Small-Cap Value has had higher returns than Total Stock Market, with lower risk

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply
JustinR
Posts: 406
Joined: Tue Apr 27, 2010 11:43 pm

Small-Cap Value has had higher returns than Total Stock Market, with lower risk

Post by JustinR » Sat Aug 26, 2017 11:12 pm

I apologize that I made a similar thread about Mid-Cap Value a few hours ago, but I just backtested Small-Cap and was wondering if people had opinions on going 100% SCV instead of TSM.

Using Simba's backtesting spreadsheet, I plugged in 80%TSM vs 80%SCV (with the rest in Total Bond).

You can download the file with my numbers already plugged in here: https://files.fm/u/tzbwgb6e

The SCV portfolio beats TSM for any time period and has less downside risk. You can try it with any time period.



Image



Image


Image



Is there something inherently better about TSM that I should keep my stocks in there and not move them to SCV? Currently I'm half TSM and half SCV for my domestic allocation.
Last edited by JustinR on Sun Aug 27, 2017 4:29 pm, edited 2 times in total.

kosomoto
Posts: 346
Joined: Tue Nov 24, 2015 8:51 pm

Re: Small-Cap Value has higher returns than Total Stock Market, with lower risk

Post by kosomoto » Sat Aug 26, 2017 11:19 pm

Because there's no guarantee that trend will continue?

I can come up with a million arguments for why it won't continue (or why it will).

Here's some for why it won't:

SCV's premium has been arbitraged away by a million SCV funds.

Larger companies are gaining more political power in the US and will use that power to benefit larger companies like themselves.

Etc, etc. Point is you can't predict which segment of the market will do best and there is a rational reason for each sector to outperform.

Theoretical
Posts: 1068
Joined: Tue Aug 19, 2014 10:09 pm

Re: Small-Cap Value has higher returns than Total Stock Market, with lower risk

Post by Theoretical » Sat Aug 26, 2017 11:50 pm

As a massive tilter personally (all Value, most small value including internationally), I completely disagree with you re: downside. Also 1970-2016 gives you all theee massive Small Value boosts. At a given equity percentage, Small Value is between 1/3 and 1/2 again as volatile as standard stocks. The reason I appreciate the Larry Portfolio is that it's a tilt that recognizes it's a lot riskier in the short term, so it has a much higher bond percentage to match. It's why I'm 32 and run a 60/40 and not an 80/20 or 90/10 portfolio.

Small value got BUTCHERED in 2008 and has underperformed the market grossly since about 2005 until 2016, where it pulled largely even after a big growth run.

The other thing is that during the Depression an obscene number of "Small Value" companies went under (even relative to other firms in the Depression) while the true blue chips tended to survive. That's a deep risk of a high-equity SCV strategy (that you're hurtling towards the abyss with no counterweight); with a low(er) equity SCV approach, you've got lots of bonds to get you through the abyss. With the large caps, a fair number will survive and most will keep some dividends (though cut).

JustinR
Posts: 406
Joined: Tue Apr 27, 2010 11:43 pm

Re: Small-Cap Value has higher returns than Total Stock Market, with lower risk

Post by JustinR » Sun Aug 27, 2017 12:06 am

Theoretical wrote:
Sat Aug 26, 2017 11:50 pm
As a massive tilter personally (all Value, most small value including internationally), I completely disagree with you re: downside. Also 1970-2016 gives you all theee massive Small Value boosts. At a given equity percentage, Small Value is between 1/3 and 1/2 again as volatile as standard stocks. The reason I appreciate the Larry Portfolio is that it's a tilt that recognizes it's a lot riskier in the short term, so it has a much higher bond percentage to match. It's why I'm 32 and run a 60/40 and not an 80/20 or 90/10 portfolio.

Small value got BUTCHERED in 2008 and has underperformed the market grossly since about 2005 until 2016, where it pulled largely even after a big growth run.

The other thing is that during the Depression an obscene number of "Small Value" companies went under (even relative to other firms in the Depression) while the true blue chips tended to survive. That's a deep risk of a high-equity SCV strategy (that you're hurtling towards the abyss with no counterweight); with a low(er) equity SCV approach, you've got lots of bonds to get you through the abyss. With the large caps, a fair number will survive and most will keep some dividends (though cut).
I tested 2005-2015 and it only did slightly worse than TSM. And the downside risk is still less. I don't think volatility matters as much as downside risk.

Image


Change 2005 to 2004 or 2015 to 2016, and SCV wins.

You can download the backtest file I'm using here: https://files.fm/u/tzbwgb6e

User avatar
privatefarmer
Posts: 276
Joined: Mon Sep 08, 2014 2:45 pm

Re: Small-Cap Value has higher returns than Total Stock Market, with lower risk

Post by privatefarmer » Sun Aug 27, 2017 1:01 am

not sure what data OP is using, SCV definitely has had higher standard deviation than TSM and a slightly worse drawdown during great recession, as we would expect. It is more volatile.

Amazon also has beaten the TSM over last 20 years, as has Apple and Microsoft. Don't use backtested data to make your decision going forward. It's depressing, I know, but the reality is that we do not and CANNOT know what companies will make the most profits going forward and thus have the highest returns. long-term returns are determined by future earnings (profits). period, end of story. if you know which specific companies are going to sell the most widgets and make the most profits then simply invest only in those companies. The P/E ratio acts as an "anchor", it keeps the stock price in line w/ future earnings. the only way for the stock price to increase in the long run is for earnings to increase by the same amount. The P/E may temporarily become overvalued or undervalued but it reverts back to it's mean over time.

In other words, unfortunately we do not know which companies are going to grow their profits the most going forward. Company size, current valuation, the favorite color of the CEO, etc etc does not tell us diddly squat about the future. This is why I just own everything at the absolute cheapest cost w/ the smallest amount of turnover, in other words total stock market index.

Valuethinker
Posts: 33143
Joined: Fri May 11, 2007 11:07 am

Re: Small-Cap Value has higher returns than Total Stock Market, with lower risk

Post by Valuethinker » Sun Aug 27, 2017 7:26 am

JustinR wrote:
Sat Aug 26, 2017 11:12 pm
I apologize that I made a similar thread about Mid-Cap Value a few hours ago, but I just backtested Small-Cap and was wondering if people had opinions on going 100% SCV instead of TSM.

Using Simba's backtesting spreadsheet, I plugged in 80%TSM vs 80%SCV (with the rest in Total Bond).

You can download the file with my numbers already plugged in here: https://files.fm/u/tzbwgb6e

The SCV portfolio beats TSM for any time period and has less downside risk. You can try it with any time period.



Is there something inherently better about TSM that I should keep my stocks in there and not move them to SCV? Currently I'm half TSM and half SCV for my domestic allocation.
Yes there is something inherent. The Small Cap Value effect is not consistent nor predictable. Past occurrence does not predict future frequency.

The great transactions costs & tax drag (the latter for taxpaying investors only) may be why SCV has a higher return. The market demands a higher return because of it. In other words it's fundamentally an illiquidity premia story.

But there's also the Fama explanation about higher risk, and I think this has legs (FWIW so does Larry Swedroe).

There was no certainty in 2008-09 that the rescues that took place, globally, would take place. Both the bailouts of financial institutions and the Keynesian fiscal impulse that was agreed among G8 Finance Ministers (under Prime Minister Gordon Brown's orchestration) in March of 2009. And the unprecedented relaxation of monetary conditions/ monetary policy (zero interest rates, quantitative easing etc.). They are not likely to be repeated: we have tested the limits of loose monetary policy, and laws re bailouts have been changed.

In the summer of 2007 the "bonfire of the quants" took place. The Goldman Sachs CFO made his famous remark about "six sigma events" (one chance in 25 million, from memory) with regard to the collapse of a Goldman fund invested in small cap value stocks. What happened is that a hedge fund had to liquidate its portfolio to cover client withdrawals, they liquidated their Small Cap Value stock portfolio, and that started a rout as the "crowded trade" all headed for the fire exit. We don't know officially, but it is believed it was a Clifford Asness/ AQR hedge fund. The rout only stopped when Goldman Sachs put some of its *own* balance sheet money into its hedge fund.

The nature of small cap value stocks is that these are highly geared, low profit margin companies (on average). Hence the low price to earnings, price to sales, price to book value (high Book/Market in econometrics) stocks. The market is not wrong to price these lowly, however it may underprice them. The Public-to-Private transactions of the Private Equity/ LBO industry is one way that this value is realized for quoted markets (but it also reduces the opportunities of an SCV portfolio of quoted stocks-- plenty of evidence there are fewer of these companies around now).

Clearly these adverse events don't only happen once in 25 million trading days. They we learned in the Great Financial Crisis was that the systemic effects are very important. One piece of the system fails, and it starts a run on all the others. Large companies with greater financial strength, access to credit markets, are in a better position to survive (one trick was simply to stop paying suppliers, which many of them did-- for any big company that sells to consumers, that's a huge source of cash flow right there).

80% SCV? Your stats won't capture the 1929s and the 2008s very well. Because they are generational events. But, just to worry you a bit, they seem to be happening more frequently ....

Jags4186
Posts: 1371
Joined: Wed Jun 18, 2014 7:12 pm

Re: Small-Cap Value has higher returns than Total Stock Market, with lower risk

Post by Jags4186 » Sun Aug 27, 2017 9:05 am

There's nothing wrong with tilting heavily as long as you stick to your guns. 25% of my portfolio is US SCV. It is not performing well this year and I can see that when I look at my "You Index" on Personal Capital vs. S&P500 and especially the Int'l markets. Every asset class has it's time in the Sun, but if you can't wait for it you will feel the pain.

User avatar
whodidntante
Posts: 2194
Joined: Thu Jan 21, 2016 11:11 pm

Re: Small-Cap Value has higher returns than Total Stock Market, with lower risk

Post by whodidntante » Sun Aug 27, 2017 9:13 am

Any time period? OK, I chose YTD. :happy

munemaker
Posts: 2103
Joined: Sat Jan 18, 2014 6:14 pm

Re: Small-Cap Value has higher returns than Total Stock Market, with lower risk

Post by munemaker » Sun Aug 27, 2017 9:40 am

privatefarmer wrote:
Sun Aug 27, 2017 1:01 am
not sure what data OP is using, SCV definitely has had higher standard deviation than TSM and a slightly worse drawdown during great recession, as we would expect. It is more volatile.

I agree. Higher return but at higher risk. You can go a little heavier on the SCV, but you should maintain good diversification.

User avatar
willthrill81
Posts: 2263
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Small-Cap Value has higher returns than Total Stock Market, with lower risk

Post by willthrill81 » Sun Aug 27, 2017 10:04 am

Valuethinker wrote:
Sun Aug 27, 2017 7:26 am
The Small Cap Value effect is not consistent nor predictable.
Considering that SCV has beaten LC in terms of returns in every twenty year period save one in the historic record, I'd say that's pretty darned consistent, at least consistent enough for the typical investor to experience the higher returns several times during their investment horizon.

Of course someone will bring up the whole independent period argument, but still SCV's outperformance of TSM has not been limited to a just a few isolated periods. Even just in recent decades, SCV just came up short of TSM from 2007-2017, but it outperformed in '97-'07, '87-'97, and '77-'87 by several percent at least.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

munemaker
Posts: 2103
Joined: Sat Jan 18, 2014 6:14 pm

Re: Small-Cap Value has higher returns than Total Stock Market, with lower risk

Post by munemaker » Sun Aug 27, 2017 10:12 am

Valuethinker wrote:
Sun Aug 27, 2017 7:26 am
The Small Cap Value effect is not consistent nor predictable.
Is anything about the stock market consistent and predictable?

User avatar
willthrill81
Posts: 2263
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Small-Cap Value has higher returns than Total Stock Market, with lower risk

Post by willthrill81 » Sun Aug 27, 2017 10:18 am

munemaker wrote:
Sun Aug 27, 2017 10:12 am
Valuethinker wrote:
Sun Aug 27, 2017 7:26 am
The Small Cap Value effect is not consistent nor predictable.
Is anything about the stock market consistent and predictable?
Over the very long-term, yes. Siegel found that across time and geography, long-term stock returns have been remarkably consistent. But over 10, 20, even 30 years, the consistency is less.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

Johnnie
Posts: 429
Joined: Sat May 28, 2016 3:18 pm
Location: Michigan

Re: Small-Cap Value has higher returns than Total Stock Market, with lower risk

Post by Johnnie » Sun Aug 27, 2017 10:21 am

How old are you, OP?

I'm a big value fan, but with an eye toward "long periods of under-performance," I'm looking to dial back my exposure a bit as I get closer to retirement (to 34% SCV, not counting what's in TSM and Total International). I don't have many "long periods" left, and don't want to wait until age 85 for my ship to come in (sorry kids).
"I know nothing."

User avatar
packer16
Posts: 884
Joined: Sat Jan 04, 2014 2:28 pm

Re: Small-Cap Value has higher returns than Total Stock Market, with lower risk

Post by packer16 » Sun Aug 27, 2017 10:34 am

What you are observing is data mining (observing an outperformance in the past due to some factor). A trend found from historical data that shows what has generated past returns but it in may not continue in the future. Historically, I was a SCV believer & wanted to test this hypothesis with data after the publication of this finding by a popular author. Larry Swedroe wrote an article in the NY Times in late 2011 observing the same thing you have, SCV has higher expected return than TSM. Thus you could have the same expected return with lower risk by investing a lower portion is SCV vs. SP500 in a balance portfolio. So I looked at the returns of such a strategy & got some pretty alarming results. The actual returns for the SCV balanced portfolio were significantly lower than the expected returns.

An example is the returns of a 60/40 SP500 post-publication (2012 to date) was 70% vs. the 50/50 SCV which was 48%. Even more std dev of the SCV portfolio was higher than the SP500 portfolio.

Even if we go back to when DFA introduced the first SCV fund, the outperformance of a spliced DFA/Vangaurd data set was only 0.8%/year. Not surprisingly about the same amount (0.9%/year) of outperformance Bogle observed in his 1993 book between value & growth mutual funds over the previous 20 years. To provide some perspective, the outperformance of SP500 vs. the Vanguard Total Bond Fund over the same period was 3%/year, or 3x the amount. So since SCV has been investable the performance delta has been 0.8% & heavily dependent upon timing.

IMO what is lost is what is a SCV security? It is a security that has a low price/accounting metric value. These characteristics are not fundamental to a firm but change over time as the price of a security changes. My issue is that the accounting metrics, while in general reflect the economics of a business they leave out some important aspects (such as valuation of internally developed intangibles and accounting for acquired intangibles) and do not incorporate future changes in the business. As a business appraiser, using accounting metrics is one method amongst many to value a security & thus this general approach to identify security returns leaves alot of open holes on the micro/accounting side that are not addressed in the SCV screens.

The other issue is market efficiency. When you are buying a SCV security, you are betting that market has mispriced this security now & what you are seeing in the past data is that they have been mispriced in the past. If we agree the market is getting more efficient especially as more money has flowed into SCV as a strategy the 0.8% of the past could easily be arbed away by now & may be as more recent returns are showing worse performance. The amount of premium is debatable but is the timing risk of SCV worth a 0.8%/year & declining SCV premium? That is a question you need to ask. For me the answer is no. There is a thread below on an article by Morgan Housel that provides some nice insights on the market & how adaptive it is that addresses the past vs. future performance issues.

Packer
Buy cheap and something good might happen

mongstradamus
Posts: 102
Joined: Mon Nov 04, 2013 11:46 pm

Re: Small-Cap Value has higher returns than Total Stock Market, with lower risk

Post by mongstradamus » Sun Aug 27, 2017 11:08 am

I have a very noobie question, what are the best metrics you use or anybody in general should use to measure downside risk or risk in general ? A lot of this seems to be going over my head

stan1
Posts: 4963
Joined: Mon Oct 08, 2007 4:35 pm

Re: Small-Cap Value has higher returns than Total Stock Market, with lower risk

Post by stan1 » Sun Aug 27, 2017 11:17 am

I realize this is not a quantitative strategy but below is why I tilt away from Total Stock Market:

Month-end ten largest holdings
(16.5% of total net assets) as of 07/31/2017
1 Apple Inc.
2 Microsoft Corp.
3 Alphabet Inc.
4 Amazon.com Inc.
5 Facebook Inc.
6 Johnson & Johnson
7 Berkshire Hathaway Inc.
8 Exxon Mobil Corp.
9 JPMorgan Chase & Co.
10 Wells Fargo & Co.

This looks like risk and lack of diversification. I would not sleep well at night. At various other times other industries have had similar dominance.

nura
Posts: 78
Joined: Mon Jun 27, 2016 11:24 am

Re: Small-Cap Value has higher returns than Total Stock Market, with lower risk

Post by nura » Sun Aug 27, 2017 11:28 am

JustinR wrote:
Sat Aug 26, 2017 11:12 pm
I apologize that I made a similar thread about Mid-Cap Value a few hours ago, but I just backtested Small-Cap and was wondering if people had opinions on going 100% SCV instead of TSM.
A while back, I back-tested 50% each SCV & LCG re-balanced quarterly offered superior risk-adjsuted returns over TSM and SCV separately for past 5,10 and 20 years. In fact you may find many other combiantion portfolios that out-perform both TSM and SCV in the past 5-20 years. But all this is historical data which is not quaranteed to repeat in future.

User avatar
JoMoney
Posts: 4397
Joined: Tue Jul 23, 2013 5:31 am

Re: Small-Cap Value has higher returns than Total Stock Market, with lower risk

Post by JoMoney » Sun Aug 27, 2017 12:01 pm

Over the past 3,5,10, and 15 year period(s) DFSVX which is typically pointed to as the premier 'Small-Cap Value' fund has had a lower Sharpe Ratio indicating higher risk and lower risk-adjusted returns using the traditional measurements
Image

The whole idea of factor investing is based on a premise that "small" and "value" are unique risk/return "factors" that are not properly measured with the traditional methods, and any difference in returns is a "risk premium" for taking ADDITIONAL RISK that has different skew and kurtosis in the way the risk/returns are distributed different from 'market beta'.
"To achieve satisfactory investment results is easier than most people realize; to achieve superior results is harder than it looks." - Benjamin Graham

User avatar
Alexa9
Posts: 391
Joined: Tue Aug 30, 2016 9:41 am

Re: Small-Cap Value has higher returns than Total Stock Market, with lower risk

Post by Alexa9 » Sun Aug 27, 2017 12:05 pm

Market weight is not sufficiently balanced for me with cap weighting or value weighting.
I wouldn't go 100% SCV (including international) but Larry Swedroe does if you're interested in reading his research.

User avatar
willthrill81
Posts: 2263
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Small-Cap Value has higher returns than Total Stock Market, with lower risk

Post by willthrill81 » Sun Aug 27, 2017 12:45 pm

mongstradamus wrote:
Sun Aug 27, 2017 11:08 am
I have a very noobie question, what are the best metrics you use or anybody in general should use to measure downside risk or risk in general ? A lot of this seems to be going over my head
"Best" is a loaded term, but risk is usually measured in standard deviation. Downside risk is usually measured with maximum drawdown (percentage drop from peak to trough).

Over the last 45 years, SCV has had both a higher standard deviation than TSM, along with a higher max. drawdown.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

mongstradamus
Posts: 102
Joined: Mon Nov 04, 2013 11:46 pm

Re: Small-Cap Value has higher returns than Total Stock Market, with lower risk

Post by mongstradamus » Sun Aug 27, 2017 1:13 pm

[quote=willthrill81 post_id=3509067 time=1503855933 user_id=116799

"Best" is a loaded term, but risk is usually measured in standard deviation. Downside risk is usually measured with maximum drawdown (percentage drop from peak to trough).

Over the last 45 years, SCV has had both a higher standard deviation than TSM, along with a higher max. drawdown.
[/quote]
Thank you for the explanation I was just curious on what I should look at when comparing funds when it comes to risk . For example if I wanted to compare Wellington vs vanguard balanced vs life strategy moderate growth fund.

User avatar
willthrill81
Posts: 2263
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Small-Cap Value has higher returns than Total Stock Market, with lower risk

Post by willthrill81 » Sun Aug 27, 2017 2:12 pm

mongstradamus wrote:
Sun Aug 27, 2017 1:13 pm
willthrill81 wrote:
Sun Aug 27, 2017 12:45 pm
"Best" is a loaded term, but risk is usually measured in standard deviation. Downside risk is usually measured with maximum drawdown (percentage drop from peak to trough).

Over the last 45 years, SCV has had both a higher standard deviation than TSM, along with a higher max. drawdown.
Thank you for the explanation I was just curious on what I should look at when comparing funds when it comes to risk . For example if I wanted to compare Wellington vs vanguard balanced vs life strategy moderate growth fund.
Portfolio Visualizer is a fantastic free tool that can be used to analyze many aspects of historic performance of both asset classes and specific securities.

You'll find that since 1995, the inception year of the LS Moderate Growth fund, all of those funds have had very comparable volatility and max. drawdown, although the LS fund has been somewhat higher than the other two and had lower returns to boot, largely due to its international component.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

garlandwhizzer
Posts: 1626
Joined: Fri Aug 06, 2010 3:42 pm

Re: Small-Cap Value has higher returns than Total Stock Market, with lower risk

Post by garlandwhizzer » Sun Aug 27, 2017 2:37 pm

Sorry I posted in the wrong space originally.

The question is not do cost-free long/short SCV indexes outperform long only TSM over long periods of time. That does tend to happen on a regular basis. The important question is whether that outperformance is harvestable on an after cost basis for investors. There are always SCV funds that can be cherry picked on backtesting over a specific period of time that beat TSM. Again, that is not the question. The question is does the average, not the best picked in retrospect, SCV fund reliably beat a low cost TSM index fund? That is a different question.

Another question worth asking is how does one define SCV? Should it be microcap (maximize the small factor) and deep value (maximize the value factor)? The real etf that does that is RZV which is a microcap deep value fund. RZV has underperformed VTI YTD, 1YR, 3YR, 5YR, 10 YR, and since its inception about 12 years ago. It has also shown considerably more volatility and suffered greater drawdowns in bear markets along the way. The fund that supposedly maximizes the SCV premium has been a consistent loser to simple beta, no small, no value for 12 years. Factor fundamentalist believers will not be deterred by this and will continue to wait patiently for the payoff. Others ask questions. Should our SCV criteria screen out junk and emphasize QUAL? Should it screen out strongly negative MOM? Should it increase average cap weight from volatile/vulnerable micro-cap and include some MCV which has been less volatile and still has provided similar, if not better returns?

How is the best way to define SCV that is harvestable after cost--what average cap weight, should we use P/B, P/E, P/CF, or DIV to define value, should be use screens (QUAL, MOM), how much portfolio turnover should we employ, keeping in mind trading costs and frictions can be expensive in the SC space, and finally should we do long/short portfolios which are quite expensive in the SC space, or long only portfolios? All of these decisions have to be made prospectively. And there are many funds with many differing approaches. On backtesting you get to pick the SCV fund that has done best over a given time frame, that is to say it has selected the magic set of parameters that worked optimally over that time frame. Will that same set work equally well going forward? IMO, we don't know that with certainty up front. P/B alone worked like a charm until it was described. Now it has been largely abandoned as a sole marker of SCV. As more factor seeking money pours into the SCV space, which has capacity restraints, no more than 3% of the market cap weight, it seems reasonable to conclude some degree of premium dilution going forward. It also seems reasonable that not all funds will pick the perfect criteria to harvest the premium after costs.

In summary, because graphs of cost-free SCV indexes, selected in the rear view mirror to optimize performance, look much more impressive than TSM, it may not be reasonable to conclude that the next 15 years will turn out likewise. Personally I don't think the certainty of outperformance on anything justifies a bet of 80% of your portfolio weight on something with only 20% bonds. That IMO is not an all weather portfolio regardless of what backtesting suggests. That is a high risk portfolio that, given the inherent uncertainty of market action, is dangerous. Larry's uses highly tilted portfolios but he advocates a large high quality bond allocation to provide a safety net. 20% bonds seems to fall well short of that, even if you're a true believer in SCV.

Garland Whizzer

User avatar
FIREchief
Posts: 1363
Joined: Fri Aug 19, 2016 6:40 pm

Re: Small-Cap Value has higher returns than Total Stock Market, with lower risk

Post by FIREchief » Sun Aug 27, 2017 2:49 pm

JustinR wrote:
Sat Aug 26, 2017 11:12 pm
I apologize that I made a similar thread about Mid-Cap Value a few hours ago, but I just backtested Small-Cap and was wondering if people had opinions on going 100% SCV instead of TSM.

Using Simba's backtesting spreadsheet, I plugged in 80%TSM vs 80%SCV (with the rest in Total Bond).
Lost me at "backtesting."
I am not a lawyer, accountant or financial advisor. Any advice or suggestions that I may provide shall be considered for entertainment purposes only.

User avatar
Robert T
Posts: 2445
Joined: Tue Feb 27, 2007 9:40 pm
Location: 1, 0.2, 0.4, 0.5
Contact:

Re: Small-Cap Value has higher returns than Total Stock Market, with lower risk

Post by Robert T » Sun Aug 27, 2017 2:55 pm

.
Small value has higher risk: e.g. 2007-08 drawdown
https://www.portfoliovisualizer.com/bac ... ion3_3=100

-51% = Vanguard 500 Index
-55% = DFA US Small Cap
-61% = DFA US Small Cap Value
-72% = Guggenheim S&P Smallcap 600 Pure Value

Fama interview may also be of interest https://www.youtube.com/watch?v=HIKO-t4vU6Q

I would just note that the Fama-French article on “Common risk factors in the returns on stocks and bonds” was published in 1993 [over 24 years ago], the same year as inception of the DFA US Small Value fund.

Annualized returns since then (from above link):

9.3% = Vanguard 500 Index [VFINX]
10.7% = DFA US Small Cap [DFSTX]
11.8% = DFA US Small Cap Value [DFSVX]

Obviously no guarantees.

Robert
.

dbr
Posts: 23709
Joined: Sun Mar 04, 2007 9:50 am

Re: Small-Cap Value has higher returns than Total Stock Market, with lower risk

Post by dbr » Sun Aug 27, 2017 2:59 pm

FIREchief wrote:
Sun Aug 27, 2017 2:49 pm
JustinR wrote:
Sat Aug 26, 2017 11:12 pm
I apologize that I made a similar thread about Mid-Cap Value a few hours ago, but I just backtested Small-Cap and was wondering if people had opinions on going 100% SCV instead of TSM.

Using Simba's backtesting spreadsheet, I plugged in 80%TSM vs 80%SCV (with the rest in Total Bond).
Lost me at "backtesting."
"Backtesting" means that you look at the performance of some holding as it
actually performed in some period in past. Backtesting is deceptively attractive because the results are fact, similar to saying that the Cowboys beat the Packers by 30-16 on Oct 16, 2106. If that sounds like too much of a one-off the data could be that the Cowboys went 13-3 and the Packers 10-6 last year. If you want something more durable than that you could cite the result that the all time record between the two teams is Packers over Cowboys 18-17-0.

Now if the question is which city should you and your brother split to to have NFL bragging rights over the other for the next twenty years, which of this data is useful to decide this? Would you make the same analysis if it were a question of getting NFL bragging rights by moving to San Diego?

User avatar
whodidntante
Posts: 2194
Joined: Thu Jan 21, 2016 11:11 pm

Re: Small-Cap Value has higher returns than Total Stock Market, with lower risk

Post by whodidntante » Sun Aug 27, 2017 3:09 pm

mongstradamus wrote:
Sun Aug 27, 2017 11:08 am
I have a very noobie question, what are the best metrics you use or anybody in general should use to measure downside risk or risk in general ? A lot of this seems to be going over my head
There are a few ways and it just depends on what you are trying to measure.

Standard deviation is probably most common.
Alpha is a measure of risk adjusted return in reference to a benchmark.
Beta measures the tendency of fund to swing with a benchmark.
R-squared if you want to compare how closely something tracks a benchmark.
Sharpe ratio is also a measure of risk adjusted return.

All of the above are backward looking. Implied volatility is forward looking. If the market "thinks" Tesla will go down hard you'll observe that a put option will be expensive to buy.

User avatar
nisiprius
Advisory Board
Posts: 34134
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: Small-Cap Value has higher returns than Total Stock Market, with lower risk

Post by nisiprius » Sun Aug 27, 2017 3:58 pm

JustinR, it would have only taken four more keystrokes to have titled the thread "Small-Cap Value has had higher returns than Total Stock Market, with lower risk." Setting aside the question of whether it really has had lower risk--I don't believe it for a minute but I'm away from my historical data sources at the moment. However, small-cap had dramatically more risk 1929-1936 and I'm pretty sure small-cap value was not any better.

The point is, always to use the correct verb tense because while the past is our most-frequently-used guide to the future, we need to be constantly alert to the realization that we need to keep asking the question: how reliable a predictor is it, and even if we assume that the stock market of 1929 or 1950 or 1990 is "the same thing" and has the same statistical parameters as the stock market of 2017, even on the face of it, how accurately do we really know the underlying parameters that describe stock market returns? For example, in a post some year ago I pointed out that even such a basic number as "the historic return of the stock market," even over periods of >80 years, varies from 9% to 11% if you just change the endpoints by a few years. The problem is that financial data is very badly behaved and constantly undergoes such extreme variation that a single historical event can be so huge and can carry so much weight that they can make huge changes in average return, standard deviation, correlation, etc. even over periods of many decades.

One thing that cannot be ignored with small-cap, let along small-cap value, is that it was almost uninvestible (as a whole, as in an index fund), up until perhaps 1980 or so. Another thing is that the accuracy of the historical data is somewhat questionable--the original paper that called attention to the "small-company effect" had serious problems with the data used, and although CRSP has since tried to correct it, I can't judge how reliable or accurate the "corrections" are.

The other thing is the "decline effect." In Berkin and Swedroe's "Your Complete Guide to Factor-Based Investing," they quote academic research as showing that factor premia have declined in size by about 30% since their discovery and publicization. They regard this optimistically--the premia declined in size but continued to persist. But it still shows that there is a decline effect. I don't know what they have to say about commodities in the form of CCF's but that's an apparent example of a much stronger effect. Two papers published in IIRC 2006 said there was a dramatic benefit to including them in a portfolio, and used mouthwatering backtested data to show that if people had been able to invest in them in the past they would have enjoyed a big portfolio boost. But what actually happened after 2006 was that interest in commodities spurred the development of commodity mutual funds and ETFs--and were widely adopted, e.g. in many fund company's target-date retirement funds--and the effect basically vanished completely just as soon as it became widely known and exploited.
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.

dbr
Posts: 23709
Joined: Sun Mar 04, 2007 9:50 am

Re: Small-Cap Value has higher returns than Total Stock Market, with lower risk

Post by dbr » Sun Aug 27, 2017 4:11 pm

mongstradamus wrote:
Sun Aug 27, 2017 11:08 am
I have a very noobie question, what are the best metrics you use or anybody in general should use to measure downside risk or risk in general ? A lot of this seems to be going over my head
There is no best. It is a question of what is helpful to answer what question one has or to best frame an understanding of how things work.

An academic approach to investing looks at investment returns in a year (or any period of time) as a random sample from a distribution of possible returns. Therefore the subject of interest is what does that distribution look like. Generally in statistics one talks about distributions according to their shape, that is what kind of distribution it is, and according to the numerical values of parameters that specify the shape. Many useful distributions can be specified by one or two or three parameters. One of the really useful shapes of distributions is the normal distribution. So if we see those invested returns as being normally distributed we can specify the whole picture by knowing the mean and the standard deviation of the distribution. Just to make sure the conversation is between academic students of finance and is intended to by privy to specialists we will not use those generic name but instead call the mean expected return and the standard deviation risk. Notice this is not really "defining" risk as standard deviation. It is really just talking about the problem in a very conventional way but using different words.

As to downside risk, if you want to know how strongly negative a return might pop out in a given year, you already have that answer if you think the return is described by a normal distribution and you know the mean and standard deviation of that distribution. It is just a standard calculation that comes along with doing the statistics. From that point of view it makes no sense to say that risk isn't downside but rather it is standard deviation or vice versa. The two are just part of one big picture. Now, if you are concerned that the real world is not well described by a normal distribution, or that our estimates of the mean and standard deviation are too approximate, or that the distribution changes from year to year, and so on, then you have some reasonable concerns. Ascertaining distributions that account well for average trends is a lot easier than ascertaining distributions that account for extremes. It might, for example, be better to just look at the extremes themselves and see if we can characterize them. Actual investment returns are known for having a higher incidence of extremes than normal distributions predict.

Still a different issue from risk in annual returns is trying to ascertain the uncertainty in returns compounded over a period of time. For normal distributions that stay the same over time and where the return is a sample independent in each year from what it was the year before one can calculate the distribution of final outcomes. For this reason standard deviation of annual returns, a short term measure, is closely linked to long term results. It is by no means to be ignored. Contained within this calculation are such controversial discussions of arithmetic and geometric returns, all of which is convoluted conversation about things that are straightforward mathematics. Trickier issues arise regarding whether returns derive from the same distribution over time and whether or not successive periods are independent.

JustinR
Posts: 406
Joined: Tue Apr 27, 2010 11:43 pm

Re: Small-Cap Value has higher returns than Total Stock Market, with lower risk

Post by JustinR » Sun Aug 27, 2017 4:16 pm

Ah oops I messed up
Last edited by JustinR on Sun Aug 27, 2017 4:21 pm, edited 1 time in total.

staythecourse
Posts: 4964
Joined: Mon Jan 03, 2011 9:40 am

Re: Small-Cap Value has higher returns than Total Stock Market, with lower risk

Post by staythecourse » Sun Aug 27, 2017 4:19 pm

mongstradamus wrote:
Sun Aug 27, 2017 11:08 am
I have a very noobie question, what are the best metrics you use or anybody in general should use to measure downside risk or risk in general ? A lot of this seems to be going over my head
I'll go with the ONLY risk that really matters and that is not having the money when you need it. So when investing it doesn't matter what ANY of the metrics say. You can't eat or live off metrics. Only thing that matters is what will you do if that risk (any risk) shows up and as murphy's law has it usually at the most inopportune time. The best diversify of ALL risk time. IF you have the time horizon for that need of money invested in XYZ it really doesn't matter then what happens. If you need any of that money to live off then one is taking on risk. Simple as that. As Shakespeare said, "Time heals all wounds".

Good luck.
"The stock market [fluctuation], therefore, is noise. A giant distraction from the business of investing.” | -Jack Bogle

JustinR
Posts: 406
Joined: Tue Apr 27, 2010 11:43 pm

Re: Small-Cap Value has higher returns than Total Stock Market, with lower risk

Post by JustinR » Sun Aug 27, 2017 4:23 pm

mongstradamus wrote:
Sun Aug 27, 2017 11:08 am
I have a very noobie question, what are the best metrics you use or anybody in general should use to measure downside risk or risk in general ? A lot of this seems to be going over my head
Look up Sortino Ratio: http://www.investopedia.com/terms/s/sortinoratio.asp
The Sortino ratio is a useful way for investors, analysts and portfolio managers to evaluate an investment's return for a given level of bad risk. Since this ratio uses the downside deviation as its risk measure, it addresses the problem of using total risk, or standard deviation, as upside volatility is beneficial to investors.

A ratio such as the Sharpe ratio punishes the investment for good risk, which provides positive returns for investors. However, determining which ratio to use depends on whether the investor wants to focus on standard deviation or downside deviation.
That's downside risk. To me, it's the best metric of risk because it factors in downwards volatility, while standard deviation and Sharpe Ratio both factor in upwards volatility (which is irrelevant when measuring risk).

So in the spreadsheet, you can see the this represented as "Sortino Ratio" and "Downside SD" in the graph and charts below.

You can see how SCV has higher returns yet lower risk than TSM for almost every time period if you play around with the spreadsheet: https://files.fm/u/tzbwgb6e

Image

Image


nisiprius wrote:
Sun Aug 27, 2017 3:58 pm
JustinR, it would have only taken four more keystrokes to have titled the thread "Small-Cap Value has had higher returns than Total Stock Market, with lower risk." Setting aside the question of whether it really has had lower risk--I don't believe it for a minute but I'm away from my historical data sources at the moment. However, small-cap had dramatically more risk 1929-1936 and I'm pretty sure small-cap value was not any better.
I changed the title, thanks.

Please review my backtesting and look at the Sortino Ratio/Downside SD metrics for SCV vs TSM.

JustinR
Posts: 406
Joined: Tue Apr 27, 2010 11:43 pm

Re: Small-Cap Value has higher returns than Total Stock Market, with lower risk

Post by JustinR » Sun Aug 27, 2017 4:45 pm

JoMoney wrote:
Sun Aug 27, 2017 12:01 pm
Over the past 3,5,10, and 15 year period(s) DFSVX which is typically pointed to as the premier 'Small-Cap Value' fund has had a lower Sharpe Ratio indicating higher risk and lower risk-adjusted returns using the traditional measurements
Image

The whole idea of factor investing is based on a premise that "small" and "value" are unique risk/return "factors" that are not properly measured with the traditional methods, and any difference in returns is a "risk premium" for taking ADDITIONAL RISK that has different skew and kurtosis in the way the risk/returns are distributed different from 'market beta'.
For the backtesting I used VTSMX (tsm) and VISVX (scv). For the last 15 years I get this:

Image

Image

Valuethinker
Posts: 33143
Joined: Fri May 11, 2007 11:07 am

Re: Small-Cap Value has higher returns than Total Stock Market, with lower risk

Post by Valuethinker » Sun Aug 27, 2017 5:07 pm

willthrill81 wrote:
Sun Aug 27, 2017 12:45 pm
mongstradamus wrote:
Sun Aug 27, 2017 11:08 am
I have a very noobie question, what are the best metrics you use or anybody in general should use to measure downside risk or risk in general ? A lot of this seems to be going over my head
"Best" is a loaded term, but risk is usually measured in standard deviation. Downside risk is usually measured with maximum drawdown (percentage drop from peak to trough).

Over the last 45 years, SCV has had both a higher standard deviation than TSM, along with a higher max. drawdown.
+1

Standard deviation of annual returns is One common measure.

Peak to trough is a good One.

For equities the recent slump is a good start.

Bonds gets harder but 1994 was bad. Inflation confuses the picture but 1980 81 was bad.

To scare oneself Japan from 1990. The UK market fell c 80 per cent in real terms 1972to 74.

dbr
Posts: 23709
Joined: Sun Mar 04, 2007 9:50 am

Re: Small-Cap Value has higher returns than Total Stock Market, with lower risk

Post by dbr » Sun Aug 27, 2017 5:21 pm

Valuethinker wrote:
Sun Aug 27, 2017 5:07 pm
willthrill81 wrote:
Sun Aug 27, 2017 12:45 pm
mongstradamus wrote:
Sun Aug 27, 2017 11:08 am
I have a very noobie question, what are the best metrics you use or anybody in general should use to measure downside risk or risk in general ? A lot of this seems to be going over my head
"Best" is a loaded term, but risk is usually measured in standard deviation. Downside risk is usually measured with maximum drawdown (percentage drop from peak to trough).

Over the last 45 years, SCV has had both a higher standard deviation than TSM, along with a higher max. drawdown.
+1

Standard deviation of annual returns is One common measure.

Peak to trough is a good One.

For equities the recent slump is a good start.

Bonds gets harder but 1994 was bad. Inflation confuses the picture but 1980 81 was bad.

To scare oneself Japan from 1990. The UK market fell c 80 per cent in real terms 1972to 74.
The "Larry" portfolio which tilts heavily to SCV and then dilutes risk by increasing allocation to bonds is supposed to have he advantage of specifically reducing the incidence of large negative returns, a concept expressed as "reduced low tail risk." This is not a recommendation in particular.

avalpert
Posts: 6193
Joined: Sat Mar 22, 2008 4:58 pm

Re: Small-Cap Value has higher returns than Total Stock Market, with lower risk

Post by avalpert » Sun Aug 27, 2017 6:14 pm

JustinR wrote:
Sun Aug 27, 2017 4:45 pm
JoMoney wrote:
Sun Aug 27, 2017 12:01 pm
Over the past 3,5,10, and 15 year period(s) DFSVX which is typically pointed to as the premier 'Small-Cap Value' fund has had a lower Sharpe Ratio indicating higher risk and lower risk-adjusted returns using the traditional measurements
Image

The whole idea of factor investing is based on a premise that "small" and "value" are unique risk/return "factors" that are not properly measured with the traditional methods, and any difference in returns is a "risk premium" for taking ADDITIONAL RISK that has different skew and kurtosis in the way the risk/returns are distributed different from 'market beta'.
For the backtesting I used VTSMX (tsm) and VISVX (scv). For the last 15 years I get this:

Image

Image
When I plug those two into portfoliovisualizer for the same period the returns are the same as yours but the stdev are 14.73 and 18.57 respectively with VTSMX being lower.

Maybe you are making a mistake in your calculations?

mongstradamus
Posts: 102
Joined: Mon Nov 04, 2013 11:46 pm

Re: Small-Cap Value has higher returns than Total Stock Market, with lower risk

Post by mongstradamus » Sun Aug 27, 2017 6:45 pm

willthrill81 wrote:
Sun Aug 27, 2017 2:12 pm
mongstradamus wrote:
Sun Aug 27, 2017 1:13 pm
willthrill81 wrote:
Sun Aug 27, 2017 12:45 pm
"Best" is a loaded term, but risk is usually measured in standard deviation. Downside risk is usually measured with maximum drawdown (percentage drop from peak to trough).

Over the last 45 years, SCV has had both a higher standard deviation than TSM, along with a higher max. drawdown.
Thank you for the explanation I was just curious on what I should look at when comparing funds when it comes to risk . For example if I wanted to compare Wellington vs vanguard balanced vs life strategy moderate growth fund.
Portfolio Visualizer is a fantastic free tool that can be used to analyze many aspects of historic performance of both asset classes and specific securities.

You'll find that since 1995, the inception year of the LS Moderate Growth fund, all of those funds have had very comparable volatility and max. drawdown, although the LS fund has been somewhat higher than the other two and had lower returns to boot, largely due to its international component.
It is an interesting tool I am just still trying to wrap my head around all the numbers. I was trying to find a good measurement of risk vs performance. I was trying to understand sharpe and sortino ratio. I also threw in the portfolio i am using into the mix just to see if i am being overly risky.

User avatar
willthrill81
Posts: 2263
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Small-Cap Value has higher returns than Total Stock Market, with lower risk

Post by willthrill81 » Sun Aug 27, 2017 6:55 pm

mongstradamus wrote:
Sun Aug 27, 2017 6:45 pm
willthrill81 wrote:
Sun Aug 27, 2017 2:12 pm
mongstradamus wrote:
Sun Aug 27, 2017 1:13 pm
willthrill81 wrote:
Sun Aug 27, 2017 12:45 pm
"Best" is a loaded term, but risk is usually measured in standard deviation. Downside risk is usually measured with maximum drawdown (percentage drop from peak to trough).

Over the last 45 years, SCV has had both a higher standard deviation than TSM, along with a higher max. drawdown.
Thank you for the explanation I was just curious on what I should look at when comparing funds when it comes to risk . For example if I wanted to compare Wellington vs vanguard balanced vs life strategy moderate growth fund.
Portfolio Visualizer is a fantastic free tool that can be used to analyze many aspects of historic performance of both asset classes and specific securities.

You'll find that since 1995, the inception year of the LS Moderate Growth fund, all of those funds have had very comparable volatility and max. drawdown, although the LS fund has been somewhat higher than the other two and had lower returns to boot, largely due to its international component.
It is an interesting tool I am just still trying to wrap my head around all the numbers. I was trying to find a good measurement of risk vs performance. I was trying to understand sharpe and sortino ratio. I also threw in the portfolio i am using into the mix just to see if i am being overly risky.
I don't find the Sharpe or Sortino ratios particularly useful from a behavioral standpoint. I personally think that maximum drawdown is the single best measure of risk because it's what generally causes people to derail their investment plans. For instance, if you could handle the maximum drawdown, paying attention to the actual dollar amount you would have lost, in periods like 2000-2002 and 2008-2009, then you're probably good to go.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

Dude2
Posts: 672
Joined: Fri Jun 08, 2007 3:40 pm

Re: Small-Cap Value has higher returns than Total Stock Market, with lower risk

Post by Dude2 » Sun Aug 27, 2017 7:07 pm

mongstradamus wrote:
Sun Aug 27, 2017 1:13 pm
willthrill81 wrote:
Sun Aug 27, 2017 12:45 pm

"Best" is a loaded term, but risk is usually measured in standard deviation. Downside risk is usually measured with maximum drawdown (percentage drop from peak to trough).

Over the last 45 years, SCV has had both a higher standard deviation than TSM, along with a higher max. drawdown.
Thank you for the explanation I was just curious on what I should look at when comparing funds when it comes to risk . For example if I wanted to compare Wellington vs vanguard balanced vs life strategy moderate growth fund.
It is unfortunate that folks use the term risk and mean volatility or standard deviation (or I guess max drawdown?). They do it because they need some measurable quantity that can be understood. However risk is far more nuanced than that.

There may be a bonus in performance for small cap value, probably due to the "cost of capital". The stock has to attract inflows by juicing its returns. Otherwise nobody would buy it, e.g. like junk bonds which need to make themselves attractive with a higher coupon or nobody would buy them.

These stocks are riskier. They pay a premium (via better returns) for you to invest in them. Whether they have higher or lower volatility could be another matter entirely. Sometimes the risks will show up. Other times you will be compensated for the risks.

User avatar
siamond
Posts: 3429
Joined: Mon May 28, 2012 5:50 am

Re: Small-Cap Value has had higher returns than Total Stock Market, with lower risk

Post by siamond » Sun Aug 27, 2017 7:47 pm

JustinR wrote:
Sat Aug 26, 2017 11:12 pm
Is there something inherently better about TSM that I should keep my stocks in there and not move them to SCV? Currently I'm half TSM and half SCV for my domestic allocation.
Hey Justin. I'm glad to see somebody making such good use of the Simba spreadsheet, this is cool. Your observations about the past are perfectly correct. I think this should provide positive reinforcement for your (already very aggressive) tilt towards SCV (50% TSM, 50% SCV). I do NOT think this should lead you to a more extreme allocation though. As multiple posters tried to explain, the SCV track record has been quite remarkable and yet went all over the place with great decades, poor decades, sideway decades, it's kind of hard to see a clear pattern in there. It's just a bad idea to put all your eggs in one (not so large, and heavily used) basket... You might actually consider diversifying your tilt between SCV and MCV (I do, and this makes me feel much more comfortable).

Let me suggest you draw a telltale chart, using 100% SCV as P1, 100% MCV as P2, and 100% TSM as benchmark. Start it early in history (e.g. 1930), then later (eg. 1970), and take your time to think about the trajectory, the time frames involved (decades!) and how you might have reacted without the benefit of hindsight. I don't know what the future will be made of, but in the past, it would have taken nerves of titanium to deal with SCV ups and downs relative to TSM.

As to a few questions that were asked in this thread:
- Justin didn't say that SCV displayed lower volatility (regular standard deviation), and this was NOT the case. He honed on other risk metrics (e.g. downside SD, Sharpe ratio, etc) and rightfully observed that SCV displayed cool properties in this respect in the past few decades (e.g. 1970+).
- The data sources are the ones used in the Simba backtesting spreadsheet. Check the spreadsheet 'Data_Sources' tab and you'll find the answer.
- Portfolio Visualizer, when it switched to the use of monthly returns, sidetracked a bit from Simba in terms of data sources. This explains the discrepancies (think of it this way: there are multiple indices for SCV, they all displayed a fascinating premium in the past, but the exact data series are all somewhat different, depending on the exact definition of a given index).

User avatar
siamond
Posts: 3429
Joined: Mon May 28, 2012 5:50 am

Re: Small-Cap Value has higher returns than Total Stock Market, with lower risk

Post by siamond » Sun Aug 27, 2017 7:57 pm

willthrill81 wrote:
Sun Aug 27, 2017 6:55 pm
I don't find the Sharpe or Sortino ratios particularly useful from a behavioral standpoint. I personally think that maximum drawdown is the single best measure of risk because it's what generally causes people to derail their investment plans. For instance, if you could handle the maximum drawdown, paying attention to the actual dollar amount you would have lost, in periods like 2000-2002 and 2008-2009, then you're probably good to go.
I would tend to agree with you. Drives me crazy to see so many people equating 'risk' with standard deviation, and worse, putting std-deviation and returns at the same level of importance while computing a ratio like Sharpe. Did you take a close look to the Ulcer Index/Ratio? This is quite an interesting idea, a creative variation on the idea of max drawdown (which has the drawback of being a single data point in history). Check this link.

JustinR
Posts: 406
Joined: Tue Apr 27, 2010 11:43 pm

Re: Small-Cap Value has higher returns than Total Stock Market, with lower risk

Post by JustinR » Sun Aug 27, 2017 8:52 pm

avalpert wrote:
Sun Aug 27, 2017 6:14 pm
JustinR wrote:
Sun Aug 27, 2017 4:45 pm
JoMoney wrote:
Sun Aug 27, 2017 12:01 pm
Over the past 3,5,10, and 15 year period(s) DFSVX which is typically pointed to as the premier 'Small-Cap Value' fund has had a lower Sharpe Ratio indicating higher risk and lower risk-adjusted returns using the traditional measurements
Image

The whole idea of factor investing is based on a premise that "small" and "value" are unique risk/return "factors" that are not properly measured with the traditional methods, and any difference in returns is a "risk premium" for taking ADDITIONAL RISK that has different skew and kurtosis in the way the risk/returns are distributed different from 'market beta'.
For the backtesting I used VTSMX (tsm) and VISVX (scv). For the last 15 years I get this:
When I plug those two into portfoliovisualizer for the same period the returns are the same as yours but the stdev are 14.73 and 18.57 respectively with VTSMX being lower.

Maybe you are making a mistake in your calculations?
I'm not sure. I'm using Simba's backtesting spreadsheet offered here: https://www.bogleheads.org/wiki/Simba%2 ... preadsheet

I don't know what the difference is between these two tools. Does PortfolioVisualizer offer downside risk metrics (Sortino Ratio)?

avalpert
Posts: 6193
Joined: Sat Mar 22, 2008 4:58 pm

Re: Small-Cap Value has higher returns than Total Stock Market, with lower risk

Post by avalpert » Sun Aug 27, 2017 8:58 pm

JustinR wrote:
Sun Aug 27, 2017 8:52 pm
avalpert wrote:
Sun Aug 27, 2017 6:14 pm
JustinR wrote:
Sun Aug 27, 2017 4:45 pm
JoMoney wrote:
Sun Aug 27, 2017 12:01 pm
Over the past 3,5,10, and 15 year period(s) DFSVX which is typically pointed to as the premier 'Small-Cap Value' fund has had a lower Sharpe Ratio indicating higher risk and lower risk-adjusted returns using the traditional measurements
Image

The whole idea of factor investing is based on a premise that "small" and "value" are unique risk/return "factors" that are not properly measured with the traditional methods, and any difference in returns is a "risk premium" for taking ADDITIONAL RISK that has different skew and kurtosis in the way the risk/returns are distributed different from 'market beta'.
For the backtesting I used VTSMX (tsm) and VISVX (scv). For the last 15 years I get this:
When I plug those two into portfoliovisualizer for the same period the returns are the same as yours but the stdev are 14.73 and 18.57 respectively with VTSMX being lower.

Maybe you are making a mistake in your calculations?
I'm not sure. I'm using Simba's backtesting spreadsheet offered here: https://www.bogleheads.org/wiki/Simba%2 ... preadsheet

I don't know what the difference is between these two tools. Does PortfolioVisualizer offer downside risk metrics (Sortino Ratio)?
Yes, the difference was much smaller - .67 to .75 and I can flip very easily with different time periods.

JustinR
Posts: 406
Joined: Tue Apr 27, 2010 11:43 pm

Re: Small-Cap Value has had higher returns than Total Stock Market, with lower risk

Post by JustinR » Sun Aug 27, 2017 9:09 pm

siamond wrote:
Sun Aug 27, 2017 7:47 pm
JustinR wrote:
Sat Aug 26, 2017 11:12 pm
Is there something inherently better about TSM that I should keep my stocks in there and not move them to SCV? Currently I'm half TSM and half SCV for my domestic allocation.
Hey Justin. I'm glad to see somebody making such good use of the Simba spreadsheet, this is cool. Your observations about the past are perfectly correct. I think this should provide positive reinforcement for your (already very aggressive) tilt towards SCV (50% TSM, 50% SCV). I do NOT think this should lead you to a more extreme allocation though. As multiple posters tried to explain, the SCV track record has been quite remarkable and yet went all over the place with great decades, poor decades, sideway decades, it's kind of hard to see a clear pattern in there. It's just a bad idea to put all your eggs in one (not so large, and heavily used) basket... You might actually consider diversifying your tilt between SCV and MCV (I do, and this makes me feel much more comfortable).

Let me suggest you draw a telltale chart, using 100% SCV as P1, 100% MCV as P2, and 100% TSM as benchmark. Start it early in history (e.g. 1930), then later (eg. 1970), and take your time to think about the trajectory, the time frames involved (decades!) and how you might have reacted without the benefit of hindsight. I don't know what the future will be made of, but in the past, it would have taken nerves of titanium to deal with SCV ups and downs relative to TSM.

As to a few questions that were asked in this thread:
- Justin didn't say that SCV displayed lower volatility (regular standard deviation), and this was NOT the case. He honed on other risk metrics (e.g. downside SD, Sharpe ratio, etc) and rightfully observed that SCV displayed cool properties in this respect in the past few decades (e.g. 1970+).
- The data sources are the ones used in the Simba backtesting spreadsheet. Check the spreadsheet 'Data_Sources' tab and you'll find the answer.
- Portfolio Visualizer, when it switched to the use of monthly returns, sidetracked a bit from Simba in terms of data sources. This explains the discrepancies (think of it this way: there are multiple indices for SCV, they all displayed a fascinating premium in the past, but the exact data series are all somewhat different, depending on the exact definition of a given index).
Thank you siamond so much for keeping Simba's spreadsheet updated! This is my first time using it and I found it much more simple to use than PortfolioVisualizer and other tools.

I agree. I think I'll stick to my 50% TSM, 50% SCV for now.

How were you recommending I work in MCV into my porfolio? By splitting it three ways like: 33% TSM, 33% MCV, 33% SCV? Or splitting my current SCV allocation so it's 50% TSM, 25% MCV, 25% SCV?

The telltale chart from 1970-2016:


Image


The valleys on SCV and MCV are scary, but looking at the left graph, SCV still shows that it has the highest return for the least downside risk. It's just weird how that is the case.


siamond wrote:
Sun Aug 27, 2017 7:57 pm
willthrill81 wrote:
Sun Aug 27, 2017 6:55 pm
I don't find the Sharpe or Sortino ratios particularly useful from a behavioral standpoint. I personally think that maximum drawdown is the single best measure of risk because it's what generally causes people to derail their investment plans. For instance, if you could handle the maximum drawdown, paying attention to the actual dollar amount you would have lost, in periods like 2000-2002 and 2008-2009, then you're probably good to go.
I would tend to agree with you. Drives me crazy to see so many people equating 'risk' with standard deviation, and worse, putting std-deviation and returns at the same level of importance while computing a ratio like Sharpe. Did you take a close look to the Ulcer Index/Ratio? This is quite an interesting idea, a creative variation on the idea of max drawdown (which has the drawback of being a single data point in history). Check this link.
Interesting, I'm just learning about max drawdown and ulcer index. Looking at the same example (1970-2016), MCV has higher return and lower numbers for both:

Image
Last edited by JustinR on Sun Aug 27, 2017 9:28 pm, edited 1 time in total.

User avatar
willthrill81
Posts: 2263
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Small-Cap Value has higher returns than Total Stock Market, with lower risk

Post by willthrill81 » Sun Aug 27, 2017 9:20 pm

siamond wrote:
Sun Aug 27, 2017 7:57 pm
willthrill81 wrote:
Sun Aug 27, 2017 6:55 pm
I don't find the Sharpe or Sortino ratios particularly useful from a behavioral standpoint. I personally think that maximum drawdown is the single best measure of risk because it's what generally causes people to derail their investment plans. For instance, if you could handle the maximum drawdown, paying attention to the actual dollar amount you would have lost, in periods like 2000-2002 and 2008-2009, then you're probably good to go.
I would tend to agree with you. Drives me crazy to see so many people equating 'risk' with standard deviation, and worse, putting std-deviation and returns at the same level of importance while computing a ratio like Sharpe. Did you take a close look to the Ulcer Index/Ratio? This is quite an interesting idea, a creative variation on the idea of max drawdown (which has the drawback of being a single data point in history). Check this link.
Thanks! I've heard of the UI but never read up on it. It's certainly intriguing and seems superior to std. dev. as a useful measure of volatility. Is there a website that calculates it for different tickers?
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

User avatar
siamond
Posts: 3429
Joined: Mon May 28, 2012 5:50 am

Re: Small-Cap Value has had higher returns than Total Stock Market, with lower risk

Post by siamond » Sun Aug 27, 2017 9:38 pm

JustinR wrote:
Sun Aug 27, 2017 9:09 pm
How were you recommending I work in MCV into my porfolio? By splitting it three ways like: 33% TSM, 33% MCV, 33% SCV? Or splitting my current SCV allocation so it's 50% TSM, 25% MCV, 25% SCV?
I would suggest you consider keeping the same amount of tilt, and split equally MCV/SCV, i.e. 50/25/25. If only to keep better continuity with what you were doing so far. Don't change your AA on a dime though. Take a few months to reflect on it, make sure whatever change you might do is truly a stable idea that you believe in. If there is one thing in the past that was required to benefit from the value premium, it was to stick to the plan for a long, long, time.
JustinR wrote:
Sun Aug 27, 2017 9:09 pm
The telltale chart from 1970-2016 [...] The valleys on SCV and MCV are scary, but looking at the left graph, SCV still shows that it has the highest return for the least downside risk. It's just weird how that is the case
Yes, it's hard to explain. Although not always true in older time periods. Anyhoo, this wasn't quite the point I wanted to make. Take a look at 50 years at a time. You'll see that either MCV or SCV made a difference due to very occasional spurts of growth, while not making much of a difference the rest of the time. And every time, it was different. I don't know if the future will rhyme with the past (although I do tilt, so I hope so!), but my guess is that the rhyme is likely to be extremely loose and not quite anything we've seen in the past. And therefore, I'd rather hedge my bets. It is just unwise to make one concentrated bet.

User avatar
siamond
Posts: 3429
Joined: Mon May 28, 2012 5:50 am

Re: Small-Cap Value has higher returns than Total Stock Market, with lower risk

Post by siamond » Sun Aug 27, 2017 9:41 pm

willthrill81 wrote:
Sun Aug 27, 2017 9:20 pm
Thanks! I've heard of the UI but never read up on it. It's certainly intriguing and seems superior to std. dev. as a useful measure of volatility. Is there a website that calculates it for different tickers?
Not that I know of. I implemented the math in the Simba spreadsheet, this isn't terribly complicated. Shouldn't be hard to generalize to an arbitrary data series downloaded from Morningstar or Yahoo Finance or else.

User avatar
FIREchief
Posts: 1363
Joined: Fri Aug 19, 2016 6:40 pm

Re: Small-Cap Value has higher returns than Total Stock Market, with lower risk

Post by FIREchief » Sun Aug 27, 2017 9:53 pm

dbr wrote:
Sun Aug 27, 2017 2:59 pm
FIREchief wrote:
Sun Aug 27, 2017 2:49 pm
JustinR wrote:
Sat Aug 26, 2017 11:12 pm
I apologize that I made a similar thread about Mid-Cap Value a few hours ago, but I just backtested Small-Cap and was wondering if people had opinions on going 100% SCV instead of TSM.

Using Simba's backtesting spreadsheet, I plugged in 80%TSM vs 80%SCV (with the rest in Total Bond).
Lost me at "backtesting."
"Backtesting" means that you look at the performance of some holding as it
actually performed in some period in past. Backtesting is deceptively attractive because the results are fact, similar to saying that the Cowboys beat the Packers by 30-16 on Oct 16, 2106. If that sounds like too much of a one-off the data could be that the Cowboys went 13-3 and the Packers 10-6 last year. If you want something more durable than that you could cite the result that the all time record between the two teams is Packers over Cowboys 18-17-0.

Now if the question is which city should you and your brother split to to have NFL bragging rights over the other for the next twenty years, which of this data is useful to decide this? Would you make the same analysis if it were a question of getting NFL bragging rights by moving to San Diego?
Thanks. Nice comparison by the way.... I didn't mean that the term "backtesting" confused me. I meant that reliance upon backtesting to justify some nirvana type "higher returns with lower risks" approach left me "lost" (i.e. unconvinced). 8-)
I am not a lawyer, accountant or financial advisor. Any advice or suggestions that I may provide shall be considered for entertainment purposes only.

User avatar
privatefarmer
Posts: 276
Joined: Mon Sep 08, 2014 2:45 pm

Re: Small-Cap Value has higher returns than Total Stock Market, with lower risk

Post by privatefarmer » Mon Aug 28, 2017 1:32 am

Robert T wrote:
Sun Aug 27, 2017 2:55 pm

Annualized returns since then (from above link):

9.3% = Vanguard 500 Index [VFINX]
10.7% = DFA US Small Cap [DFSTX]
11.8% = DFA US Small Cap Value [DFSVX]

Obviously no guarantees.

Robert
.

I apologize if I am regurgitating what has already been said many times on this forum, but we know that if one sliver of the market has outperformed (SCV) then another equally-sized sliver of the TSM must have underperformed by the same amount. In this case, maybe LCG or SCG underperformed the total market by 2.5%/year? The point is, does anyone believe that a LCG investor (ie momentum investor) believes that LCG will underperform going forward? of course not. They have their reasons for believing that LCG will outperform just as many here believe SCV will outperform.

For one to win another must lose. I have decided not to play that game. Owning the total market is the only solution to not betting against other active investors. Smart beta is a form of active management disguised by "index funds", but it is active management. And just as active managers have all come and gone in the past, their strategies that they thought would work 30 years ago are now forgotten about, so will "smart beta". I am a young investor but even I have seen "smart beta" transform from small/value to now momentum, quality, low vol... if there was a strategy that had worked and WILL work going forward it would remain popular and everyone would do it - thus destroying it. It's impossible to have a winning strategy that is publicly known continue for any long period of time as once it's publically known massive amounts of funds flow into it making it more expensive and a bubble waiting to "pop".

sorry to ramble, I am sure all this has been already stated in the thread.

WanderingDoc
Posts: 288
Joined: Sat Aug 05, 2017 8:21 pm

Re: Small-Cap Value has higher returns than Total Stock Market, with lower risk

Post by WanderingDoc » Mon Aug 28, 2017 3:17 am

willthrill81 wrote:
Sun Aug 27, 2017 10:04 am
Valuethinker wrote:
Sun Aug 27, 2017 7:26 am
The Small Cap Value effect is not consistent nor predictable.
Considering that SCV has beaten LC in terms of returns in every twenty year period save one in the historic record, I'd say that's pretty darned consistent, at least consistent enough for the typical investor to experience the higher returns several times during their investment horizon.

Of course someone will bring up the whole independent period argument, but still SCV's outperformance of TSM has not been limited to a just a few isolated periods. Even just in recent decades, SCV just came up short of TSM from 2007-2017, but it outperformed in '97-'07, '87-'97, and '77-'87 by several percent at least.
Don't worry. They plan on living to 450 years old! When I bring up sustained year over year appreciation of 9% and rent growth of 6% in Honolulu and San Francisco since the 1960s until present, 50+ years. They will have you believe that 50 years is too short to "draw any real conclusions". Cause, like Case-Shiller said so, bro! ;)
Meanwhile, I will just smile when my rents and property values double every 10 years, and I earn my entire down payment in profit after 2-4 years.
One day it suddenly dawned on me that I had won the real estate lottery. | I'm not looking to get rich quickly. I'm not looking to get rich slowly. I'm looking to get rich for sure.

Post Reply