and small-cap, but when you account for that in portfoliovisualizer the mid/small mix should've actually beaten the large blend by a solid 1%/yr over the '86-'14 period.
I am starting to suspect that perhaps the small/value "factors" work in theory only. When you actually invest in index funds/ETFs that track these they don't do what you would expect them to do. Any thoughts?
- Phineas J. Whoopee
- Posts: 9675
- Joined: Sun Dec 18, 2011 5:18 pm
Re: Houston, we have a problem...
The claim goes that the factors have worked (which is to say, today's backtesting shows they would have worked) over the last 70 to 80 years (which is around as long as we have reliable and detailed stock price records).
Assuming the claim is true, a period of three decades is a mere pittance. Of course one must hold on longer. Anybody who bails out after such a small interval of underperformance is no true factor investor.
Oh, and one more thing: most investors who are building their own portfolios out of their own efforts don't live long enough to significantly invest for the better part of a century.
PJW
Assuming the claim is true, a period of three decades is a mere pittance. Of course one must hold on longer. Anybody who bails out after such a small interval of underperformance is no true factor investor.
Oh, and one more thing: most investors who are building their own portfolios out of their own efforts don't live long enough to significantly invest for the better part of a century.
PJW
Re: Houston, we have a problem...
This always puzzled me a bit. I have a relatively long time horizon of 40+ years. Some of the information I read uses numbers that seem to assume that I have 50+ years to loaf around. Nonetheless it is interesting information, I'm just not to sure what to do with it sometimes.Phineas J. Whoopee wrote:The claim goes that the factors have worked (which is to say, today's backtesting shows they would have worked) over the last 70 to 80 years (which is around as long as we have reliable and detailed stock price records).
Assuming the claim is true, a period of three decades is a mere pittance. Of course one must hold on longer. Anybody who bails out after such a small interval of underperformance is no true factor investor.
Oh, and one more thing: most investors who are building their own portfolios out of their own efforts don't live long enough to significantly invest for the better part of a century.
PJW
-Joey
- privatefarmer
- Posts: 779
- Joined: Mon Sep 08, 2014 2:45 pm
Re: Houston, we have a problem...
Thanks. But that still doesn't answer why the actual index funds (Vang small cap vs vanguard S/P 500) differ from what the "historical data" says. If you backrest just small vs large over the last 30 years they basically should end at a tie. But if you compare vanguard small index fund vs large index fund, the large fund majorly out performs.Phineas J. Whoopee wrote:The claim goes that the factors have worked (which is to say, today's backtesting shows they would have worked) over the last 70 to 80 years (which is around as long as we have reliable and detailed stock price records).
Assuming the claim is true, a period of three decades is a mere pittance. Of course one must hold on longer. Anybody who bails out after such a small interval of underperformance is no true factor investor.
Oh, and one more thing: most investors who are building their own portfolios out of their own efforts don't live long enough to significantly invest for the better part of a century.
PJW
So the real question is: can the index funds which supposedly track these factors actually capture the "premium" or are we all just wasting our time?
- Phineas J. Whoopee
- Posts: 9675
- Joined: Sun Dec 18, 2011 5:18 pm
Re: Houston, we have a problem...
There is no consensus about what constitutes large-, mid-, and small-cap, and that's one of the problems. Somebody could, I'm not accusing anybody of, but somebody could choose thresholds to make the backtesting work. I accuse nobody of malfeasance in this area. Still, the possibility reduces my confidence in their results. Of course, with my low confidence and too-short likely remaining lifetime to benefit even if the results predict the future, I've chosen not to use factor tilts, but that's just me. I won't argue with anybody who, for their own reasons, makes the opposite decision.privatefarmer wrote:...
Thanks. But that still doesn't answer why the actual index funds (Vang small cap vs vanguard S/P 500) differ from what the "historical data" says. If you backrest just small vs large over the last 30 years they basically should end at a tie. But if you compare vanguard small index fund vs large index fund, the large fund majorly out performs.
I only argue with people who say statistics (for which they concede they do not have enough data even to calculate a t-stat) clearly show that if we all used the tilts, we could all beat the market.
I think, given the lack of consensus, that depends very much on what indices the index funds track. One good thing about total-market index funds, even if for practical reasons they must use sampling on the smaller end, is at least one knows what they're investing in and why.privatefarmer wrote:So the real question is: can the index funds which supposedly track these factors actually capture the "premium" or are we all just wasting our time?
Once again, to reiterate, I do not mean to pick any fights with people who make choices different than mine; and I do not work as an investment adviser, or investing book author, suggesting to others what they ought to do.
PJW
-
- Posts: 1007
- Joined: Sat Aug 06, 2011 7:01 am
Re: Houston, we have a problem...
NAESX hasn't always been an index fund....privatefarmer wrote: Thanks. But that still doesn't answer why the actual index funds (Vang small cap vs vanguard S/P 500) differ from what the "historical data" says.
Re: Houston, we have a problem... [Small cap historical performance]
I think that's the case that Jack Bogle made with the telltale chart--that the evidence doesn't support the benefit of tilting. http://www.vanguard.com/bogle_site/sp20020626.html
I have overweighted small cap in our portfolio and the return has been better than TSM over the last 10 years. But that has everything to do with the timing of contributions. It's evidence of nothing and just my personal data point.
I have overweighted small cap in our portfolio and the return has been better than TSM over the last 10 years. But that has everything to do with the timing of contributions. It's evidence of nothing and just my personal data point.
-
- Posts: 16022
- Joined: Thu Feb 22, 2007 7:28 am
- Location: St Louis MO
Re: Houston, we have a problem... [Small cap historical performance]
Unfortunately Vanguard's small cap fund used to be tied to the R2k, an awful index which underperformed by something like 2% a year because it was easily exploited
Here's data for live period 1/82-6/15 which is now over 30 years and of course includes trading costs and fund fees--
DFA Micro cap 12.26 vs 11.66 for S&P 500
Now using DFA small fund instead we have data from 4/92-6/15 so well over 20 years of live data and here you go
DFA Small 11.02, DFA Micro 11.89. S&P 500 9.45
Note R2k returned 9.63----showing again that choice of benchmarks and fund construction matters a great deal
Note these are all compound returns of course---risk premiums on other hand are always stated as ANNUAL AVERAGE returns, so the premium was larger than the gaps in annualized returns.
Would not logically we should expect the small premium to have fallen due to much improved liquidity, much lower bid offer spreads, which should lower the premium investors demand.
Hope that helpful
Larry
Here's data for live period 1/82-6/15 which is now over 30 years and of course includes trading costs and fund fees--
DFA Micro cap 12.26 vs 11.66 for S&P 500
Now using DFA small fund instead we have data from 4/92-6/15 so well over 20 years of live data and here you go
DFA Small 11.02, DFA Micro 11.89. S&P 500 9.45
Note R2k returned 9.63----showing again that choice of benchmarks and fund construction matters a great deal
Note these are all compound returns of course---risk premiums on other hand are always stated as ANNUAL AVERAGE returns, so the premium was larger than the gaps in annualized returns.
Would not logically we should expect the small premium to have fallen due to much improved liquidity, much lower bid offer spreads, which should lower the premium investors demand.
Hope that helpful
Larry
Re: Houston, we have a problem... [Small cap historical performance]
With respect to the difference between the fund and the data you're looking at, higher transaction costs and higher fund fees are to be expected in small caps.
Best to look at live fund data if possible, though in some cases the track record is pretty short.
Best to look at live fund data if possible, though in some cases the track record is pretty short.
- FrugalInvestor
- Posts: 6214
- Joined: Thu Nov 06, 2008 11:20 pm
Re: Houston, we have a problem...
See here....FillorKill wrote:NAESX hasn't always been an index fund....privatefarmer wrote: Thanks. But that still doesn't answer why the actual index funds (Vang small cap vs vanguard S/P 500) differ from what the "historical data" says.
viewtopic.php?t=35929
Have a plan, stay the course and simplify. Then ignore the noise!
Re: Houston, we have a problem... [Small cap historical performance]
I don't believe this applies to Vanguard's small value fund though. From the wiki;larryswedroe wrote:Unfortunately Vanguard's small cap fund used to be tied to the R2k, an awful index which underperformed by something like 2% a year because it was easily exploited
Hope that helpful
Larry
The Vanguard Small Cap, Small Growth and Small Value index funds have changed benchmark tracking indices two times over their history. The Small cap fund changed from Russell to MSCI indices and then to CRSP indices; the two style funds shifted from S&P indices to MSCI indices and then to CRSP indices.
Re: Houston, we have a problem... [Small cap historical performance]
And Vanguard's index regularly outperformed the Russell 2000 as well.larryswedroe wrote:Unfortunately Vanguard's small cap fund used to be tied to the R2k, an awful index which underperformed by something like 2% a year because it was easily exploited
The main reason for the poor performance of the Russell 2000 is that its construction rules caused it to do unusually badly in the Internet bubble; it missed many new tech stops on the way up and added them on the way down. The S&P 600, holding stocks in the same cap range, had construction rules which avoided that bubble. Thus, it performed much better than the Russell index in 2000-2002, while it didn't have the same advantage at other times. In 2008, for example, all small-cap indexes lost about the same amount; the next market crash may favor a different index.
- Rx 4 investing
- Posts: 735
- Joined: Sat Apr 25, 2015 11:03 am
Re: Houston, we have a problem... [Small cap historical performance]
A more recent development is throwing a wrench into your long-term strategy. Since the so-called "recovery" started in '09, interest rates have been falling. Usually in an economic recovery, interest rates start to rise, but the opposite has happened. Lower rates tend to benefit the companies that have the most debt, which happens to be the large caps. If this had been a "normal" recovery, small caps would have out-performed and your strategy would have most likely beared out. The falling interest rate environment has benefited the large caps significantly.I am starting to suspect that perhaps the small/value "factors" work in theory only. When you actually invest in index funds/ETFs that track these they don't do what you would expect them to do. Any thoughts?
I like the way Joel Greenblatt once explained why his version of value investing works. "It works because it does not always work." It's human nature to walk away from a sound strategy when it doesn't seem to work, but ironically that is usually just before it begins to outperform. Investors tend to chase performance, and the fund flow data shows the biggest inflows usually go into last year's out-performers If an investor is truly committed to a long- term "tilting" strategy, he/she has to understand there will be periods of under-performance.
A study which made an impression on me many years ago was Ibbotson's and Kaplan's paper: "Does Asset Allocation Policy Explain 40, 90, or 100 Percent of Performance? " The answer was "yes". I think the concern expressed by the OP is typical of investor behavior when less diversified strategies begin to under-perform. Probably to avoid the risk of behavioral mistakes, most of us should spread our bets across all capitalization sizes. We all say we are long-term investors, but "tilt" investors have to be strongly committed, or they are likely to have second thoughts during periods of under-performance.
http://corporate.morningstar.com/ib/doc ... xplain.pdf
Now before you walk away from the long-term strategy, perhaps you can look to some recent developments for encouragement. The world appears to be uncoupling. With China's and Europe's economies sputtering along, domestic growth appears to be the next opportunity in this aging bull market. The "hot money" might soon recognize this and you could be in like Flynn. Small- caps and mid- caps are positioned well because even though the US economy sputters along, its "relative" performance is still better than the rest of the world. And, small and mid-cap tend to have more domestic business, while the large cap multi-nationals are now facing earnings headwinds from a strong $USD. Prices for many large caps in the S & P 500 have been under pressure the past few weeks.
At this point, I'd just keep up your D-C-A into your small-and mid cap "tilt" strategy, and "stay the course " In the meantime here's some wisdom and encouragement from a great investor, Ben Graham:
"It requires strength of character in order to think and to act in opposite fashion from the crowd--- and also patience to wait for opportunities that may be spaced years apart."
Best wishes for continuing success, and good luck along your investing journey !
Last edited by Rx 4 investing on Sun Aug 02, 2015 3:02 pm, edited 1 time in total.
“Everyone is a disciplined, long-term investor until the market goes down.” – Steve Forbes
Re: Houston, we have a problem... [Small cap historical performance]
Were there any small-cap index funds before VG converted NAESX to one in 1994?
Re: Houston, we have a problem... [Small cap historical performance]
One of the interesting things about the smallcap "premium" is how much the questioning of it has started up after a long-period of small cap *outpeformance.* Usually you see a lot of additional interest in asset classes/sub-classes after such outperformance, with a lot of academic/sophisticated analysis explaining why it is some type of economic law that such outpeformance must exist for time and eternity. Instead, among the people who think about such stuff a lot, the conclusion has increasingly become that the small-cap premium is bunk, even before there is any consideration of theoretical vs. real world experience.
For example: http://aswathdamodaran.blogspot.com/201 ... n-and.html or https://www.researchaffiliates.com/Our% ... _Size.aspx
For example: http://aswathdamodaran.blogspot.com/201 ... n-and.html or https://www.researchaffiliates.com/Our% ... _Size.aspx
-
- Posts: 5682
- Joined: Sat Aug 11, 2012 8:44 am
Re: Houston, we have a problem... [Small cap historical performance]
Goingup,goingup wrote:I think that's the case that Jack Bogle made with the telltale chart--that the evidence doesn't support the benefit of tilting. http://www.vanguard.com/bogle_site/sp20020626.html
I have overweighted small cap in our portfolio and the return has been better than TSM over the last 10 years. But that has everything to do with the timing of contributions. It's evidence of nothing and just my personal data point.
I'd like to check that you're not comparing a money-weighted rate of return (XIRR) with a time-weighted rate of return. (Which would be an invalid comparison*).
* See wiki:Calculating personal returns.
When you say that your portfolio had a better return than TSM, are you comparing the XIRR return of your portfolio and the XIRR return of a hypothetical portfolio with identical timing of contributions but without but a tilt? (Which would be a valid comparison).
Or, maybe, have you calculated the time-weighted return of your portfolio and compared it to the return of TSM? (Which is the recommended approach to compare performance).
Last edited by longinvest on Sun Aug 02, 2015 3:26 pm, edited 1 time in total.
Variable Percentage Withdrawal (bogleheads.org/wiki/VPW) | One-Fund Portfolio (bogleheads.org/forum/viewtopic.php?t=287967)
Re: Houston, we have a problem... [Small cap historical performance]
Despite having data that goes back to 1978, the Russell 2000 index didn't even exist until 1984, the older data was recreated. As others have pointed out, Vanguard's Small-Cap fund wasn't an index fund until the mid 1990's. There's lots of arguments about small-cap data and small-cap funds.
DFSCX is probably a better fund if you want to look at micro-cap performance, but that isn't quite the same as how some small-caps are defined, so actual small-cap indices and funds will vary (and they do - allot).
Since inception, DFSCX (DFA's U.S. micro-cap fund) has had about .50% annualized return more than VFINX (Vanguard's S&P500 fund).
Maybe the .50% would have paid for the fees you'd have to pay an adviser to buy the fund.
But even that performance is period dependent. If you had bought the fund in the early 1980's you might have held it 20+ years and never saw returns higher than the broad market, and even with the recent period of small-cap out-performance, there's no saying things won't change going forward. These things are always sensitive to start/stop dates.
DFSCX is probably a better fund if you want to look at micro-cap performance, but that isn't quite the same as how some small-caps are defined, so actual small-cap indices and funds will vary (and they do - allot).
Since inception, DFSCX (DFA's U.S. micro-cap fund) has had about .50% annualized return more than VFINX (Vanguard's S&P500 fund).
Maybe the .50% would have paid for the fees you'd have to pay an adviser to buy the fund.
But even that performance is period dependent. If you had bought the fund in the early 1980's you might have held it 20+ years and never saw returns higher than the broad market, and even with the recent period of small-cap out-performance, there's no saying things won't change going forward. These things are always sensitive to start/stop dates.
"To achieve satisfactory investment results is easier than most people realize; to achieve superior results is harder than it looks." - Benjamin Graham
Re: Houston, we have a problem... [Small cap historical performance]
Hi longinvest:longinvest wrote:Goingup,goingup wrote:I think that's the case that Jack Bogle made with the telltale chart--that the evidence doesn't support the benefit of tilting. http://www.vanguard.com/bogle_site/sp20020626.html
I have overweighted small cap in our portfolio and the return has been better than TSM over the last 10 years. But that has everything to do with the timing of contributions. It's evidence of nothing and just my personal data point.
I'd like to check that you're not comparing a money-weighted rate of return with a time-weighted rate of return. (Which would be an invalid comparison).
When you say that your portfolio had a better return than TSM, are you comparing the XIRR return of your portfolio and the XIRR return of a hypothetical portfolio with identical timing of contributions but without but a tilt? (Which would be a valid comparison).
Or, maybe, have you calculated the time-weighted return of your portfolio and compared it to the return of TSM? (Which is the recommended approach to compare performance).
I'm getting the IRR calculations from my Personal Returns tab on my Vanguard account. I agree that this is not relevant to the tilting debate. My 10-yr returns for VSMAX have averaged 11.3% vs my 10-yr returns for VTSAX at 10%. As I said these are my personal data points, and can't be extrapolated to an argument for tilting. (I really shouldn't have waded into this...)
-
- Posts: 5682
- Joined: Sat Aug 11, 2012 8:44 am
Re: Houston, we have a problem... [Small cap historical performance]
OK. Thanks for the clarification.goingup wrote:Hi longinvest:
I'm getting the IRR calculations from my Personal Returns tab on my Vanguard account. I agree that this is not relevant to the tilting debate. My 10-yr returns for VSMAX have averaged 11.3% vs my 10-yr returns for VTSAX at 10%. As I said these are my personal data points, and can't be extrapolated to an argument for tilting. (I really shouldn't have waded into this...)
If you're interested to do a fair comparison, you could calculate the time-weighted returns of your portfolio using the wiki's spreadsheet and compare them to VTSAX's. (No need to report back the results here; past returns should not be a guide for investing!).
Variable Percentage Withdrawal (bogleheads.org/wiki/VPW) | One-Fund Portfolio (bogleheads.org/forum/viewtopic.php?t=287967)
- nisiprius
- Advisory Board
- Posts: 52216
- Joined: Thu Jul 26, 2007 9:33 am
- Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry
Re: Houston, we have a problem... [Small cap historical performance]
privatefarmer, given that you (think you are) following Fama and French, a reasonably interesting chart to look at is the one below.
Fama and French work for Dimensional Fund Advisors. It's my personal opinion (I'm a factor skeptic) that if the effects claimed for factor-informed investing were robust, you really ought to be able to see them as long as you use reasonably appropriate funds for the job. If caffeine keeps you alert, it shouldn't take estate Kona to show the effect--Dunkin' Donuts ought to work even if Kona works better. If Jupiter has moons, you shouldn't need to use Galileo's personal telescope to see them.
Nevertheless, it seems plausible that Dimensional is the place with the best understanding of Fama-French factors and the best tools for factor-based investing.
And the most relevant one here is DFA US Micro Cap, DFSCX. It's "passive" as DFA understands the term, and it is surely the oldest passive small-cap fund. It was launched in 1981, the same year that Rolf Banz published "The relationship between return and market value of common stock," the paper that introduced the "size effect" to the investing world.
I think it's a fair test of the "size effect" and of course people will read different things into it.
The small cap fund (DFSCX, blue) outperformed the S&P 500 (orange) for a couple of years and then--as so often seems to be the case--the expected outperformance vanished, and the connoisseur's small cap fund fell behind the S&P 500. It took until 2002 to catch up.
The final score, over 33-2/3 years: small-cap fund, about 12.05% annualized; S&P 500, 11.56%.
To get that extra gain of 0.51% annualized, you had to watch your fund underperforming the S&P 500 for nearly 20 years. At the end you had $460,000 instead of $400,000. The scary thing is how much the outcome depend on the accident of the particular choice of endpoints you choose.
Source: Morningstar
Fama and French work for Dimensional Fund Advisors. It's my personal opinion (I'm a factor skeptic) that if the effects claimed for factor-informed investing were robust, you really ought to be able to see them as long as you use reasonably appropriate funds for the job. If caffeine keeps you alert, it shouldn't take estate Kona to show the effect--Dunkin' Donuts ought to work even if Kona works better. If Jupiter has moons, you shouldn't need to use Galileo's personal telescope to see them.
Nevertheless, it seems plausible that Dimensional is the place with the best understanding of Fama-French factors and the best tools for factor-based investing.
And the most relevant one here is DFA US Micro Cap, DFSCX. It's "passive" as DFA understands the term, and it is surely the oldest passive small-cap fund. It was launched in 1981, the same year that Rolf Banz published "The relationship between return and market value of common stock," the paper that introduced the "size effect" to the investing world.
I think it's a fair test of the "size effect" and of course people will read different things into it.
The small cap fund (DFSCX, blue) outperformed the S&P 500 (orange) for a couple of years and then--as so often seems to be the case--the expected outperformance vanished, and the connoisseur's small cap fund fell behind the S&P 500. It took until 2002 to catch up.
The final score, over 33-2/3 years: small-cap fund, about 12.05% annualized; S&P 500, 11.56%.
To get that extra gain of 0.51% annualized, you had to watch your fund underperforming the S&P 500 for nearly 20 years. At the end you had $460,000 instead of $400,000. The scary thing is how much the outcome depend on the accident of the particular choice of endpoints you choose.
Source: Morningstar
Annual income twenty pounds, annual expenditure nineteen nineteen and six, result happiness; Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery.
Re: Houston, we have a problem... [Small cap historical performance]
Between 1/1/1984 - 1/1/1999
VFINX outperformed DFSCX +6% a year (annualized over the 15 year period)
Between 1/1/1999 - 1/1/2014
DFSCX outperformed VFINX +6.7% a year (annualized over the 15 year period)
Over the period 1/1/1984 to present (8/1/2015) they've had just about the same returns.
I really don't think this is indicative of one type of stock having any persistent premium. To me, it's just noise and demonstrates that there can be wide variations in returns among different groups of stocks.
VFINX outperformed DFSCX +6% a year (annualized over the 15 year period)
Between 1/1/1999 - 1/1/2014
DFSCX outperformed VFINX +6.7% a year (annualized over the 15 year period)
Over the period 1/1/1984 to present (8/1/2015) they've had just about the same returns.
I really don't think this is indicative of one type of stock having any persistent premium. To me, it's just noise and demonstrates that there can be wide variations in returns among different groups of stocks.
"To achieve satisfactory investment results is easier than most people realize; to achieve superior results is harder than it looks." - Benjamin Graham
Re: Houston, we have a problem... [Small cap historical performance]
Impressive that the micro cap had the same return despite 35 bp higher fees. That means if we can just get the expenses down (VBR is pretty low in expenses) then we can harvest the premium.JoMoney wrote:Between 1/1/1984 - 1/1/1999
VFINX outperformed DFSCX +6% a year (annualized over the 15 year period)
Between 1/1/1999 - 1/1/2014
DFSCX outperformed VFINX +6.7% a year (annualized over the 15 year period)
Over the period 1/1/1984 to present (8/1/2015) they've had just about the same returns.
I really don't think this is indicative of one type of stock having any persistent premium. To me, it's just noise and demonstrates that there can be wide variations in returns among different groups of stocks.
Monthly or yearly movements of stocks are often erratic and not indicative of changes in intrinsic value. Over time, however, stock prices and intrinsic value almost invariably converge. ~ WB
-
- Posts: 16022
- Joined: Thu Feb 22, 2007 7:28 am
- Location: St Louis MO
Re: Houston, we have a problem... [Small cap historical performance]
I'd just add this, when looking at returns of asset classes have to look at the valuations at start and end
The DFA micro fund was created right at the tail end of the "small cap bubble" when small caps outperformed large by massive margin from 75-83,a bout 14% a year for that 9 years.
Since 75, choosing that as starting point before the bubble, DFA small cap index outperformed their large index by about 3% a year
Also as I noted we should expect smaller size premium as liquidity costs are now much lower
Larry
The DFA micro fund was created right at the tail end of the "small cap bubble" when small caps outperformed large by massive margin from 75-83,a bout 14% a year for that 9 years.
Since 75, choosing that as starting point before the bubble, DFA small cap index outperformed their large index by about 3% a year
Also as I noted we should expect smaller size premium as liquidity costs are now much lower
Larry
- privatefarmer
- Posts: 779
- Joined: Mon Sep 08, 2014 2:45 pm
Re: Houston, we have a problem... [Small cap historical performance]
Thanks everyone for the input. I did not realize NAESX was actively managed until '94, that definitely explains why it was not matching up w/ the historical small cap data that portfoliovisualizer gave.stlutz wrote:Were there any small-cap index funds before VG converted NAESX to one in 1994?
It is interesting to see that the DFA micro cap fund "only" outperformed by 0.5%/yr over those several decades. You'd have to think that you're taking on a lot more risk being in micro caps vs S/P 500 and maybe not being well compensated for it?
Anyhow, I do tilt to small/value heavily but I've been using vanguard ETFs so the costs for doing so have been negligible. I was offered DFA access at 30bp/yr by FPL capital management, but I am not sold that I would recoup that through the more "pure" tilting that DFA does. I think I'll just stay the course, focus on keeping expenses near or under 20bp/yr, and globally diversified.
Thanks everyone for the help
-
- Posts: 5682
- Joined: Sat Aug 11, 2012 8:44 am
Re: Houston, we have a problem... [Small cap historical performance]
Larry,larryswedroe wrote:I'd just add this, when looking at returns of asset classes have to look at the valuations at start and end
The DFA micro fund was created right at the tail end of the "small cap bubble" when small caps outperformed large by massive margin from 75-83,a bout 14% a year for that 9 years.
Since 75, choosing that as starting point before the bubble, DFA small cap index outperformed their large index by about 3% a year
Also as I noted we should expect smaller size premium as liquidity costs are now much lower
Larry
According to http://cawidgets.morningstar.ca/Article ... ture=en-CA, "[f]or years, Canadian small-cap equities have been a disappointing asset class. Returns have lagged the broader Canadian stock market and volatility, as always, has been higher."
So, doesn't a small cap premium come with a higher risk? Why should small-caps have higher returns than the entire stock market, unless accompanied with additional risk?
Or, is the small cap premium limited to some countries with specific characteristics, and Canada does not qualify?
Variable Percentage Withdrawal (bogleheads.org/wiki/VPW) | One-Fund Portfolio (bogleheads.org/forum/viewtopic.php?t=287967)
Re: Houston, we have a problem... [Small cap historical performance]
Does the idea of a bubble impact a "risk premium" ? If we acknowledge that the markets are subject to behavioral swings where some stocks may become priced above what a rational price should be... how do we know we're not in a "small-cap bubble" now?
"To achieve satisfactory investment results is easier than most people realize; to achieve superior results is harder than it looks." - Benjamin Graham
Re: Houston, we have a problem... [Small cap historical performance]
Many people are saying small cap valuations are a bit high now, for what it's worth.JoMoney wrote:Does the idea of a bubble impact a "risk premium" ? If we acknowledge that the markets are subject to behavioral swings where some stocks may become priced above what a rational price should be... how do we know we're not in a "small-cap bubble" now?
Of course an average doesn't say if the current price is fair or not. But it's something. Are current and expected future conditions friendlier for small caps than before (i.e. is it justified)?
If nothing else, as mentioned, higher liquidity and the ability to easily diversify with thousands of stocks should mean risks are lower, so if the market might well accept higher prices to be reasonable.
- Maynard F. Speer
- Posts: 2139
- Joined: Wed Mar 18, 2015 10:31 am
Re: Houston, we have a problem... [Small cap historical performance]
If you look up Norbert Keimling's Twitter, he's recently calculated CAPE ratios for US small and large caps - and small looks very expensive now .. My own analysis with PEG ratios, however, makes them all look quite similarly valued
One reason you might expect a small-cap premium is because once a stock's grown, and left the index, you've essentially banked your profits - so there's a stronger momentum effect .. In a large-cap index, your largest holdings - over the long term - have been more likely to drop back down the index
Another consideration is small-caps are usually more domestically focused .. So whether there's an advantage to holding them may come down to the health of local economies - another reason small-caps may be popular among US investors at the moment .. But too popular and valuations and unrealistic expectations are likely to be a greater drag than any potential premium
I don't know whether it's un-BH-like - as Malkiel uses active management where markets are less efficient - but active micro-caps represent about a third of my equities exposure at the moment
Micro-caps are often too small for it to be worth institutional investors' or large fund managers' energies researching, and almost fly under the radars of most indexes .. Essentially I see them as an analog for private equity - albeit with better liquidity and without the 1.5/15 fee structures .. A space that may be somewhat immune from the headwinds facing the broader indexes
One reason you might expect a small-cap premium is because once a stock's grown, and left the index, you've essentially banked your profits - so there's a stronger momentum effect .. In a large-cap index, your largest holdings - over the long term - have been more likely to drop back down the index
Another consideration is small-caps are usually more domestically focused .. So whether there's an advantage to holding them may come down to the health of local economies - another reason small-caps may be popular among US investors at the moment .. But too popular and valuations and unrealistic expectations are likely to be a greater drag than any potential premium
I don't know whether it's un-BH-like - as Malkiel uses active management where markets are less efficient - but active micro-caps represent about a third of my equities exposure at the moment
Micro-caps are often too small for it to be worth institutional investors' or large fund managers' energies researching, and almost fly under the radars of most indexes .. Essentially I see them as an analog for private equity - albeit with better liquidity and without the 1.5/15 fee structures .. A space that may be somewhat immune from the headwinds facing the broader indexes
"Economics is a method rather than a doctrine, an apparatus of the mind, a technique of thinking, which helps its possessor to draw correct conclusions." - John Maynard Keynes
Re: Houston, we have a problem... [Small cap historical performance]
I saw you post previously regarding a PEG ratio using forward-prospective P/E's, and projected earnings growthMaynard F. Speer wrote: .. My own analysis with PEG ratios, however, makes them all look quite similarly valued ..
viewtopic.php?f=10&t=169920&p=2561521#p2561521
I think this may misrepresent it, as you're double counting earnings growth doing it that way.
The forward P/E is already assuming the future earnings growth at the current price.
But there's problems if you use the trailing P/E with the future expected earnings growth as well. With that you're not really showing anything different than what the forward P/E is already showing (unless you're using your own unique growth estimates). Using trailing earnings for small-caps, and especially small-value seems problematic because so many of these stocks have extremely low/negative earnings and depending on how you decide to account for that you can get all sorts of crazy numbers.
Vanguard might put the trailing Russell 2000 P/E at 34.5
but the WSJ puts it at 79.76
I believe the difference is just because of how they deal with zero or negative earnings which just can't be used to show a P/E.
"To achieve satisfactory investment results is easier than most people realize; to achieve superior results is harder than it looks." - Benjamin Graham
- Maynard F. Speer
- Posts: 2139
- Joined: Wed Mar 18, 2015 10:31 am
Re: Houston, we have a problem... [Small cap historical performance]
I'm not using forward P/E, so I'm calculating trailing P/E divided by forecast earnings growth - so you're getting a sort of vector, rather than a static point, that tells you something you could extrapolate only from both trailing and forward P/EJoMoney wrote:I saw you post previously regarding a PEG ratio using forward-prospective P/E's, and projected earnings growthMaynard F. Speer wrote: .. My own analysis with PEG ratios, however, makes them all look quite similarly valued ..
viewtopic.php?f=10&t=169920&p=2561521#p2561521
I think this may misrepresent it, as you're double counting earnings growth doing it that way.
The forward P/E is already assuming the future earnings growth at the current price.
But there's problems if you use the trailing P/E with the future expected earnings growth as well. With that you're not really showing anything different than what the forward P/E is already showing (unless you're using your own unique growth estimates). Using trailing earnings for small-caps, and especially small-value seems problematic because so many of these stocks have extremely low/negative earnings and depending on how you decide to account for that you can get all sorts of crazy numbers.
Vanguard might put the trailing Russell 2000 P/E at 34.5
but the WSJ puts it at 79.76
I believe the difference is just because of how they deal with zero or negative earnings which just can't be used to show a P/E.
It was used famously by one of our top small companies investors, Jim Slater - and according to Peter Lynch (within reason) a correctly valued stock's P/E ratio should match its earnings growth .. a P/E of 30 could be justified by earnings growth of 30%, whereas a P/E of 10 wouldn't be very appealing with earnings growth of only 5% (very much a rule of thumb)
Obviously there are weaknesses .. Analyst forecasts can change sharply (to an extent, you can only know as much as the market knows about itself); P/E can be deceptive (as you say); this isn't generally used over indexes, so you could have stocks with poor PEG ratios, but with a mix of value traps and expensive high growth .. Dividends may need to be added to earnings forecasts .. Sector weightings can distort valuations .. So it's a tool among many that tells you something other tools don't, but to an extent I'm surprised how evenly valued US indices appear to be using such a crude measure
"Economics is a method rather than a doctrine, an apparatus of the mind, a technique of thinking, which helps its possessor to draw correct conclusions." - John Maynard Keynes
-
- Posts: 9883
- Joined: Mon Sep 07, 2009 2:57 pm
- Location: Milky Way
Re: Houston, we have a problem... [Small cap historical performance]
So, let's see; we all agree that small-cap stocks are (in aggregate) riskier than large cap stocks - right? Why then would it not follow that they should also have higher expected return?stlutz wrote:One of the interesting things about the smallcap "premium" is how much the questioning of it has started up after a long-period of small cap *outpeformance.* Usually you see a lot of additional interest in asset classes/sub-classes after such outperformance, with a lot of academic/sophisticated analysis explaining why it is some type of economic law that such outpeformance must exist for time and eternity. Instead, among the people who think about such stuff a lot, the conclusion has increasingly become that the small-cap premium is bunk, even before there is any consideration of theoretical vs. real world experience.
Best regards, -Op |
|
"In the middle of difficulty lies opportunity." Einstein
- just frank
- Posts: 1949
- Joined: Sun Nov 02, 2014 3:13 pm
- Location: Philly Metro
Re: Houston, we have a problem... [Small cap historical performance]
The fly in my punchbowl for small cap is private equity. If there is a higher return to be had investing in smaller more entrepreneurial companies, relative to large-caps, why wouldn't a bunch of rich people jump on it for themselves (after the Fama/French paper)??
If you look at the huge growth in venture capital (now a lot of it outside the major VC firms, i.e. angel investors) over the last 20 years, you could claim that the VCs are taking a lot of the profits of the most important/profitable 'small caps', and those things are IPO-ing and showing up on the index only after they've already got $1B+ in 'cap' (i.e. 'a unicorn').
Am I missing something here?
If you look at the huge growth in venture capital (now a lot of it outside the major VC firms, i.e. angel investors) over the last 20 years, you could claim that the VCs are taking a lot of the profits of the most important/profitable 'small caps', and those things are IPO-ing and showing up on the index only after they've already got $1B+ in 'cap' (i.e. 'a unicorn').
Am I missing something here?
Re: Houston, we have a problem... [Small cap historical performance]
Some people explicitly claim all sorts of risks are "uncompensated risk", so I'm not sure that the theory even claims higher risk automatically means higher expected returns unless you assume risk = some specific model they use for measuring it. While it would be rational to only take compensated risk, assuming the market is completely rational may not be something everyone agrees on.Call_Me_Op wrote:...
So, let's see; we all agree that small-cap stocks are (in aggregate) riskier than large cap stocks - right? Why then would it not follow that they should also have higher expected return?
I think you might need a common definition of the risk before you can say can really even ask if everyone agrees that it's more risky.
If you asked me to answer the question: "Do you believe you would lose more/gain less money being invested in small cap stocks right now than large cap stocks over the next decade?" ... Then I might say yes, I think the risk of losing more money is higher in small-caps.
"To achieve satisfactory investment results is easier than most people realize; to achieve superior results is harder than it looks." - Benjamin Graham
Re: Houston, we have a problem... [Small cap historical performance]
1) The Fama/French seminal work is not focused on the out-performance of small-value stocks as many people seem to think. They provided a method to decompose a portfolio into various factors, by comparing the returns of a portfolio with the returns of some base portfolios. The issue of performance of one or more of these factors is an entirely separate matter.privatefarmer wrote:I am a small/value "tilter". I have believed the evidence presented by Fama/French.
However, can someone answer me this: when you compare Vanguard's small cap blend index (NAESX) to vanguards s/p 500 index (VFINX), going back nearly 30 years, on portfoliovisualizer.com the s/p 500 DESTROYS the small cap index over that 28-year period.
When you look at what the index's "should" have done, according to porftoliovisualizer, there basically should've been a tie over that period.
They both are nearly identical in expenses. NAESX is not pure small, it is a mix of mid-cap and small-cap, but when you account for that in portfoliovisualizer the mid/small mix should've actually beaten the large blend by a solid 1%/yr over the '86-'14 period.
I am starting to suspect that perhaps the small/value "factors" work in theory only. When you actually invest in index funds/ETFs that track these they don't do what you would expect them to do. Any thoughts?
2) Even if there is small-cap premium, it doesn't mean that investors will capture that premium in every time period. For some periods you may and in other periods you may not. If there is a premium, it just means that in the majority of time periods there will be a premium.
Re: Houston, we have a problem... [Small cap historical performance]
Any time you look back, even over long time periods, your results will be heavily skewed by recent performance. Over the last few years, the large cap indexes have done much better than small caps. This will make the small cap strategy look worse when looking back.
The Espresso portfolio: |
|
20% US TSM, 20% Small Value, 10% US REIT, 10% Dev Int'l, 10% EM, 10% Commodities, 20% Inter-term US Treas |
|
"A journey of a thousand miles begins with a single step."
Re: Houston, we have a problem... [Small cap historical performance]
FWIW, the total return is comparable for the longest period available for the two funds on Morningstar:
The comparison is apparently very sensitive to begin and end dates, as many are.
Avi
The comparison is apparently very sensitive to begin and end dates, as many are.
Avi
- Maynard F. Speer
- Posts: 2139
- Joined: Wed Mar 18, 2015 10:31 am
Re: Houston, we have a problem... [Small cap historical performance]
This is Norbert Keimling's chart on Small vs Large-cap CAPE ratiosjust frank wrote:The fly in my punchbowl for small cap is private equity. If there is a higher return to be had investing in smaller more entrepreneurial companies, relative to large-caps, why wouldn't a bunch of rich people jump on it for themselves (after the Fama/French paper)??
If you look at the huge growth in venture capital (now a lot of it outside the major VC firms, i.e. angel investors) over the last 20 years, you could claim that the VCs are taking a lot of the profits of the most important/profitable 'small caps', and those things are IPO-ing and showing up on the index only after they've already got $1B+ in 'cap' (i.e. 'a unicorn').
Am I missing something here?
And this might suggest people have piled into small-caps en masse, all chasing the same premiums .. Or there may be a 'ship of Theseus' problem (perhaps more IPOs skewing the metric vs averaged earnings?)
I don't know the US market as well, but if PE and VCTs have effectively pushed the small-cap bracket up, then I'd imagine much of the premium will be pushed down into the unlisted space .. It would sort of echo El-Erian's advice: Invest less in stocks, more in tech and private equity
http://www.cnbc.com/2015/07/23/invest-l ... erian.html
"Economics is a method rather than a doctrine, an apparatus of the mind, a technique of thinking, which helps its possessor to draw correct conclusions." - John Maynard Keynes
Re: Houston, we have a problem... [Small cap historical performance]
Small caps have higher earnings growth potential so it makes sense that their PE would be higher. PE changes alone can't explain the return premium going back to 1926.Maynard F. Speer wrote:This is Norbert Keimling's chart on Small vs Large-cap CAPE ratiosjust frank wrote:The fly in my punchbowl for small cap is private equity. If there is a higher return to be had investing in smaller more entrepreneurial companies, relative to large-caps, why wouldn't a bunch of rich people jump on it for themselves (after the Fama/French paper)??
If you look at the huge growth in venture capital (now a lot of it outside the major VC firms, i.e. angel investors) over the last 20 years, you could claim that the VCs are taking a lot of the profits of the most important/profitable 'small caps', and those things are IPO-ing and showing up on the index only after they've already got $1B+ in 'cap' (i.e. 'a unicorn').
Am I missing something here?
And this might suggest people have piled into small-caps en masse, all chasing the same premiums .. Or there may be a 'ship of Theseus' problem (perhaps more IPOs skewing the metric vs averaged earnings?)
I don't know the US market as well, but if PE and VCTs have effectively pushed the small-cap bracket up, then I'd imagine much of the premium will be pushed down into the unlisted space .. It would sort of echo El-Erian's advice: Invest less in stocks, more in tech and private equity
http://www.cnbc.com/2015/07/23/invest-l ... erian.html
Monthly or yearly movements of stocks are often erratic and not indicative of changes in intrinsic value. Over time, however, stock prices and intrinsic value almost invariably converge. ~ WB
- Maynard F. Speer
- Posts: 2139
- Joined: Wed Mar 18, 2015 10:31 am
Re: Houston, we have a problem... [Small cap historical performance]
At the moment those P/Es are about 2x forecast earnings growth, which suggests those parts of the market are simply too crowdedrca1824 wrote:Small caps have higher earnings growth potential so it makes sense that their PE would be higher. PE changes alone can't explain the return premium going back to 1926.
One thing to remember with 1926 is markets were presumably much less efficient then - but the simple explanation could be that in a large-cap index you're giving highest weightings to companies that (historically) have tended to fall down the index within 10-20 years, whereas with small-caps you're selling your winners long before they reach that stage
"Economics is a method rather than a doctrine, an apparatus of the mind, a technique of thinking, which helps its possessor to draw correct conclusions." - John Maynard Keynes
Re: Houston, we have a problem... [Small cap historical performance]
Do you have any data on this? Why do you believe it can't?rca1824 wrote:... PE changes alone can't explain the return premium going back to 1926.
Changes in the price paid compounded with dividends over 90 years seems like it could explain a bit to me. It's certainly been pretty explanative over the past 15 or so years. I can see plenty of reasons why things have changed in the markets and in businesses themselves over time.
Before changes in the laws starting with ERISA in the 1970's, there used to be a "prudent man rule" in how a trustee had to manage assets. Some investments, like small-cap stocks, would have been by default considered too speculative for the majority of institutional investors and put them at risk of liability if they invested in them. That certainly isn't the case anymore, and I'm sure there's dozens of other differences between small-caps today and small-caps going back to 1926.
"To achieve satisfactory investment results is easier than most people realize; to achieve superior results is harder than it looks." - Benjamin Graham
Re: Houston, we have a problem... [Small cap historical performance]
Only the intuition that PE changes can't persist in long term returns over a period of 90 years. Even if you bought small caps at a 50% discount below their intrinsic value in 1926, you wouldn't get the generous return they've yielded since then.JoMoney wrote:Do you have any data on this? Why do you believe it can't?rca1824 wrote:... PE changes alone can't explain the return premium going back to 1926.
Changes in the price paid compounded with dividends over 90 years seems like it could explain a bit to me. It's certainly been pretty explanative over the past 15 or so years. I can see plenty of reasons why things have changed in the markets and in businesses themselves over time.
Before changes in the laws starting with ERISA in the 1970's, there used to be a "prudent man rule" in how a trustee had to manage assets. Some investments, like small-cap stocks, would have been by default considered too speculative for the majority of institutional investors and put them at risk of liability if they invested in them. That certainly isn't the case anymore, and I'm sure there's dozens of other differences between small-caps today and small-caps going back to 1926.
Using Rodc's charts http://www.bogleheads.org/forum/viewtopic.php?t=9445
The SCV premium since 1927 has been 5% CAGR (15% vs. 10%). The small cap premium alone was roughly half of that, or 2.5%. I The data only spans 80 years. Over that period, $1 in SB would have grown to be be 6x greater than $1 invested in TSM. This means for changes in PE alone to explain this variation, SB would have been offered at a 84% discount in 1927. An SCV would have had to be an offered at a 97% discount.
So either markets were so inefficient in 1927 that investors could get 84-97% discounts just by shopping around small cap value, or more likely, they were probably getting no more than 50% discounts and the remainder of the excess performance is because of earnings growth.
If I had data on small cap value earnings in 1927 I could answer more confidently but with just total return data, arithmetic, common sense, and intuition I think we can rule out the hypothesis that the SCV premium is entirely attributed to changes in valuations. There must have been real earnings to back them up.
Monthly or yearly movements of stocks are often erratic and not indicative of changes in intrinsic value. Over time, however, stock prices and intrinsic value almost invariably converge. ~ WB
Re: Houston, we have a problem... [Small cap historical performance]
I've shown it before, but the 6% annualized difference between a small-cap index fund and a large-cap index fund over the past 15 year period can be entirely explained by relative changes in the P/E multiple over that time. I don't think it's unfeasible to believe the same situation could exist when looking at a 90 year period.
"To achieve satisfactory investment results is easier than most people realize; to achieve superior results is harder than it looks." - Benjamin Graham
Re: Houston, we have a problem... [Small cap historical performance]
Seems to be affected by the starting point in 1986. If you change it to 1988, there isn't much difference. These kind of comparisons are always dependent on the starting and ending point. You have to also remember that large cap had a really fine time of it in the 1990s and beat the kkkkrap out of everything else on the planet. But also, the Small Cap factor went the way of the dodo after it was discovered in the wild. Now you have to search in the deepest, darkest part of small cap space to find the "value" stocks which may be going extinct too.
We don't know where we are, or where we're going -- but we're making good time.
Re: Houston, we have a problem... [Small cap historical performance]
Some quotes from Andrew Ang "Asset Management":
"The size effect was discovered by Benz (1981), with similar results in Reinganum (1981), and refers to the fact that small stocks tended to do better than large stocks, after adjusting for their betas. The past tense is appropriate here, because since the mid-1980s there has not been any significant size effect." (p228)
"That is, small stocks do have higher returns than large stocks but not after taking out their exposure to the market factor. The creation of small-stock mutual funds allowed the ordinary investor to participate and bear size-related risk. Thus the risk-bearing capacity of the economy changed after industry created new products to capitalize on the size premium. Those industry developments caused size to disappear. Those early investors in small caps experienced a bonanza. Prices of small stocks rose back to long-run equilibrium and early investors enjoyed a tidy risk-adjusted capital gain." (p457)
"The size effect was discovered by Benz (1981), with similar results in Reinganum (1981), and refers to the fact that small stocks tended to do better than large stocks, after adjusting for their betas. The past tense is appropriate here, because since the mid-1980s there has not been any significant size effect." (p228)
"That is, small stocks do have higher returns than large stocks but not after taking out their exposure to the market factor. The creation of small-stock mutual funds allowed the ordinary investor to participate and bear size-related risk. Thus the risk-bearing capacity of the economy changed after industry created new products to capitalize on the size premium. Those industry developments caused size to disappear. Those early investors in small caps experienced a bonanza. Prices of small stocks rose back to long-run equilibrium and early investors enjoyed a tidy risk-adjusted capital gain." (p457)
Re: Houston, we have a problem... [Small cap historical performance]
I'm not surprised, there are several charts floating around like this showing relative small/large PE ratios over time.Browser wrote:Seems to be affected by the starting point in 1986. If you change it to 1988, there isn't much difference. These kind of comparisons are always dependent on the starting and ending point. You have to also remember that large cap had a really fine time of it in the 1990s and beat the kkkkrap out of everything else on the planet. But also, the Small Cap factor went the way of the dodo after it was discovered in the wild. Now you have to search in the deepest, darkest part of small cap space to find the "value" stocks which may be going extinct too.
If you pick two points where the relative valuation was about the same, you seem to find similar returns
Morningstar 1995-2005
...now if I just knew what the spread in future valuations of large/small will be
"To achieve satisfactory investment results is easier than most people realize; to achieve superior results is harder than it looks." - Benjamin Graham
- Maynard F. Speer
- Posts: 2139
- Joined: Wed Mar 18, 2015 10:31 am
Re: Houston, we have a problem... [Small cap historical performance]
That's exactly why - rather than just equal-weighting Micro, Small, Mid-cap, etc (which I believe works well) - I look at ETFs and work out where's least crowded, and if there's an opportunity to pick up a sector near the bottom end of its valuations
What doesn't tend to happen is a cheap part of the market getting perpetually cheaper, or vice versa, expensive (it can endure for a few years - e.g. Russia or Biotech - but things get very distorted) ... So even with random markets, and some boundary effect, autoregression (mean reversion) is what you'd expect
Buy in that bottom half of the graph and you've always found yourself in the top after a while .. And when valuations across most asset classes look like they may be a drag on future returns, I think mean reversion's a very interesting thing to think about
What doesn't tend to happen is a cheap part of the market getting perpetually cheaper, or vice versa, expensive (it can endure for a few years - e.g. Russia or Biotech - but things get very distorted) ... So even with random markets, and some boundary effect, autoregression (mean reversion) is what you'd expect
Buy in that bottom half of the graph and you've always found yourself in the top after a while .. And when valuations across most asset classes look like they may be a drag on future returns, I think mean reversion's a very interesting thing to think about
"Economics is a method rather than a doctrine, an apparatus of the mind, a technique of thinking, which helps its possessor to draw correct conclusions." - John Maynard Keynes
Re: Houston, we have a problem... [Small cap historical performance]
Thanks for those links. I've seen other papers that demoted Small-Cap Factor to Small-Cap Anomaly. Damodaran's article The Small Cap Premium: Where is the beef? provides additional confirmation of my bias against small-cap. Damodaran is a highly respected finance professor and researcher at the Stern School of Business at NYU, and a leading authority on valuations and expected returns, so I give weight to what he writes.stlutz wrote:One of the interesting things about the smallcap "premium" is how much the questioning of it has started up after a long-period of small cap *outpeformance.* Usually you see a lot of additional interest in asset classes/sub-classes after such outperformance, with a lot of academic/sophisticated analysis explaining why it is some type of economic law that such outpeformance must exist for time and eternity. Instead, among the people who think about such stuff a lot, the conclusion has increasingly become that the small-cap premium is bunk, even before there is any consideration of theoretical vs. real world experience.
For example: http://aswathdamodaran.blogspot.com/201 ... n-and.html or https://www.researchaffiliates.com/Our% ... _Size.aspx
For anyone interested in the subject of Small-Cap Premium, I strongly recommend reading Damodarin's article linked above.
So it sounds to me like it's time to bury the small-cap anomaly. But apparently the value factor is still alive and kicking. And, for closet stock pickers, there are plenty of other factors in the factor zoo to tilt towards: Low Volatility, Momentum, Profitability, ROI, etc.
Re: Houston, we have a problem... [Small cap historical performance]
I just showed the math. The argument that changes in P/E multiple over time doesn't scale to explain large return differences over large periods of time. You need a 33x change in P/E multiple between 1927 and 2007 to explain the 5% return premium of SCV and a 6x change in P/E multiple to explain the 2.5% return premium of SB. It's very easy to get large changes because of P/E over 1 or 2 decades but not 8 decades. The math doesn't scale.JoMoney wrote:I've shown it before, but the 6% annualized difference between a small-cap index fund and a large-cap index fund over the past 15 year period can be entirely explained by relative changes in the P/E multiple over that time. I don't think it's unfeasible to believe the same situation could exist when looking at a 90 year period.
Monthly or yearly movements of stocks are often erratic and not indicative of changes in intrinsic value. Over time, however, stock prices and intrinsic value almost invariably converge. ~ WB
Re: Houston, we have a problem... [Small cap historical performance]
I wonder why PXSV has done so badly this year, much worse than the S&P500...
-
- Posts: 16022
- Joined: Thu Feb 22, 2007 7:28 am
- Location: St Louis MO
Re: Houston, we have a problem... [Small cap historical performance]
longinvest
While value has been debated as to source being risk vs. anomaly (behavioral) there has never been any such debate about small stocks with the small factor being considered a risk story. And the size premium has exited basically all over the world as well
Now the size premium is complicated because of the awful performance of small growth stocks with low profitability---which can be screened out-- but of course reduce the size premium
I hope that is helpful
Larry
While value has been debated as to source being risk vs. anomaly (behavioral) there has never been any such debate about small stocks with the small factor being considered a risk story. And the size premium has exited basically all over the world as well
Now the size premium is complicated because of the awful performance of small growth stocks with low profitability---which can be screened out-- but of course reduce the size premium
I hope that is helpful
Larry