How much tilting to SCV (rule of thumb)?

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
User avatar
vineviz
Posts: 6637
Joined: Tue May 15, 2018 1:55 pm

Re: How much tilting to SCV (rule of thumb)?

Post by vineviz » Thu Jun 25, 2020 2:24 pm

international001 wrote:
Thu Jun 25, 2020 2:00 pm
Uncorrelated wrote:
Tue Jun 23, 2020 8:17 am
Unless an investor is leverage constrained, the optimal tilt does not vary throughout time. If you believe that it is optimal to tilt 1/3rd of your portfolio to SCV, then you can expect that to be very close to optimal for portfolio's ranging from 100% equities to 10% equities. I don't know why Paul Merriman recommends a tilt based on age, but it sounds like a crude and simple attempt to approximate lifecycle investing.
Well, SCV has more return and more stdev, so it would seem normal to have more SCV if you are further from retirement. I just don't trust his exacts numbers.
The further you are from retirement, the more portfolio volatility you can probably (or should probably) accept.

Within that portfolio, however, there's no reason that the ratio of SCV to total stock market should decrease over time. If anything, as I mentioned earlier I think, the optimal amount of SCV tilt is probably highest right after retirement.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

User avatar
Uncorrelated
Posts: 727
Joined: Sun Oct 13, 2019 3:16 pm

Re: How much tilting to SCV (rule of thumb)?

Post by Uncorrelated » Thu Jun 25, 2020 2:28 pm

international001 wrote:
Thu Jun 25, 2020 1:43 pm
Uncorrelated wrote:
Tue Jun 23, 2020 8:17 am
Technically, the optimal tilt is approximately equal parts of the factors MKT, HmL and SmB (among others) because all factors have approximately equal sharpe ratio's. However, obtaining this tilt requires funds that do not exist and would be to expensive, if they existed. The next best theoretical option is to go 100% SVC, but that also has disadvantages because extreme tilts increase idiosyncratic (uncompensated) risk in the market factor in ways that are very hard to measure. To my knowledge no accurate models exist that allow investors to calculate which tilt is optimal.
I'm not familiar with sharpe ratios for factors. You consider betas separately, but you use the same stdev? That certainly wouldn't give you similar sharpes. Any link appreciated

I suggest tilting to a maximum of 1/3 market, 1/3 SVC and 1/3 value. As you can tell by the round numbers, this suggestion is not based on mathematical rigor, but is a best-effort estimate that sounds reasonable. If you are really leverage constrained, you might want an even higher tilt. If you're not so convinced, you probably want a lower tilt.
This is more exposure to value than to size, no?
See https://papers.ssrn.com/sol3/papers.cfm ... id=2287202, table 4 on page 7. The sharpe ratio (return divided by stddev) of each factor (including market beta) is approximately the same.

The size factor appears weaker than others. This is not apparent from the above research paper, but is documented in this paper: https://www.aqr.com/Insights/Research/J ... ize-Effect.

With a 1/3 split between mkt, small value and value, you'll have approximately equal exposure to size and value. Obtaining size exposure is just much easier than obtaining value exposure. The source above suggests that it would be better to get higher value exposure and lower size exposure, but since most of the value-y stocks are small cap stocks, it is pretty much impossible to obtain significant value exposure without obtaining small exposure.

User avatar
vineviz
Posts: 6637
Joined: Tue May 15, 2018 1:55 pm

Re: How much tilting to SCV (rule of thumb)?

Post by vineviz » Thu Jun 25, 2020 2:49 pm

international001 wrote:
Thu Jun 25, 2020 1:37 pm

I don't want to compare time periods separately. I want to compare what is the chance that in a time period of X years SCV underperforms LCB. My understanding is that the greater X, that chance is lower
The shorter the time period, the lower the probability of SCV outperformance. But the probability is >50% even for very short periods of time (i.e. a month or more), which is why the argument for not tilting based on time horizon is nonsensical: any investment horizon long enough to justify an allocation to stocks is long enough to justify a tilt to SCV.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

rkhusky
Posts: 9439
Joined: Thu Aug 18, 2011 8:09 pm

Re: How much tilting to SCV (rule of thumb)?

Post by rkhusky » Thu Jun 25, 2020 3:14 pm

vineviz wrote:
Thu Jun 25, 2020 2:49 pm
international001 wrote:
Thu Jun 25, 2020 1:37 pm

I don't want to compare time periods separately. I want to compare what is the chance that in a time period of X years SCV underperforms LCB. My understanding is that the greater X, that chance is lower
The shorter the time period, the lower the probability of SCV outperformance. But the probability is >50% even for very short periods of time (i.e. a month or more), which is why the argument for not tilting based on time horizon is nonsensical: any investment horizon long enough to justify an allocation to stocks is long enough to justify a tilt to SCV.
Assuming that past percentages hold in the future. No one knows the future probabilities.

User avatar
Uncorrelated
Posts: 727
Joined: Sun Oct 13, 2019 3:16 pm

Re: How much tilting to SCV (rule of thumb)?

Post by Uncorrelated » Thu Jun 25, 2020 4:23 pm

rkhusky wrote:
Thu Jun 25, 2020 3:14 pm
vineviz wrote:
Thu Jun 25, 2020 2:49 pm
international001 wrote:
Thu Jun 25, 2020 1:37 pm

I don't want to compare time periods separately. I want to compare what is the chance that in a time period of X years SCV underperforms LCB. My understanding is that the greater X, that chance is lower
The shorter the time period, the lower the probability of SCV outperformance. But the probability is >50% even for very short periods of time (i.e. a month or more), which is why the argument for not tilting based on time horizon is nonsensical: any investment horizon long enough to justify an allocation to stocks is long enough to justify a tilt to SCV.
Assuming that past percentages hold in the future. No one knows the future probabilities.
The same holds true for market beta itself. It just so happens that the confidence intervals for market beta and factors are approximately equal. Uncertainty about the underlying factor return is a pretty poor reason to avoid a tilt.

User avatar
Steve Reading
Posts: 1856
Joined: Fri Nov 16, 2018 10:20 pm

Re: How much tilting to SCV (rule of thumb)?

Post by Steve Reading » Thu Jun 25, 2020 4:47 pm

Uncorrelated wrote:
Thu Jun 25, 2020 4:23 pm
Uncertainty about the underlying factor return is a pretty poor reason to avoid a tilt.
I disagree completely and in fact suggest that uncertainty about the underlying returns of any investment, factor, sector, etc is an entirely appropriate and relevant consideration when thinking about whether to avoid it or not.

rkhusky
Posts: 9439
Joined: Thu Aug 18, 2011 8:09 pm

Re: How much tilting to SCV (rule of thumb)?

Post by rkhusky » Thu Jun 25, 2020 5:36 pm

Uncorrelated wrote:
Thu Jun 25, 2020 4:23 pm
The same holds true for market beta itself. It just so happens that the confidence intervals for market beta and factors are approximately equal. Uncertainty about the underlying factor return is a pretty poor reason to avoid a tilt.
The Market-RiskFree return is substantially greater than SmB or HmL. Uncertainty in the betas is only one piece. There is also the uncertainty in the returns.

Topic Author
international001
Posts: 1463
Joined: Thu Feb 15, 2018 7:31 pm

Re: How much tilting to SCV (rule of thumb)?

Post by international001 » Thu Jun 25, 2020 7:23 pm

vineviz wrote:
Thu Jun 25, 2020 2:24 pm


The further you are from retirement, the more portfolio volatility you can probably (or should probably) accept.

Within that portfolio, however, there's no reason that the ratio of SCV to total stock market should decrease over time. If anything, as I mentioned earlier I think, the optimal amount of SCV tilt is probably highest right after retirement.
??
Adding SCV will increase stdev.

Investment A: TSM (stdev 15%)
Investment B: SCV (stdev 20%)

To reach the same stdev after N years you could hold B for NB years or A for NA years (NB/NA = 20**2/15**2), no?

User avatar
vineviz
Posts: 6637
Joined: Tue May 15, 2018 1:55 pm

Re: How much tilting to SCV (rule of thumb)?

Post by vineviz » Thu Jun 25, 2020 7:47 pm

international001 wrote:
Thu Jun 25, 2020 7:23 pm
vineviz wrote:
Thu Jun 25, 2020 2:24 pm


The further you are from retirement, the more portfolio volatility you can probably (or should probably) accept.

Within that portfolio, however, there's no reason that the ratio of SCV to total stock market should decrease over time. If anything, as I mentioned earlier I think, the optimal amount of SCV tilt is probably highest right after retirement.
??
Adding SCV will increase stdev.

Investment A: TSM (stdev 15%)
Investment B: SCV (stdev 20%)

To reach the same stdev after N years you could hold B for NB years or A for NA years (NB/NA = 20**2/15**2), no?
Or simply adjust your stock/bond ratio to get the portfolio volatility you want.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

User avatar
Uncorrelated
Posts: 727
Joined: Sun Oct 13, 2019 3:16 pm

Re: How much tilting to SCV (rule of thumb)?

Post by Uncorrelated » Fri Jun 26, 2020 8:11 am

Steve Reading wrote:
Thu Jun 25, 2020 4:47 pm
Uncorrelated wrote:
Thu Jun 25, 2020 4:23 pm
Uncertainty about the underlying factor return is a pretty poor reason to avoid a tilt.
I disagree completely and in fact suggest that uncertainty about the underlying returns of any investment, factor, sector, etc is an entirely appropriate and relevant consideration when thinking about whether to avoid it or not.
I should clarify. Statistically, there is no real reason to have more faith in the market premium than in value. I understand that the underlying behaviorally/risk based reasoning for the market factor are much stronger than for value, but I find it hard to argue that that makes sense from a statistical viewpoint.

There are other concerns, such as questions about reason the factors exist in the first place, concern of over-fitting and data-mining, difficulty to implement a factor tilt, unknown amounts of idiosyncratic risk in a factor portfolio, and cost reasons that can be used convincingly against a factor tilt. But as far as concerns go about the returns of the factors themselves, I don't see any convincing reason to see factor returns as weaker than market beta.
rkhusky wrote:
Thu Jun 25, 2020 5:36 pm
Uncorrelated wrote:
Thu Jun 25, 2020 4:23 pm
The same holds true for market beta itself. It just so happens that the confidence intervals for market beta and factors are approximately equal. Uncertainty about the underlying factor return is a pretty poor reason to avoid a tilt.
The Market-RiskFree return is substantially greater than SmB or HmL. Uncertainty in the betas is only one piece. There is also the uncertainty in the returns.
The t-stat is a measure of statistical significance (probability that underlying return != 0). These are approximately the same for all factors including beta. The absolute magnitude doesn't matter much.

rkhusky
Posts: 9439
Joined: Thu Aug 18, 2011 8:09 pm

Re: How much tilting to SCV (rule of thumb)?

Post by rkhusky » Fri Jun 26, 2020 10:15 am

Uncorrelated wrote:
Fri Jun 26, 2020 8:11 am
The t-stat is a measure of statistical significance (probability that underlying return != 0). These are approximately the same for all factors including beta. The absolute magnitude doesn't matter much.
The stock market doesn't obey any known statistics, so I'm not sure of the validity of statistical tests.

As I said before, there are the betas for Market, SmB, HmL, and also the actual Market, SmB, HmL returns. One needs the uncertainty in both the betas and the uncertainty in the returns to compute the uncertainty in the contribution that each component makes to the overall return.

User avatar
vineviz
Posts: 6637
Joined: Tue May 15, 2018 1:55 pm

Re: How much tilting to SCV (rule of thumb)?

Post by vineviz » Fri Jun 26, 2020 10:23 am

Uncorrelated wrote:
Fri Jun 26, 2020 8:11 am
rkhusky wrote:
Thu Jun 25, 2020 5:36 pm

The Market-RiskFree return is substantially greater than SmB or HmL. Uncertainty in the betas is only one piece. There is also the uncertainty in the returns.
The t-stat is a measure of statistical significance (probability that underlying return != 0). These are approximately the same for all factors including beta. The absolute magnitude doesn't matter much.
It's also worth remembering that the returns of the market premium factor (Rmf) are almost universally overstated due to misspecification of the risk-free rate.

Most estimates for Rmf use 30-day Treasury bills as the assumed risk-free rate, which is only valid for investors with a 30-day investment horizon. The average equity investor does NOT have a 30-day investment horizon. David Blitz recently authored a paper which concluded that the market-implied risk-free rate corresponds to intermediate-term Treasuries.
In empirical tests of the CAPM, the theoretical risk-free asset is typically assumed to be 1-month Treasury bills. This paper examines the implications of a mis-specified risk-free asset, i.e. the possibility that the ‘true’ risk-free asset is a longer-maturity Treasury bond. A simple theoretical derivation leads to the testable prediction that low-beta (high-beta) stocks should then exhibit positive (negative) bond betas. We find strong empirical confirmation for these predictions. The market-implied risk-free asset can be pinpointed at medium-term (5-year) bonds. Concrete implications of this finding are a lower equity risk premium and a less steep security market line.
link

This has important implications for discussions about both the relative size and confidence intervals of the various factor returns.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

User avatar
Steve Reading
Posts: 1856
Joined: Fri Nov 16, 2018 10:20 pm

Re: How much tilting to SCV (rule of thumb)?

Post by Steve Reading » Fri Jun 26, 2020 10:50 am

rkhusky wrote:
Fri Jun 26, 2020 10:15 am
As I said before, there are the betas for Market, SmB, HmL, and also the actual Market, SmB, HmL returns. One needs the uncertainty in both the betas and the uncertainty in the returns to compute the uncertainty in the contribution that each component makes to the overall return.
I would not be as concerned about the "betas" that you talk about there. Funds that track S&P 600 Value, CRSP SCV, etc typically have t-statistics on their SmB and HmL of several orders of magnitude below statistical significance. Odds are incredibly high that these funds will continue to load on these factors. And their methodologies should logically load on these factors too. The size of that beta might change in the future, but probably not enough to be concerning. Funds of diversified stocks that have shown market betas of 1, will probably have very similar market beta of 1 in the future, etc etc.

As for the uncertainty about the future premia of factors, however, I think that's an appropriate concern and probably one of the biggest. And that includes uncertainty about premia like the market, term, etc too.

JamesDean44
Posts: 149
Joined: Sat Jan 13, 2018 7:36 pm

Re: How much tilting to SCV (rule of thumb)?

Post by JamesDean44 » Fri Jun 26, 2020 11:22 am

vineviz wrote:
Fri Jun 26, 2020 10:23 am
Uncorrelated wrote:
Fri Jun 26, 2020 8:11 am
rkhusky wrote:
Thu Jun 25, 2020 5:36 pm

The Market-RiskFree return is substantially greater than SmB or HmL. Uncertainty in the betas is only one piece. There is also the uncertainty in the returns.
The t-stat is a measure of statistical significance (probability that underlying return != 0). These are approximately the same for all factors including beta. The absolute magnitude doesn't matter much.
It's also worth remembering that the returns of the market premium factor (Rmf) are almost universally overstated due to misspecification of the risk-free rate.

Most estimates for Rmf use 30-day Treasury bills as the assumed risk-free rate, which is only valid for investors with a 30-day investment horizon. The average equity investor does NOT have a 30-day investment horizon. David Blitz recently authored a paper which concluded that the market-implied risk-free rate corresponds to intermediate-term Treasuries.
In empirical tests of the CAPM, the theoretical risk-free asset is typically assumed to be 1-month Treasury bills. This paper examines the implications of a mis-specified risk-free asset, i.e. the possibility that the ‘true’ risk-free asset is a longer-maturity Treasury bond. A simple theoretical derivation leads to the testable prediction that low-beta (high-beta) stocks should then exhibit positive (negative) bond betas. We find strong empirical confirmation for these predictions. The market-implied risk-free asset can be pinpointed at medium-term (5-year) bonds. Concrete implications of this finding are a lower equity risk premium and a less steep security market line.
link

This has important implications for discussions about both the relative size and confidence intervals of the various factor returns.
Interesting.

User avatar
nisiprius
Advisory Board
Posts: 41003
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: How much tilting to SCV (rule of thumb)?

Post by nisiprius » Fri Jun 26, 2020 12:26 pm

Here is something I did for my own interest. It is what it is. It refers to a specific period of time. It's past performance. In this particular thread, I don't care to discuss the use of the arithmetic mean (appropriate here but I don't want to argue it), or the possibility of a rebalancing bonus (annual rebalancing is assumed), or the use of standard deviation as a measure of risk.

Because I personally am retired and very conservatively invested, I chose 40/60 as my starting point rather than 60/40 but will mention the 60/40 results afterwards.

It is based on real-world investments in actual mutual funds and thus could have been easily implemented by an ordinary retail investor using real dollars. The funds chosen for consideration were an initial portfolio of 40% Vanguard Total Stock, 60% Vanguard Total Bond. I explored the results of replacing some of Total Stock with the DFA US Small-cap Value Portfolio, DFSVX. I chose this fund because it is old, it is from a firm that has Fama and French among their directors, and because factor mavens in the forum have praised it. The time period covered is 1994-2019 inclusive, limited to whole calendar years and by the inception date of DFSVX.

1994-2019 includes a couple of business cycles. It includes about 26/28ths of the time since the "discovery" of the Fama-French three-factor novel. It includes one period (2000-2003) of stellar small-cap value performance.

The red dot is Total Stock, the green dot is Total Bond, the yellow dot is a 40/60 mix. The average (arithmetic mean of annual returns) is µ = 7.60%, and the standard deviation is σ = 7.73%.

Image

Since 40/60 represents the amount of risk I wish to take, this is how I formulated the question. For a fixed "cost" of 7.73% in standard deviation, how much could I have increased return by including DFSVX in the portfolio without increasing the risk, and what mixture of Total Stock and DVSVX would have given the greatest improvement?

In this chart, the aqua dot at right is DFSVX, and the maroon curve shows all of the possible µ and σ combinations obtained by mixtures of Total Stock and DFSVX. Interestingly, both Total Stock and DFSVX had almost identical Sharpe ratios, about 0.5, so DFSVX offered higher return at the cost of higher risk, and virtually identical risk-adjusted return.

However, because Total Stock and DFSVX had an imperfect correlation of ρ = 0.74. This means the efficient frontier of the two is not a straight line, but bulges upward.

Image

Going back to the yellow line, then, we can choose to anchor the right end, not at the red Total Stock dot, but at some mixture of Total Stock and DFSVX. That will raise the right end of the yellow curve higher. Since DFSVX has higher risk than Total Stock, a mixture of 60% bonds with 40% combination-of-stocks will have higher risk; in order to bring hold risk constant, we need to reduce the total stock allocation, use a little less than 40% of a slightly-riskier combination of stocks.

Image

It turns out that:
  • The optimum mix of Total Stock and DFSVX, for this particular kind of optimum over this particular past time period, would have been 45% Total Stock, 55% DFSVX, i.e. more than half of the stock allocation is invested in small-cap value.
  • To hold total portfolio risk constant, it is theoretically necessary to compensate by cutting the overall stock allocation, but in this particular case the reduction need turns out to be microscopic: 39.4% stocks, 60.6% bonds.
  • The return of the portfolio was improved by 0.29%, 29 basis points, from 7.60% to 7.89%.
If we perform the same analysis but start with a 60/40 allocation, the return is improved by 0.43%, 43 basis points, from 8.84% to 9.27%. The optimum mix within the stock allocation was 56.5% to small-cap value. And in order to keep the risk constant at 7.73% it was necessary to raise the bond allocation to 41.1% and cut the stock allocation to 58.9%.

So, that's what I did and those are the numbers I got. They suggest to me that if all the parameters were stable going forward--which I don't believe!--the theoretical optimum allocation to small-cap value would be quite high... but the amount of improvement is not worth having unless it is all-but-guaranteed.
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.

garlandwhizzer
Posts: 2846
Joined: Fri Aug 06, 2010 3:42 pm

Re: How much tilting to SCV (rule of thumb)?

Post by garlandwhizzer » Fri Jun 26, 2020 1:23 pm

vineviz wrote:

The shorter the time period, the lower the probability of SCV outperformance. But the probability is >50% even for very short periods of time (i.e. a month or more), which is why the argument for not tilting based on time horizon is nonsensical: any investment horizon long enough to justify an allocation to stocks is long enough to justify a tilt to SCV.
As to the promised SCV outperformance, let's take a look at just how robust that gravy train of excess returns is in practice. Let's don't look at the numerous SCV funds that fell by the wayside and didn't live up to the hype. Instead, let's look at DFSVX, touted by factor mavens to be state of the art SCV harvesting. DFA who creates this fund has on its payroll the 2 Nobel Prize winners who originally described the small and value factors, as well as Merton of MIT who also won the Nobel Prize. That's right DFA employs 3 Nobel Prize winners and DFSVX is their flagship fund. So let's look at how this fund has performed since it's inception more than 27 years ago, March 1993 through May of 2020. The following results are from Portfolio Visualizer comparing DFSVX to VFINX (LCB index with massive negative loadings to size and no emphasis on value at all) and also to VTMSX, TSM, a one factor fund merely ultra cheap beta.

The results show that over this 27 yr. 2 month period (certainly that qualifies as long term), DFSVX did in fact outperform both of Vanguard's offerings but only by a hair. It beat the S&P 500 fund by 0.16%/yr in compound annual return but it did so with vastly greater risk as measured by maximal drawdown and vastly greater volatility as measured by standard deviation. As a result the risk per unit of reward for the S&P 500 Fund as measured by the Sharpe ratio was much better than for this state of the art SCV fund over its entire lifetime. The story is the same for the TSM fund which underperformed by only 0.17%/yr. but had did so with much less risk and volatility and consequently had a much better return per unit of risk. So after tolerating the increased risk and volatility for 27+years the big SCV payoff was less than 1/5 of 1% per year.

I believe that vineviz quote overstates the case for SCV.
the argument for not tilting based on time horizon is nonsensical: any investment horizon long enough to justify an allocation to stocks is long enough to justify a tilt to SCV.
In view of these backtesting results for DFSVX the necessity for tilting to SCV in all portfolios is what seems nonsensical to me.

https://www.portfoliovisualizer.com/bac ... ion3_3=100

Garland Whizzer

billy269
Posts: 117
Joined: Mon Nov 18, 2013 2:50 pm

Re: How much tilting to SCV (rule of thumb)?

Post by billy269 » Fri Jun 26, 2020 4:26 pm

I have about a third of my domestic allocation in SCV (2/3 in VTI total market) and plan to leave it in for 30 years. We'll see if that was a smart decision in 2050. Started this three years ago. Hasn't been a good start.

User avatar
Taylor Larimore
Advisory Board
Posts: 29507
Joined: Tue Feb 27, 2007 8:09 pm
Location: Miami FL

Re: How much tilting to SCV (rule of thumb)?

Post by Taylor Larimore » Fri Jun 26, 2020 4:42 pm

Garland Whizzer:

I am reluctant to put much faith in past performance, nevertheless I am wondering if your post above, showing almost identical returns for VFINX (Vanguard S&P 500 Index Fund) and DFSVX (Dimensional Small-Cap Value Fund), includes the advisor fee required for individual DFSVX funds? What about tax-efficiency?

Thank you and best wishes.
Taylor
Jack Bogle's Words of Wisdom: "No analysis of the past, no matter how painstaking, assures future superiority."
"Simplicity is the master key to financial success." -- Jack Bogle

rkhusky
Posts: 9439
Joined: Thu Aug 18, 2011 8:09 pm

Re: How much tilting to SCV (rule of thumb)?

Post by rkhusky » Fri Jun 26, 2020 5:52 pm

Steve Reading wrote:
Fri Jun 26, 2020 10:50 am
rkhusky wrote:
Fri Jun 26, 2020 10:15 am
As I said before, there are the betas for Market, SmB, HmL, and also the actual Market, SmB, HmL returns. One needs the uncertainty in both the betas and the uncertainty in the returns to compute the uncertainty in the contribution that each component makes to the overall return.
I would not be as concerned about the "betas" that you talk about there. Funds that track S&P 600 Value, CRSP SCV, etc typically have t-statistics on their SmB and HmL of several orders of magnitude below statistical significance. Odds are incredibly high that these funds will continue to load on these factors. And their methodologies should logically load on these factors too. The size of that beta might change in the future, but probably not enough to be concerning. Funds of diversified stocks that have shown market betas of 1, will probably have very similar market beta of 1 in the future, etc etc.

As for the uncertainty about the future premia of factors, however, I think that's an appropriate concern and probably one of the biggest. And that includes uncertainty about premia like the market, term, etc too.
From PortfolioVisualizer, the range for the betas (loadings) for DFSVX under default 3-factor settings looks to be about Mkt-RF: 1.0 +- 0.1, SmB: 0.85 +- 0.15, HmL: 0.55 +- 0.15 (My account expired so I can't download the data and calculate the SD).

From Ken French's data library, the average monthly factor returns (premiums) for 1993-2020 was Mkt-RF: 0.7 +- 4.4, SmB: 0.1 +- 3.2, HmL: 0.0 +- 3.1. The average yearly factor returns (premiums) for 1993-2019 was Mkt-RF: 9 +- 18, SmB: 1 +- 10, HmL: 2 +- 15

User avatar
Steve Reading
Posts: 1856
Joined: Fri Nov 16, 2018 10:20 pm

Re: How much tilting to SCV (rule of thumb)?

Post by Steve Reading » Fri Jun 26, 2020 6:12 pm

rkhusky wrote:
Fri Jun 26, 2020 5:52 pm
From PortfolioVisualizer, the range for the betas (loadings) for DFSVX under default 3-factor settings looks to be about Mkt-RF: 1.0 +- 0.1, SmB: 0.85 +- 0.15, HmL: 0.55 +- 0.15 (My account expired so I can't download the data and calculate the SD).
I don't know what you're referring to here. Are you talking about the 36 month rolling regressions? Over a period of 3 years, the factor loadings of the fund could vary significantly. The methodology of VBR, VIOV, DFSVX isn't identical to FF so it is perfectly normal to see, over short periods, a loading of FF that varies, sometimes significantly.

What I'm saying is that whatever loading a fund obtains over a long period of time (decades) most likely is very close to whatever loading it will have for the following long period of time. I assume we're all investing long term here. Ex: I expect the loading of DFSVX from 2010 to 2020 to be very similar to that from 1993-2010. The latter was 1.0, 0.8 and 0.67 in order. The former was 1.08, 0.85 and 0.55. Not perfect by any means, but probably close enough that it just doesn't matter much. You can see that in that the 60-month rolling regression is much smoother.

The future of factor premiums, OTOH, is the big elephant in the room that I think is by far more important. A SCV fund will almost certainly load on SCV in the future. But whether SCV will actually outperform is, in my opinion, far more uncertain comparatively.

User avatar
Uncorrelated
Posts: 727
Joined: Sun Oct 13, 2019 3:16 pm

Re: How much tilting to SCV (rule of thumb)?

Post by Uncorrelated » Sat Jun 27, 2020 8:32 am

garlandwhizzer wrote:
Fri Jun 26, 2020 1:23 pm
vineviz wrote:

The shorter the time period, the lower the probability of SCV outperformance. But the probability is >50% even for very short periods of time (i.e. a month or more), which is why the argument for not tilting based on time horizon is nonsensical: any investment horizon long enough to justify an allocation to stocks is long enough to justify a tilt to SCV.
As to the promised SCV outperformance, let's take a look at just how robust that gravy train of excess returns is in practice. Let's don't look at the numerous SCV funds that fell by the wayside and didn't live up to the hype. Instead, let's look at DFSVX, touted by factor mavens to be state of the art SCV harvesting. DFA who creates this fund has on its payroll the 2 Nobel Prize winners who originally described the small and value factors, as well as Merton of MIT who also won the Nobel Prize. That's right DFA employs 3 Nobel Prize winners and DFSVX is their flagship fund. So let's look at how this fund has performed since it's inception more than 27 years ago, March 1993 through May of 2020. The following results are from Portfolio Visualizer comparing DFSVX to VFINX (LCB index with massive negative loadings to size and no emphasis on value at all) and also to VTMSX, TSM, a one factor fund merely ultra cheap beta.

The results show that over this 27 yr. 2 month period (certainly that qualifies as long term), DFSVX did in fact outperform both of Vanguard's offerings but only by a hair. It beat the S&P 500 fund by 0.16%/yr in compound annual return but it did so with vastly greater risk as measured by maximal drawdown and vastly greater volatility as measured by standard deviation. As a result the risk per unit of reward for the S&P 500 Fund as measured by the Sharpe ratio was much better than for this state of the art SCV fund over its entire lifetime. The story is the same for the TSM fund which underperformed by only 0.17%/yr. but had did so with much less risk and volatility and consequently had a much better return per unit of risk. So after tolerating the increased risk and volatility for 27+years the big SCV payoff was less than 1/5 of 1% per year.

I believe that vineviz quote overstates the case for SCV.
the argument for not tilting based on time horizon is nonsensical: any investment horizon long enough to justify an allocation to stocks is long enough to justify a tilt to SCV.
In view of these backtesting results for DFSVX the necessity for tilting to SCV in all portfolios is what seems nonsensical to me.

https://www.portfoliovisualizer.com/bac ... ion3_3=100

Garland Whizzer
This is not the correct way to measure fund-level outperformance. The correct way to measure fund-level outperformance is to either use the arithmetic average, or to run a factor regression and project those results on the historical factor premia.

nisiprius did the analysis the right way. It's not perfect because he optimizes for return at a specific volatility rather than for max utility, but his analysis proves that tilters with this specific fund in this specific time period have been rewarded with higher returns at the same risk (or any variation thereof). Of course that doesn't tell you whether that will continue in the future, but it's kind of ignorant to claim that tilters received slightly more return at vastly higher risk when the post above claims the exact opposite.

Please stop linking to portfolio visualizer and/or citing geometric averages and sharpe ratio's. Those are misleading figures. A higher geometric average or sharpe ratio does not imply higher utility for the individual investor. Poor performance for a 100% SCV portfolio does not imply that tilting a small amount also results in poor performance. If you want to show that tilting doesn't work, you need to show that the efficient frontier with and without factor funds is exactly the same.

User avatar
abuss368
Posts: 19926
Joined: Mon Aug 03, 2009 2:33 pm
Location: Where the water is warm, the drinks are cold, and I don't know the names of the players!
Contact:

Re: How much tilting to SCV (rule of thumb)?

Post by abuss368 » Sat Jun 27, 2020 9:17 am

Taylor Larimore wrote:
Fri Jun 26, 2020 4:42 pm
Garland Whizzer:

I am reluctant to put much faith in past performance, nevertheless I am wondering if your post above, showing almost identical returns for VFINX (Vanguard S&P 500 Index Fund) and DFSVX (Dimensional Small-Cap Value Fund), includes the advisor fee required for individual DFSVX funds? What about tax-efficiency?

Thank you and best wishes.
Taylor
Jack Bogle's Words of Wisdom: "No analysis of the past, no matter how painstaking, assures future superiority."
That is interesting. Vanguard has always stated that it is after tax returns that count.
John C. Bogle: Two Fund Portfolio - Total Stock & Total Bond - “Simplicity is the master key to financial success."

garlandwhizzer
Posts: 2846
Joined: Fri Aug 06, 2010 3:42 pm

Re: How much tilting to SCV (rule of thumb)?

Post by garlandwhizzer » Sat Jun 27, 2020 8:13 pm

Taylor wrote:

Garland Whizzer:

I am reluctant to put much faith in past performance, nevertheless I am wondering if your post above, showing almost identical returns for VFINX (Vanguard S&P 500 Index Fund) and DFSVX (Dimensional Small-Cap Value Fund), includes the advisor fee required for individual DFSVX funds? What about tax-efficiency?
Two great points, Taylor, thanks for posting. These figures do not include advisor fees which would have negatively impacted DFSVX, making it the clear loser, not do they include tax-efficiency concerns. TSM has very little portfolio turnover, no short term capital gains generated, and no long term gains not generated until shares are sold. In addition close to 100% of TSM dividends are qualified for tax purposes instead of being taxed as ordinary income. I do not know the specifics of DFSVX in terms of tax treatment but it can't be better than that. TSM offers rock bottom cost, great tax efficiency, and very wide diversification. It is in my portfolio the major core holding in US equity by far (75%+).

Garland Whizzer

garlandwhizzer
Posts: 2846
Joined: Fri Aug 06, 2010 3:42 pm

Re: How much tilting to SCV (rule of thumb)?

Post by garlandwhizzer » Sat Jun 27, 2020 9:22 pm

This is not the correct way to measure fund-level outperformance. The correct way to measure fund-level outperformance is to either use the arithmetic average, or to run a factor regression and project those results on the historical factor premia.
The way to measure fund performance IMO is how much money the fund puts in investor's pockets over the time frame involved. I personally believe that factor models are unrealistic ignoring negatives (trading frictions, trading costs, management fees) and magnify positives (cost-free long short). The models are designed to magnify significance and make a big splash. Their utility in the current real world investing is open to question.In addition the time spas when factor results actually did very well on backtesting 1929 -1993 were in markets (very inefficient vs. more efficient now) and economies (bricks and mortar/industrial manufacturing then vs information/tech now) that were so unlike ours that these long term backtesting results may fall into the junk-in/junk-out category. In addition to all that quants have been arbitraging away all the described factor premiums for years and factor premiums have been steadily declining. I really don't care what factor regressions show if the money doesn't wind up in the investor's pocket. I believe the models are flawed and not to be trusted to put money in investors pockets as has been the case for more than a decade.

In the Portfolio Visualizer model which I used, it assumed that $10K was invested each fund in March 1993 and each fund was followed over the 27 year 3 month time span to see how much was in each fund at the end. After all that time, all three funds (DFSVX, VFINX, VTSMX) had massive gains and the 10K initial investment had turned into more than 114.4 K for all 3. DFSVX at 119.585K did outperform the other two by about 5K in total after 27+ years. That however does not take into account paying advisor access fees for DFSVX and it assumes that the factor investor had the nerve to hold on without panic selling when the fund had a 61% drawdown. Yes, DFSVX did outperform but not significantly so, nothing like the robust premiums described in the literature and the 27 year ride was a lot more difficult emotionally. Was putting up with it for 27 years worth that 5K assuming you were patient enough to hold on through the tough times? Each of us must answer that question when we look at this time period. It appears to be a risk/reward tradeoff.
If you choose to use SCV, which I believe to be optional, it makes more sense at least to me at least to add a modest slice (25%?) of SCV to a big slice (75%?) of beta (TSM). Beta is by far my favorite factor and, unlike SCV where considerable controversy exists on how exactly to harvest it, harvesting beta is simple, fool proof, and dirt cheap with TSM.

Garland Whizzer

User avatar
Forester
Posts: 1264
Joined: Sat Jan 19, 2019 2:50 pm
Location: UK

Re: How much tilting to SCV (rule of thumb)?

Post by Forester » Sun Jun 28, 2020 12:18 am

You could say 2 bull markets for TSM/growth vs 1 bull market for SCV, and SCV was level :confused also a period of historically low inflation. And SCV level at a point in time, when the spread between value & growth is extreme by objective measures. At least it didn't hurt to take the punt on SCV.

muffins14
Posts: 206
Joined: Wed Oct 26, 2016 4:14 am

Re: How much tilting to SCV (rule of thumb)?

Post by muffins14 » Sun Jun 28, 2020 1:11 am

Can we stop talking about the dimensional advisor fee without acknowledging that many people can invest in it for free in a 401k?

The fee question has been raised and answered 100 times by the same people

User avatar
Uncorrelated
Posts: 727
Joined: Sun Oct 13, 2019 3:16 pm

Re: How much tilting to SCV (rule of thumb)?

Post by Uncorrelated » Sun Jun 28, 2020 7:20 am

garlandwhizzer wrote:
Sat Jun 27, 2020 9:22 pm
This is not the correct way to measure fund-level outperformance. The correct way to measure fund-level outperformance is to either use the arithmetic average, or to run a factor regression and project those results on the historical factor premia.
The way to measure fund performance IMO is how much money the fund puts in investor's pockets over the time frame involved.
Do you have only one fund in your portfolio? No, of course not. Your portfolio consists of multiple funds, you care about the characteristics of your entire portfolio.

If you measure fund performance by arithmetic return, then the total portfolio performance is a linear combination of the individual funds. If you measure fund performance by "how much money the fund put's in the investor's pockets" (aka geometric return), then the total portfolio performance cannot be calculated from the individual funds. This is especially misleading when comparing funds that don't have the same risk. Your portfoliovisualizer link suggests that there is only a 0.2% difference is annual return between DFSVX and VTSMX, but the actual difference is 1.2%.

nisiprius did the analysis the right way.

Sure you can use "how much money the fund puts in investor's pockets over the time frame involved" as a performance metric, but only after you have determined the optimal ratio between different funds, and only if you have a logaritmic utility function. If those conditions are not met, the metric is misleading.

sean.mcgrath
Posts: 318
Joined: Thu Dec 29, 2016 6:15 am
Location: US in NL

Re: How much tilting to SCV (rule of thumb)?

Post by sean.mcgrath » Sun Jun 28, 2020 7:46 am

Uncorrelated wrote:
Tue Jun 23, 2020 8:17 am
I suggest tilting to a maximum of 1/3 market, 1/3 SVC and 1/3 value.
Hi UC,

Can you help me to understand this? If I run my portfolio, say, in Portfolio Visualizer and end up with 1.0 MKT, 0.3 HML and 0.2 SMB are you really saying that the goal is to get to 1.0 / 1.0 / 1.0?

thanks,
Sean

User avatar
Uncorrelated
Posts: 727
Joined: Sun Oct 13, 2019 3:16 pm

Re: How much tilting to SCV (rule of thumb)?

Post by Uncorrelated » Sun Jun 28, 2020 9:17 am

sean.mcgrath wrote:
Sun Jun 28, 2020 7:46 am
Uncorrelated wrote:
Tue Jun 23, 2020 8:17 am
I suggest tilting to a maximum of 1/3 market, 1/3 SVC and 1/3 value.
Hi UC,

Can you help me to understand this? If I run my portfolio, say, in Portfolio Visualizer and end up with 1.0 MKT, 0.3 HML and 0.2 SMB are you really saying that the goal is to get to 1.0 / 1.0 / 1.0?

thanks,
Sean
1.0 / 1.0 / 1.0 is theoretically optimal for a mean-variance investor, but unobtainable in practice (You would need long-short funds, which exist but are very expensive). I think splitting your funds equally between TSM, SVC and value is a good compromise between factor exposure, costs, and idiosyncratic risk. As you noted, that results in a rather disappointing factor exposure, but I believe it's the best we can do.

If you want to take even more risk in exchange for more reward, it's probably more effective to use leverage than to use increasingly concentrated SVC funds to try to increase the raw factor exposure further.

sean.mcgrath
Posts: 318
Joined: Thu Dec 29, 2016 6:15 am
Location: US in NL

Re: How much tilting to SCV (rule of thumb)?

Post by sean.mcgrath » Sun Jun 28, 2020 9:23 am

Uncorrelated wrote:
Sun Jun 28, 2020 9:17 am
1.0 / 1.0 / 1.0 is theoretically optimal for a mean-variance investor, but unobtainable in practice (You would need long-short funds, which exist but are very expensive). I think splitting your funds equally between TSM, SVC and value is a good compromise between factor exposure, costs, and idiosyncratic risk. As you noted, that results in a rather disappointing factor exposure, but I believe it's the best we can do.

If you want to take even more risk in exchange for more reward, it's probably more effective to use leverage than to use increasingly concentrated SVC funds to try to increase the raw factor exposure further.
Thank you -- I wasn't sure whether "my" 1.0s were the same as "yours," just because they are so far off from what I can achieve. It turns out that they are. :D

Yes, leverage is out for psychological / behavioral reasons (I don't like it), and I end up Mkt 1.0, SMB 0.2, HML 0.3. Although one nasty surprise I found out during the drops in March was that I am quite over-allocated to Japan. I guess they had to hunt very hard to find Intl Value.

rkhusky
Posts: 9439
Joined: Thu Aug 18, 2011 8:09 pm

Re: How much tilting to SCV (rule of thumb)?

Post by rkhusky » Sun Jun 28, 2020 4:39 pm

sean.mcgrath wrote:
Sun Jun 28, 2020 7:46 am
Uncorrelated wrote:
Tue Jun 23, 2020 8:17 am
I suggest tilting to a maximum of 1/3 market, 1/3 SVC and 1/3 value.
Hi UC,

Can you help me to understand this? If I run my portfolio, say, in Portfolio Visualizer and end up with 1.0 MKT, 0.3 HML and 0.2 SMB are you really saying that the goal is to get to 1.0 / 1.0 / 1.0?

thanks,
Sean
One advantage of the low factor loadings is that if the factor returns continue to be negative, your portfolio won’t underperform the markets by as much.

Topic Author
international001
Posts: 1463
Joined: Thu Feb 15, 2018 7:31 pm

Re: How much tilting to SCV (rule of thumb)?

Post by international001 » Mon Jun 29, 2020 5:33 pm

vineviz wrote:
Thu Jun 25, 2020 7:47 pm
international001 wrote:
Thu Jun 25, 2020 7:23 pm
vineviz wrote:
Thu Jun 25, 2020 2:24 pm


The further you are from retirement, the more portfolio volatility you can probably (or should probably) accept.

Within that portfolio, however, there's no reason that the ratio of SCV to total stock market should decrease over time. If anything, as I mentioned earlier I think, the optimal amount of SCV tilt is probably highest right after retirement.
??
Adding SCV will increase stdev.

Investment A: TSM (stdev 15%)
Investment B: SCV (stdev 20%)

To reach the same stdev after N years you could hold B for NB years or A for NA years (NB/NA = 20**2/15**2), no?
Or simply adjust your stock/bond ratio to get the portfolio volatility you want.
Fair enough. But just looking at PV, efficient frontier tells me that for higher risk, I need fewer bonds (obviously) and a higher percentage of SCV

User avatar
Taylor Larimore
Advisory Board
Posts: 29507
Joined: Tue Feb 27, 2007 8:09 pm
Location: Miami FL

Re: How much tilting to SCV in or near retirement?

Post by Taylor Larimore » Mon Jun 29, 2020 7:18 pm

vineviz wrote:
Thu Jun 25, 2020 2:24 pm
there's no reason that the ratio of SCV to total stock market should decrease over time. If anything, as I mentioned earlier I think, the optimal amount of SCV tilt is probably highest right after retirement.
vineviz:

It is obvious to me that overweighting Small Cap Value stocks in or near retirement can be a major mistake. Consider the current retirement portfolios of investors who tried to beat the market by overweighting Small Cap Value stocks 10 years ago. They are way behind investors using low-cost, tax-efficient, total market index funds. Can there be any doubt that many of these SCV fund investors must either postpone their retirement or substantially cut their retirement budget.

In or near retirement is not the time to speculate in one small cap value (3%) corner of the stock market.

Best wishes.
Taylor
Jack Bogle's Words of Wisdom: "The beauty of owning the market is that you eliminate individual stock risk, you eliminate market sector risk, and you eliminate manager risk."
"Simplicity is the master key to financial success." -- Jack Bogle

User avatar
vineviz
Posts: 6637
Joined: Tue May 15, 2018 1:55 pm

Re: How much tilting to SCV in or near retirement?

Post by vineviz » Mon Jun 29, 2020 7:42 pm

Taylor Larimore wrote:
Mon Jun 29, 2020 7:18 pm

It is obvious to me that overweighting Small Cap Value stocks in or near retirement can be a major mistake. Consider the current retirement portfolios of investors who tried to beat the market by overweighting Small Cap Value stocks 10 years ago. They are way behind investors using low-cost, tax-efficient, total market index funds.
With all due respect, this is a point on which you and I seem to differ: I don't subscribe to the theory that concentrating on asset classes with the greatest recent performance is a smart way to invest.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

JamesDean44
Posts: 149
Joined: Sat Jan 13, 2018 7:36 pm

Re: How much tilting to SCV in or near retirement?

Post by JamesDean44 » Mon Jun 29, 2020 9:41 pm

Taylor Larimore wrote:
Mon Jun 29, 2020 7:18 pm

It is obvious to me that overweighting Small Cap Value stocks in or near retirement can be a major mistake. Consider the current retirement portfolios of investors who tried to beat the market by overweighting Small Cap Value stocks 10 years ago. They are way behind investors using low-cost, tax-efficient, total market index funds. Can there be any doubt that many of these SCV fund investors must either postpone their retirement or substantially cut their retirement budget.

In or near retirement is not the time to speculate in one small cap value (3%) corner of the stock market.

Best wishes.
Taylor
Taylor, I've seen you make similar statements repeatedly. I don't really understand what you're saying. It isn't as if domestic small cap value stocks repeatedly went to 0 in the last ten years. There was no disaster. Here is a simple backtest comparing a 90/10 split between Vanguard's 2020 TDF and VTI with Vanguard's 2020 TDF and DFSVX to approximate a tilt. The backtest uses 2010-YTD to mirror the 10 year period you reference.

90/10 VTI, CAGR 8.84%, ending balance $4,832,390
90/10 DFSVX, CAGR 8.41%, ending balance $4,639,777

https://www.portfoliovisualizer.com/bac ... tion3_2=10

You can obviously change the assumptions and time periods to change the results, including showing results that favor a SCV tilt.

Topic Author
international001
Posts: 1463
Joined: Thu Feb 15, 2018 7:31 pm

Re: How much tilting to SCV (rule of thumb)?

Post by international001 » Wed Jul 01, 2020 4:34 pm

sean.mcgrath wrote:
Sun Jun 28, 2020 9:23 am
Uncorrelated wrote:
Sun Jun 28, 2020 9:17 am
1.0 / 1.0 / 1.0 is theoretically optimal for a mean-variance investor, but unobtainable in practice (You would need long-short funds, which exist but are very expensive). I think splitting your funds equally between TSM, SVC and value is a good compromise between factor exposure, costs, and idiosyncratic risk. As you noted, that results in a rather disappointing factor exposure, but I believe it's the best we can do.

If you want to take even more risk in exchange for more reward, it's probably more effective to use leverage than to use increasingly concentrated SVC funds to try to increase the raw factor exposure further.
Thank you -- I wasn't sure whether "my" 1.0s were the same as "yours," just because they are so far off from what I can achieve. It turns out that they are. :D

Yes, leverage is out for psychological / behavioral reasons (I don't like it), and I end up Mkt 1.0, SMB 0.2, HML 0.3. Although one nasty surprise I found out during the drops in March was that I am quite over-allocated to Japan. I guess they had to hunt very hard to find Intl Value.
For the more ignorant, can you explain what all those coefficients mean?

1.0 Mkt can be changed?
1.0/ 1.0/1.0 isn't that SCV?
How do you determine 1.0/0.2/0.3. and how it translates into a portfolio?

User avatar
nedsaid
Posts: 13437
Joined: Fri Nov 23, 2012 12:33 pm

Re: How much tilting to SCV in or near retirement?

Post by nedsaid » Wed Jul 01, 2020 4:55 pm

vineviz wrote:
Mon Jun 29, 2020 7:42 pm
Taylor Larimore wrote:
Mon Jun 29, 2020 7:18 pm

It is obvious to me that overweighting Small Cap Value stocks in or near retirement can be a major mistake. Consider the current retirement portfolios of investors who tried to beat the market by overweighting Small Cap Value stocks 10 years ago. They are way behind investors using low-cost, tax-efficient, total market index funds.
With all due respect, this is a point on which you and I seem to differ: I don't subscribe to the theory that concentrating on asset classes with the greatest recent performance is a smart way to invest.
Yes, I am sooo old, though not as old as Taylor obviously, I am sooo old that I remember the Death of Value back in 1999. I am old enough to remember articles on Warren Buffett saying that the old boy didn't quite have it anymore, time had passed him by. The next decade, I recall, was pretty darned good for Small Cap Value and Value in general. The 2000's also vindicated Buffett. I am soooo old, that I remember when the S&P 500 was flat from 2000-2012, whatever return you had was from dividends. But no Value junkies that I can remember were posting about the death of the S&P 500 or the Total US Stock Market. No Value junkies that I am aware of telling the 3 funders that their investment strategy was bad during the 2000's. Indeed, I have sprung to the defense of the 3 funders here on the forum many times. The death of Value is greatly exaggerated.

Bogleheads are falling victim to recency bias. As I have joked, we are racing to the Zero Fund portfolio around here. TIPS, REITs, and International Stocks have bitten the dust. There were threads complaining about bonds though a good 2019 silenced a lot of that. So we are down from 5 to 2. Total Bond was in danger for a while and if we get into bear territory again, Total Stock might be on the chopping block again as well. We will be arguing over what the best mattress is out there to stuff our cash into. I am pretty sure it will be a low cost mattress. :wink:

I myself say, just ditch Total Stock Market and just own the FAANG stocks: Facebook, Apple, Amazon, Netflix, and Google. Maybe go real simple and just own Google. The majesty of simplicity! ;o)
A fool and his money are good for business.

JamesDean44
Posts: 149
Joined: Sat Jan 13, 2018 7:36 pm

Re: How much tilting to SCV in or near retirement?

Post by JamesDean44 » Wed Jul 01, 2020 4:58 pm

nedsaid wrote:
Wed Jul 01, 2020 4:55 pm
vineviz wrote:
Mon Jun 29, 2020 7:42 pm
Taylor Larimore wrote:
Mon Jun 29, 2020 7:18 pm

It is obvious to me that overweighting Small Cap Value stocks in or near retirement can be a major mistake. Consider the current retirement portfolios of investors who tried to beat the market by overweighting Small Cap Value stocks 10 years ago. They are way behind investors using low-cost, tax-efficient, total market index funds.
With all due respect, this is a point on which you and I seem to differ: I don't subscribe to the theory that concentrating on asset classes with the greatest recent performance is a smart way to invest.
Yes, I am sooo old, though not as old as Taylor obviously, I am sooo old that I remember the Death of Value back in 1999. I am old enough to remember articles on Warren Buffett saying that the old boy didn't quite have it anymore, time had passed him by. The next decade, I recall, was pretty darned good for Small Cap Value and Value in general. The 2000's also vindicated Buffett. I am soooo old, that I remember when the S&P 500 was flat from 2000-2012, whatever return you had was from dividends. But no Value junkies that I can remember were posting about the death of the S&P 500 or the Total US Stock Market. No Value junkies that I am aware of telling the 3 funders that their investment strategy was bad during the 2000's. Indeed, I have sprung to the defense of the 3 funders here on the forum many times. The death of Value is greatly exaggerated.

Bogleheads are falling victim to recency bias. As I have joked, we are racing to the Zero Fund portfolio around here. TIPS, REITs, and International Stocks have bitten the dust. There were threads complaining about bonds though a good 2019 silenced a lot of that. So we are down from 5 to 2. Total Bond was in danger for a while and if we get into bear territory again, Total Stock might be on the chopping block again as well. We will be arguing over what the best mattress is out there to stuff our cash into. I am pretty sure it will be a low cost mattress. :wink:

I myself say, just ditch Total Stock Market and just own the FAANG stocks: Facebook, Apple, Amazon, Netflix, and Google. Maybe go real simple and just own Google. The majesty of simplicity! ;o)
Well said.

sean.mcgrath
Posts: 318
Joined: Thu Dec 29, 2016 6:15 am
Location: US in NL

Re: How much tilting to SCV (rule of thumb)?

Post by sean.mcgrath » Thu Jul 02, 2020 6:56 am

international001 wrote:
Wed Jul 01, 2020 4:34 pm
For the more ignorant, can you explain what all those coefficients mean?

1.0 Mkt can be changed?
1.0/ 1.0/1.0 isn't that SCV?
How do you determine 1.0/0.2/0.3. and how it translates into a portfolio?
Hi Int,
Basically they are the co-efficients of a number of independent variables in a linear regression, if that means anything to you. The idea is to find variables (they call them "factors") that can explain more of a stock's movement than just the overall market. Here's the most famous attempt, although there are many.

What I do is take an ETF and run a linear regression in Excel against a number of variables that come from this Fama-French study (they publish the data). This creates an equation with the variables SCM, HML, etc. The numbers people talk about are the co-efficients to those variables. You can also go to websites like Portfolio visualizer which will do it for you.

I am not even sure that all this effort brings an additional investment return, but that's how the mechanics work, anyway. :happy

Post Reply