SCV tilt advocates, please suggest a percentage for me to use in a backtest

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply
User avatar
Topic Author
nisiprius
Advisory Board
Posts: 41097
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

SCV tilt advocates, please suggest a percentage for me to use in a backtest

Post by nisiprius » Mon Jun 17, 2019 10:08 am

I want to compare a portfolio of

60% large-cap blend stocks
40% intermediate-term investment-grade bonds

to a comparable portfolio with a reasonable small-cap value tilt.

X% large-cap blend stocks
Y% small-cap value stocks
Z% intermediate-term investment-grade bonds

I want to set X to a "fair" value, one that would be thought sensible and a "typical" recommendations by factor tilters. I don't want to set X so high as to be crazy, nor so low that it cannot move the needle much. Since small-cap value stocks have higher volatility, I would like some adjustment of overall stock/bond allocations so as to roughly equalize the volatility of the two portfolios, so I think Z% ought to be higher than 40%.

Please suggest some numbers. Not necessarily some optimum, but a good, workmanlike tilt, with enough heft for the hoped-for diversification effect to make itself visible.

In an exercise like this that I tried a long time ago, using SBBI numbers for "large-company stocks" and "small-company stocks" (i.e. small-cap blend, not small-cap tilt), 25% large-cap blend, 25% small-cap blend, 50% bonds was pretty comparable, and those are the numbers I was thinking of using this time.
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.

marcopolo
Posts: 3068
Joined: Sat Dec 03, 2016 10:22 am

Re: SCV tilt advocates, please suggest a percentage for me to use in a backtest

Post by marcopolo » Mon Jun 17, 2019 10:12 am

nisiprius wrote:
Mon Jun 17, 2019 10:08 am
I want to compare a portfolio of

60% large-cap blend stocks
40% intermediate-term investment-grade bonds

to a comparable portfolio with a reasonable small-cap value tilt.

X% large-cap blend stocks
Y% small-cap value stocks
Z% intermediate-term investment-grade bonds

I want to set X to a "fair" value, one that would be thought sensible and a "typical" recommendations by factor tilters. I don't want to set X so high as to be crazy, nor so low that it cannot move the needle much. Since small-cap value stocks have higher volatility, I would like some adjustment of overall stock/bond allocations so as to roughly equalize the volatility of the two portfolios, so I think Z% ought to be higher than 40%.

Please suggest some numbers. Not necessarily some optimum, but a good, workmanlike tilt, with enough heft for the hoped-for diversification effect to make itself visible.

In an exercise like this that I tried a long time ago, using SBBI numbers for "large-company stocks" and "small-company stocks" (i.e. small-cap blend, not small-cap tilt), 25% large-cap blend, 25% small-cap blend, 50% bonds was pretty comparable, and those are the numbers I was thinking of using this time.
Not really tilt advocate here, but why not use the "Larry portfolio"?

X=0
Y=30
Z=70
Once in a while you get shown the light, in the strangest of places if you look at it right.

dbr
Posts: 32864
Joined: Sun Mar 04, 2007 9:50 am

Re: SCV tilt advocates, please suggest a percentage for me to use in a backtest

Post by dbr » Mon Jun 17, 2019 10:15 am

I would think one would want to make a chart or a heat map across as wide a range as possible.

One question, is it less complicated to not vary the bond allocation or is it important to explore the "Larry" portfolio of 100% of stocks in small caps but a large bond allocation? I think that is a different question from keeping the bond allocation fixed and just varying X from 0% to 60% and Y from 60% to 0%.

PS Noted was typing while above reply was posted.

marcopolo
Posts: 3068
Joined: Sat Dec 03, 2016 10:22 am

Re: SCV tilt advocates, please suggest a percentage for me to use in a backtest

Post by marcopolo » Mon Jun 17, 2019 10:17 am

dbr wrote:
Mon Jun 17, 2019 10:15 am
I would think one would want to make a chart or a heat map across as wide a range as possible.

One question, is it less complicated to not vary the bond allocation or is it important to explore the "Larry" portfolio of 100% of stocks in small caps but a large bond allocation? I think that is a different question from keeping the bond allocation fixed and just varying X from 0% to 60% and Y from 60% to 0%.

PS Noted was typing while above reply was posted.
I think one of the arguments for tilting is that it allows you to lower equity allocation and still maintain the same expected returns.
Separating the tilt from the equity allocation might not capture the intended benefits.
Once in a while you get shown the light, in the strangest of places if you look at it right.

dbr
Posts: 32864
Joined: Sun Mar 04, 2007 9:50 am

Re: SCV tilt advocates, please suggest a percentage for me to use in a backtest

Post by dbr » Mon Jun 17, 2019 10:23 am

marcopolo wrote:
Mon Jun 17, 2019 10:17 am
dbr wrote:
Mon Jun 17, 2019 10:15 am
I would think one would want to make a chart or a heat map across as wide a range as possible.

One question, is it less complicated to not vary the bond allocation or is it important to explore the "Larry" portfolio of 100% of stocks in small caps but a large bond allocation? I think that is a different question from keeping the bond allocation fixed and just varying X from 0% to 60% and Y from 60% to 0%.

PS Noted was typing while above reply was posted.
I think one of the arguments for tilting is that it allows you to lower equity allocation and still maintain the same expected returns.
Separating the tilt from the equity allocation might not capture the intended benefits.
So the right answer for the study is to vary X, Y, and Z from 0% to 100% under the constraint that the sum is 100% and find a clever way to display the results for CAGR, SD, and Sharpe Ratio, for example. If it turns out the results start to become insensitive in some domain of the inputs the test can be done in large steps. I am not being very sensitive to how much time is required to do all this work.

livesoft
Posts: 72078
Joined: Thu Mar 01, 2007 8:00 pm

Re: SCV tilt advocates, please suggest a percentage for me to use in a backtest

Post by livesoft » Mon Jun 17, 2019 10:33 am

Start with
X = 30
Y = 30
Z = 40
and tell us what the results are. Then go from there.
Wiki This signature message sponsored by sscritic: Learn to fish.

User avatar
willthrill81
Posts: 19207
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: SCV tilt advocates, please suggest a percentage for me to use in a backtest

Post by willthrill81 » Mon Jun 17, 2019 10:35 am

marcopolo wrote:
Mon Jun 17, 2019 10:17 am
dbr wrote:
Mon Jun 17, 2019 10:15 am
I would think one would want to make a chart or a heat map across as wide a range as possible.

One question, is it less complicated to not vary the bond allocation or is it important to explore the "Larry" portfolio of 100% of stocks in small caps but a large bond allocation? I think that is a different question from keeping the bond allocation fixed and just varying X from 0% to 60% and Y from 60% to 0%.

PS Noted was typing while above reply was posted.
I think one of the arguments for tilting is that it allows you to lower equity allocation and still maintain the same expected returns.
Well, that's the idea behind the Larry Portfolio. The problem is that the SCV premium needed for this to work is far from guaranteed. From 2004 until now, TSM and SCV have had nearly identical returns. If you weren't assuming that SCV would have higher returns, then you should be satisfied with this outcome, but if you were anticipating a SCV premium and had lowered your equity allocation in anticipation of it, you would probably be disappointed.
“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


marcopolo
Posts: 3068
Joined: Sat Dec 03, 2016 10:22 am

Re: SCV tilt advocates, please suggest a percentage for me to use in a backtest

Post by marcopolo » Mon Jun 17, 2019 10:45 am

willthrill81 wrote:
Mon Jun 17, 2019 10:35 am
marcopolo wrote:
Mon Jun 17, 2019 10:17 am
dbr wrote:
Mon Jun 17, 2019 10:15 am
I would think one would want to make a chart or a heat map across as wide a range as possible.

One question, is it less complicated to not vary the bond allocation or is it important to explore the "Larry" portfolio of 100% of stocks in small caps but a large bond allocation? I think that is a different question from keeping the bond allocation fixed and just varying X from 0% to 60% and Y from 60% to 0%.

PS Noted was typing while above reply was posted.
I think one of the arguments for tilting is that it allows you to lower equity allocation and still maintain the same expected returns.
Well, that's the idea behind the Larry Portfolio. The problem is that the SCV premium needed for this to work is far from guaranteed. From 2004 until now, TSM and SCV have had nearly identical returns. If you weren't assuming that SCV would have higher returns, then you should be satisfied with this outcome, but if you were anticipating a SCV premium and had lowered your equity allocation in anticipation of it, you would probably be disappointed.
Well, if you weren't expecting higher returns, what would be the point?
Once in a while you get shown the light, in the strangest of places if you look at it right.

User avatar
willthrill81
Posts: 19207
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: SCV tilt advocates, please suggest a percentage for me to use in a backtest

Post by willthrill81 » Mon Jun 17, 2019 10:55 am

marcopolo wrote:
Mon Jun 17, 2019 10:45 am
willthrill81 wrote:
Mon Jun 17, 2019 10:35 am
marcopolo wrote:
Mon Jun 17, 2019 10:17 am
dbr wrote:
Mon Jun 17, 2019 10:15 am
I would think one would want to make a chart or a heat map across as wide a range as possible.

One question, is it less complicated to not vary the bond allocation or is it important to explore the "Larry" portfolio of 100% of stocks in small caps but a large bond allocation? I think that is a different question from keeping the bond allocation fixed and just varying X from 0% to 60% and Y from 60% to 0%.

PS Noted was typing while above reply was posted.
I think one of the arguments for tilting is that it allows you to lower equity allocation and still maintain the same expected returns.
Well, that's the idea behind the Larry Portfolio. The problem is that the SCV premium needed for this to work is far from guaranteed. From 2004 until now, TSM and SCV have had nearly identical returns. If you weren't assuming that SCV would have higher returns, then you should be satisfied with this outcome, but if you were anticipating a SCV premium and had lowered your equity allocation in anticipation of it, you would probably be disappointed.
Well, if you weren't expecting higher returns, what would be the point?
Some might view such a strategy as giving one the opportunity to outperform TSM. But one doesn't have to plan on outperforming TSM.

Besides that, even if you did outperform TSM, you don't know that the outperformance would be enough to justify reducing your stock exposure as much as you did.

Let's take a look at the Larry Portfolio's performance since 2004 in PV compared to a 3-fund portfolio with 15% TSM/15% ex-U.S. stock/70% TBM. Both had nearly identical returns (LP = 5.17%, 3F = 5.25%), but the LP was more volatile (SD = 6.08% vs. 4.97%), had a bigger drawdown (18.29% vs. 15.01%), and a worse Sharpe (.66 vs. 80).
“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

DecumulatorDoc
Posts: 191
Joined: Sun Jan 28, 2018 7:48 am

Re: SCV tilt advocates, please suggest a percentage for me to use in a backtest

Post by DecumulatorDoc » Mon Jun 17, 2019 11:03 am

X=50
Y=10
Z=40

YRT70
Posts: 536
Joined: Sat Apr 27, 2019 8:51 am

Re: SCV tilt advocates, please suggest a percentage for me to use in a backtest

Post by YRT70 » Mon Jun 17, 2019 11:10 am

DecumulatorDoc wrote:
Mon Jun 17, 2019 11:03 am
X=50
Y=10
Z=40
From what I've read that wouldn't give a significant SCV tilt because it holds more large cap blend stocks.

I could be wrong though, I'm quite new to this.

MotoTrojan
Posts: 9952
Joined: Wed Feb 01, 2017 8:39 pm

Re: SCV tilt advocates, please suggest a percentage for me to use in a backtest

Post by MotoTrojan » Mon Jun 17, 2019 11:12 am

I think 50/50 on equity side is reasonable for this and look forward to your results. What has been most impressive to me though is the impact SCV has had historically on portfolio performance in withdrawal stage.

marcopolo
Posts: 3068
Joined: Sat Dec 03, 2016 10:22 am

Re: SCV tilt advocates, please suggest a percentage for me to use in a backtest

Post by marcopolo » Mon Jun 17, 2019 11:12 am

willthrill81 wrote:
Mon Jun 17, 2019 10:55 am
marcopolo wrote:
Mon Jun 17, 2019 10:45 am
willthrill81 wrote:
Mon Jun 17, 2019 10:35 am
marcopolo wrote:
Mon Jun 17, 2019 10:17 am
dbr wrote:
Mon Jun 17, 2019 10:15 am
I would think one would want to make a chart or a heat map across as wide a range as possible.

One question, is it less complicated to not vary the bond allocation or is it important to explore the "Larry" portfolio of 100% of stocks in small caps but a large bond allocation? I think that is a different question from keeping the bond allocation fixed and just varying X from 0% to 60% and Y from 60% to 0%.

PS Noted was typing while above reply was posted.
I think one of the arguments for tilting is that it allows you to lower equity allocation and still maintain the same expected returns.
Well, that's the idea behind the Larry Portfolio. The problem is that the SCV premium needed for this to work is far from guaranteed. From 2004 until now, TSM and SCV have had nearly identical returns. If you weren't assuming that SCV would have higher returns, then you should be satisfied with this outcome, but if you were anticipating a SCV premium and had lowered your equity allocation in anticipation of it, you would probably be disappointed.
Well, if you weren't expecting higher returns, what would be the point?
Some might view such a strategy as giving one the opportunity to outperform TSM. But one doesn't have to plan on outperforming TSM.

Besides that, even if you did outperform TSM, you don't know that the outperformance would be enough to justify reducing your stock exposure as much as you did.

Let's take a look at the Larry Portfolio's performance since 2004 in PV compared to a 3-fund portfolio with 15% TSM/15% ex-U.S. stock/70% TBM. Both had nearly identical returns (LP = 5.17%, 3F = 5.25%), but the LP was more volatile (SD = 6.08% vs. 4.97%), had a bigger drawdown (18.29% vs. 15.01%), and a worse Sharpe (.66 vs. 80).
You are "preaching to the choir". I am not sold on the advantage of factor tilting, prospectively. I admit it looks nice historically.
Larry would say you are "resulting".

My point was that for the analysis, it would seems one should try to mimic what the factor tilting advocates are proposing.
Once in a while you get shown the light, in the strangest of places if you look at it right.

User avatar
willthrill81
Posts: 19207
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: SCV tilt advocates, please suggest a percentage for me to use in a backtest

Post by willthrill81 » Mon Jun 17, 2019 11:21 am

marcopolo wrote:
Mon Jun 17, 2019 11:12 am
willthrill81 wrote:
Mon Jun 17, 2019 10:55 am
marcopolo wrote:
Mon Jun 17, 2019 10:45 am
willthrill81 wrote:
Mon Jun 17, 2019 10:35 am
marcopolo wrote:
Mon Jun 17, 2019 10:17 am


I think one of the arguments for tilting is that it allows you to lower equity allocation and still maintain the same expected returns.
Well, that's the idea behind the Larry Portfolio. The problem is that the SCV premium needed for this to work is far from guaranteed. From 2004 until now, TSM and SCV have had nearly identical returns. If you weren't assuming that SCV would have higher returns, then you should be satisfied with this outcome, but if you were anticipating a SCV premium and had lowered your equity allocation in anticipation of it, you would probably be disappointed.
Well, if you weren't expecting higher returns, what would be the point?
Some might view such a strategy as giving one the opportunity to outperform TSM. But one doesn't have to plan on outperforming TSM.

Besides that, even if you did outperform TSM, you don't know that the outperformance would be enough to justify reducing your stock exposure as much as you did.

Let's take a look at the Larry Portfolio's performance since 2004 in PV compared to a 3-fund portfolio with 15% TSM/15% ex-U.S. stock/70% TBM. Both had nearly identical returns (LP = 5.17%, 3F = 5.25%), but the LP was more volatile (SD = 6.08% vs. 4.97%), had a bigger drawdown (18.29% vs. 15.01%), and a worse Sharpe (.66 vs. 80).
You are "preaching to the choir". I am not sold on the advantage of factor tilting, prospectively. I admit it looks nice historically.
Larry would say you are "resulting".
It wouldn't be appropriate to accuse me of resulting as I've not condemned the LP in any way nor the process underlying its development. The point I'm trying to make is that one's assumption that SCV or any factor tilt will outperform TSM over your specific investing time frame may well prove false. Even the equity premium over bonds may be negative for a meaningful period of time (e.g. 2000-2009).
“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

chisey
Posts: 275
Joined: Mon Apr 16, 2007 10:47 am

Re: SCV tilt advocates, please suggest a percentage for me to use in a backtest

Post by chisey » Mon Jun 17, 2019 11:27 am

livesoft wrote:
Mon Jun 17, 2019 10:33 am
Start with
X = 30
Y = 30
Z = 40
and tell us what the results are. Then go from there.
MotoTrojan wrote:
Mon Jun 17, 2019 11:12 am
I think 50/50 on equity side is reasonable for this and look forward to your results. What has been most impressive to me though is the impact SCV has had historically on portfolio performance in withdrawal stage.
I'm with these two. Split it evenly, 30/30/40.

DecumulatorDoc
Posts: 191
Joined: Sun Jan 28, 2018 7:48 am

Re: SCV tilt advocates, please suggest a percentage for me to use in a backtest

Post by DecumulatorDoc » Mon Jun 17, 2019 11:34 am

YRT70 wrote:
Mon Jun 17, 2019 11:10 am
DecumulatorDoc wrote:
Mon Jun 17, 2019 11:03 am
X=50
Y=10
Z=40
From what I've read that wouldn't give a significant SCV tilt because it holds more large cap blend stocks.

I could be wrong though, I'm quite new to this.
Nisiprius asked for "reasonable", "sensible" and "typical" SCV tilt suggestions, not necessarily "significant". This is approximately what I do, SCV represents about 15-20% of equities.

Up to a 25% tilt used to be a common recommendation here (Rick Ferri included), and one I am comfortable with. Didn't realize there would be wrong answers to this post.
Last edited by DecumulatorDoc on Mon Jun 17, 2019 1:21 pm, edited 1 time in total.

MotoTrojan
Posts: 9952
Joined: Wed Feb 01, 2017 8:39 pm

Re: SCV tilt advocates, please suggest a percentage for me to use in a backtest

Post by MotoTrojan » Mon Jun 17, 2019 11:38 am

I said 50/50 but my actual split is 27% SCV 73% Total US, so maybe call that 30% SCV relative to large blend.

User avatar
willthrill81
Posts: 19207
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: SCV tilt advocates, please suggest a percentage for me to use in a backtest

Post by willthrill81 » Mon Jun 17, 2019 11:40 am

dbr wrote:
Mon Jun 17, 2019 10:15 am
I would think one would want to make a chart or a heat map across as wide a range as possible.
I entirely agree. Many will argue that any single percentage is too low/high. And a chart or heat map would provide greater insight as well.
“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

marcopolo
Posts: 3068
Joined: Sat Dec 03, 2016 10:22 am

Re: SCV tilt advocates, please suggest a percentage for me to use in a backtest

Post by marcopolo » Mon Jun 17, 2019 11:40 am

willthrill81 wrote:
Mon Jun 17, 2019 11:21 am
marcopolo wrote:
Mon Jun 17, 2019 11:12 am
willthrill81 wrote:
Mon Jun 17, 2019 10:55 am
marcopolo wrote:
Mon Jun 17, 2019 10:45 am
willthrill81 wrote:
Mon Jun 17, 2019 10:35 am


Well, that's the idea behind the Larry Portfolio. The problem is that the SCV premium needed for this to work is far from guaranteed. From 2004 until now, TSM and SCV have had nearly identical returns. If you weren't assuming that SCV would have higher returns, then you should be satisfied with this outcome, but if you were anticipating a SCV premium and had lowered your equity allocation in anticipation of it, you would probably be disappointed.
Well, if you weren't expecting higher returns, what would be the point?
Some might view such a strategy as giving one the opportunity to outperform TSM. But one doesn't have to plan on outperforming TSM.

Besides that, even if you did outperform TSM, you don't know that the outperformance would be enough to justify reducing your stock exposure as much as you did.

Let's take a look at the Larry Portfolio's performance since 2004 in PV compared to a 3-fund portfolio with 15% TSM/15% ex-U.S. stock/70% TBM. Both had nearly identical returns (LP = 5.17%, 3F = 5.25%), but the LP was more volatile (SD = 6.08% vs. 4.97%), had a bigger drawdown (18.29% vs. 15.01%), and a worse Sharpe (.66 vs. 80).
You are "preaching to the choir". I am not sold on the advantage of factor tilting, prospectively. I admit it looks nice historically.
Larry would say you are "resulting".
It wouldn't be appropriate to accuse me of resulting as I've not condemned the LP in any way nor the process underlying its development. The point I'm trying to make is that one's assumption that SCV or any factor tilt will outperform TSM over your specific investing time frame may well prove false. Even the equity premium over bonds may be negative for a meaningful period of time (e.g. 2000-2009).
Maybe I am confused about what you said upthread.

I think you said that if people expected higher returns from SCV, then they would have been disappointed.

I think the factor tilting argument is that you do it precisely because SCV has a higher expected return. Those expectations may not matierlize, but in that case you should not be disappointed because the strategy based on the previous expected returns is more important than the outcome, which can not be known at the time you set about with your strategy.

Being disappointed, and maybe even looking to see if expectations were met, would be "resulting". At least, that is the way understood the argument from various threads here.
Once in a while you get shown the light, in the strangest of places if you look at it right.

User avatar
willthrill81
Posts: 19207
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: SCV tilt advocates, please suggest a percentage for me to use in a backtest

Post by willthrill81 » Mon Jun 17, 2019 11:48 am

marcopolo wrote:
Mon Jun 17, 2019 11:40 am
willthrill81 wrote:
Mon Jun 17, 2019 11:21 am
marcopolo wrote:
Mon Jun 17, 2019 11:12 am
willthrill81 wrote:
Mon Jun 17, 2019 10:55 am
marcopolo wrote:
Mon Jun 17, 2019 10:45 am


Well, if you weren't expecting higher returns, what would be the point?
Some might view such a strategy as giving one the opportunity to outperform TSM. But one doesn't have to plan on outperforming TSM.

Besides that, even if you did outperform TSM, you don't know that the outperformance would be enough to justify reducing your stock exposure as much as you did.

Let's take a look at the Larry Portfolio's performance since 2004 in PV compared to a 3-fund portfolio with 15% TSM/15% ex-U.S. stock/70% TBM. Both had nearly identical returns (LP = 5.17%, 3F = 5.25%), but the LP was more volatile (SD = 6.08% vs. 4.97%), had a bigger drawdown (18.29% vs. 15.01%), and a worse Sharpe (.66 vs. 80).
You are "preaching to the choir". I am not sold on the advantage of factor tilting, prospectively. I admit it looks nice historically.
Larry would say you are "resulting".
It wouldn't be appropriate to accuse me of resulting as I've not condemned the LP in any way nor the process underlying its development. The point I'm trying to make is that one's assumption that SCV or any factor tilt will outperform TSM over your specific investing time frame may well prove false. Even the equity premium over bonds may be negative for a meaningful period of time (e.g. 2000-2009).
Maybe I am confused about what you said upthread.

I think you said that if people expected higher returns from SCV, then they would have been disappointed.

I think the factor tilting argument is that you do it precisely because SCV has a higher expected return. Those expectations may not matierlize, but in that case you should not be disappointed because the strategy based on the previous expected returns is more important than the outcome, which can not be known at the time you set about with your strategy.

Being disappointed, and maybe even looking to see if expectations were met, would be "resulting". At least, that is the way understood the argument from various threads here.
Being disappointed is not the same as resulting. Resulting refers to conflating the quality of a strategy or process with the outcome. If one was disappointed with SCV's performance, it does not necessarily mean that they believe a SCV tilt to be a poor strategy. Poster nedsaid is a great example of this.
“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

marcopolo
Posts: 3068
Joined: Sat Dec 03, 2016 10:22 am

Re: SCV tilt advocates, please suggest a percentage for me to use in a backtest

Post by marcopolo » Mon Jun 17, 2019 11:50 am

willthrill81 wrote:
Mon Jun 17, 2019 11:48 am
marcopolo wrote:
Mon Jun 17, 2019 11:40 am
willthrill81 wrote:
Mon Jun 17, 2019 11:21 am
marcopolo wrote:
Mon Jun 17, 2019 11:12 am
willthrill81 wrote:
Mon Jun 17, 2019 10:55 am


Some might view such a strategy as giving one the opportunity to outperform TSM. But one doesn't have to plan on outperforming TSM.

Besides that, even if you did outperform TSM, you don't know that the outperformance would be enough to justify reducing your stock exposure as much as you did.

Let's take a look at the Larry Portfolio's performance since 2004 in PV compared to a 3-fund portfolio with 15% TSM/15% ex-U.S. stock/70% TBM. Both had nearly identical returns (LP = 5.17%, 3F = 5.25%), but the LP was more volatile (SD = 6.08% vs. 4.97%), had a bigger drawdown (18.29% vs. 15.01%), and a worse Sharpe (.66 vs. 80).
You are "preaching to the choir". I am not sold on the advantage of factor tilting, prospectively. I admit it looks nice historically.
Larry would say you are "resulting".
It wouldn't be appropriate to accuse me of resulting as I've not condemned the LP in any way nor the process underlying its development. The point I'm trying to make is that one's assumption that SCV or any factor tilt will outperform TSM over your specific investing time frame may well prove false. Even the equity premium over bonds may be negative for a meaningful period of time (e.g. 2000-2009).
Maybe I am confused about what you said upthread.

I think you said that if people expected higher returns from SCV, then they would have been disappointed.

I think the factor tilting argument is that you do it precisely because SCV has a higher expected return. Those expectations may not matierlize, but in that case you should not be disappointed because the strategy based on the previous expected returns is more important than the outcome, which can not be known at the time you set about with your strategy.

Being disappointed, and maybe even looking to see if expectations were met, would be "resulting". At least, that is the way understood the argument from various threads here.
Being disappointed is not the same as resulting. Resulting refers to conflating the quality of a strategy or process with the outcome. If one was disappointed with SCV's performance, it does not necessarily mean that they believe a SCV tilt to be a poor strategy. Poster nedsaid is a great example of this.
Fair point. Thanks.
Once in a while you get shown the light, in the strangest of places if you look at it right.

stlutz
Posts: 5555
Joined: Fri Jan 02, 2009 1:08 am

Re: SCV tilt advocates, please suggest a percentage for me to use in a backtest

Post by stlutz » Mon Jun 17, 2019 8:10 pm

nisiprius wrote:
Mon Jun 17, 2019 10:08 am
I want to compare a portfolio of

60% large-cap blend stocks
40% intermediate-term investment-grade bonds

to a comparable portfolio with a reasonable small-cap value tilt.
...

Please suggest some numbers. Not necessarily some optimum, but a good, workmanlike tilt, with enough heft for the hoped-for diversification effect to make itself visible.
Back when Rick Ferri was more of a pro-tilting guy, he used to break the stock portion up into 75% total market and 25% SV (e.g. https://www.etf.com/sections/index-inve ... nopaging=1). I think he usually left the bond side alone. So, 45/15/40 seem to be a fairly "official" answer.

SpideyIndexer
Posts: 526
Joined: Thu Apr 02, 2015 10:13 pm

Re: SCV tilt advocates, please suggest a percentage for me to use in a backtest

Post by SpideyIndexer » Tue Jun 18, 2019 11:33 am

If you simply backtest, you'll probably rediscover that best is 100% SCV for maximizing return, or something like 60% SCV / 40% long term treasuries for tempering return with lower volatility. But hindsight is always 20 /20. Few of us expect the stock and bond markets to exactly repeat past results.

I suggest doing Monte Carlo rather than just backtesting, and still adjust with some conventional wisdom. It just isn't an exact science ala Newtonian physics.

If you discover something better, please let us know. Good luck.

marcopolo
Posts: 3068
Joined: Sat Dec 03, 2016 10:22 am

Re: SCV tilt advocates, please suggest a percentage for me to use in a backtest

Post by marcopolo » Tue Jun 18, 2019 11:37 am

SpideyIndexer wrote:
Tue Jun 18, 2019 11:33 am
If you simply backtest, you'll probably rediscover that best is 100% SCV for maximizing return, or something like 60% SCV / 40% long term treasuries for tempering return with lower volatility. But hindsight is always 20 /20. Few of us expect the stock and bond markets to exactly repeat past results.

I suggest doing Monte Carlo rather than just backtesting, and still adjust with some conventional wisdom. It just isn't an exact science ala Newtonian physics.

If you discover something better, please let us know. Good luck.
What would you suggest for parameters of the Monte Carlo.
Based, on the relative parameters you assign to the various asset classes, you could get whatever result you want.
If you just use historical returns and variances, then you will have the same built in bias you are (rightfully) concerned about above.
Once in a while you get shown the light, in the strangest of places if you look at it right.


User avatar
vineviz
Posts: 6749
Joined: Tue May 15, 2018 1:55 pm

Re: SCV tilt advocates, please suggest a percentage for me to use in a backtest

Post by vineviz » Wed Jun 19, 2019 12:47 am

nisiprius wrote:
Mon Jun 17, 2019 10:08 am

Please suggest some numbers. Not necessarily some optimum, but a good, workmanlike tilt, with enough heft for the hoped-for diversification effect to make itself visible.
If we’re only talking US equities , I’d say 40-50% SCV is entirely reasonable. The remainder being TSM or maybe large cap growth.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

User avatar
JoMoney
Posts: 9338
Joined: Tue Jul 23, 2013 5:31 am

Re: SCV tilt advocates, please suggest a percentage for me to use in a backtest

Post by JoMoney » Wed Jun 19, 2019 8:15 am

The FundAdvice Ultimate Buy & Hold uses 6% SCV, but that's on top of the 6% to Large-Value and 6% to small blend
https://www.marketwatch.com/lazyportfol ... e-buy-hold

So maybe something in the 10-20% range.
When Mr. Bogle wrote the Tell-Tale chart, SCV wasn't nearly as in vogue, but it seems splitting into 4 allocation with 25% to each of the major 'slices' was the thing. Maybe he would suggest no more than 15% a slice in Slice-&-Dice portfolio:
https://www.vanguard.com/bogle_site/sp20020626.html
... If we take the extra risk into account, there's a real question about whether the game is worth the candle. And even if you don't accept my challenge to S&D, I urge you, before you plunge into a 4x25 portfolio, to put more than 25% in the total market—say 55%. Then put just 15% in the three slices that you dice, thereby taking much of the risk out of your decision...
"To achieve satisfactory investment results is easier than most people realize; to achieve superior results is harder than it looks." - Benjamin Graham

User avatar
willthrill81
Posts: 19207
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: SCV tilt advocates, please suggest a percentage for me to use in a backtest

Post by willthrill81 » Wed Jun 19, 2019 3:08 pm

JoMoney wrote:
Wed Jun 19, 2019 8:15 am
The FundAdvice Ultimate Buy & Hold uses 6% SCV, but that's on top of the 6% to Large-Value and 6% to small blend
https://www.marketwatch.com/lazyportfol ... e-buy-hold
Paul Merriman's UBH 'versions' #6, #7, and #8 also include an allocation to international SCV, 10%, 10%, and 20%, respectively.
“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

benway
Posts: 74
Joined: Thu Jun 30, 2011 5:17 pm

Re: SCV tilt advocates, please suggest a percentage for me to use in a backtest

Post by benway » Wed Jun 19, 2019 7:48 pm

Back in 2016, Larry Swedroe wrote that the large majority of his clients have roughly a .3 value and size loading- viewtopic.php?f=10&t=193537&start=50

According to Portfolio Visualizer, the following ETF combination gives a .3 size and value loading.

IVV (S&P 500): 29%
IVE (S&P 500 Large Value): 27%
IJS (S&P 400 Small Value): 44%

PV Link: https://www.portfoliovisualizer.com/fac ... total1=100

User avatar
vineviz
Posts: 6749
Joined: Tue May 15, 2018 1:55 pm

Re: SCV tilt advocates, please suggest a percentage for me to use in a backtest

Post by vineviz » Thu Jun 20, 2019 11:48 am

benway wrote:
Wed Jun 19, 2019 7:48 pm
Back in 2016, Larry Swedroe wrote that the large majority of his clients have roughly a .3 value and size loading- viewtopic.php?f=10&t=193537&start=50

According to Portfolio Visualizer, the following ETF combination gives a .3 size and value loading.

IVV (S&P 500): 29%
IVE (S&P 500 Large Value): 27%
IJS (S&P 400 Small Value): 44%

PV Link: https://www.portfoliovisualizer.com/fac ... total1=100
A cost-conscious investor could achieve similar factor loadings with just two two funds:

65.00% iShares Core S&P US Value ETF (IUSV)
35.00% iShares Core S&P Small-Cap ETF (IJR)

Total ER of this combination is just 5bps

PV Link: https://www.portfoliovisualizer.com/fac ... total1=100
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

heyyou
Posts: 3702
Joined: Tue Feb 20, 2007 4:58 pm

Re: SCV tilt advocates, please suggest a percentage for me to use in a backtest

Post by heyyou » Thu Jun 20, 2019 5:15 pm

Not an answer to your specific question, but I just use equal slices of several equity sub-asset classes, then hope for adequacy instead of seeking historical optimization in my future, total returns. I'll name that the "don't know, but don't care" method.

I wonder if the affect of reducing Large Growth exposure at the portfolio level, indirectly helps SCV advocates?
Michael H. McClung in Living Off Your Money on page 184 in the paper version, uses a 50/50 stock/bond mix in his recommended retirement portfolios, but this was some of the data used for choosing his stock allocations.

Max SWR at 100% confidence level, 30 year retirements
US SCV 5.20 %
US Large Value 4.90 %
70%TSM,10%Sm,10%LV,10%SCV 4.40 %
TSM 3.50 %
US Large Growth 3.20 %
US Sm Growth 2.70 %

Perhaps the proportions of the various results could be helpful to you.

Post Reply