Most Trusted Risk Premia Factors?
-
- Posts: 236
- Joined: Sun Feb 09, 2020 11:30 am
Most Trusted Risk Premia Factors?
Last edited by Register44 on Mon Jan 11, 2021 7:36 pm, edited 1 time in total.
Re: Most Trusted Risk Premia Factors?
I trust low vol (SPLV, USMV, ACWV etc) the most.
1) low turnover 2) boring 3) disliked by the FF factor academic crew, can't be explained as a risk premium
1) low turnover 2) boring 3) disliked by the FF factor academic crew, can't be explained as a risk premium
Amateur Self-Taught Senior Macro Strategist
- asset_chaos
- Posts: 2629
- Joined: Tue Feb 27, 2007 5:13 pm
- Location: Melbourne
Re: Most Trusted Risk Premia Factors?
Market, term, and credit are also risk factors that I believe almost all accept as real, have premia over the risk-free asset, and are expected to persist.
Regards, |
|
Guy
Re: Most Trusted Risk Premia Factors?
I recommend the book Your Complete Guide to Factor‑Based Investing by Andrew Berkin and Larry Swedroe
Swedroe was a huge contributor to this forum until a couple years ago
Swedroe was a huge contributor to this forum until a couple years ago
I'm just a fan of the person I got my user name from
-
- Posts: 1397
- Joined: Mon Apr 24, 2017 11:16 am
Re: Most Trusted Risk Premia Factors?
Odd silver lining. Isn't that the same as saying, "At least with value if the strategy fails you end up with [less money]"??? If any factor strategy fails, you end up with less than had you not chased the factor. Why is failing with value any better than failing with momentum?Register44 wrote: ↑Tue Oct 20, 2020 2:31 pm
I know quality, size, and momentum are other popular ones. I question though how they will hold up in the future. At least with value if the strategy fails you end up with a stock fund that isn't grossly overpriced as you would possibly with momentum.
- whodidntante
- Posts: 13114
- Joined: Thu Jan 21, 2016 10:11 pm
- Location: outside the echo chamber
Re: Most Trusted Risk Premia Factors?
You can also go to the original research and judge for yourself, or you can through an interpreter like Larry Swedroe or Ben Felix. I don't consider Bogleheads a useful factor learning resource for a beginner. There is simply too much misinformation here and your filter is not tuned.
-
- Posts: 236
- Joined: Sun Feb 09, 2020 11:30 am
removed
Last edited by Register44 on Mon Jan 11, 2021 7:35 pm, edited 1 time in total.
-
- Posts: 15363
- Joined: Fri Apr 10, 2015 12:29 am
Re: Most Trusted Risk Premia Factors?
Market beta is the most robust and reliable equity risk premium factor.Register44 wrote: ↑Tue Oct 20, 2020 2:31 pm From what I have read here at bogleheads and elsewhere that there is evidence that small value may offer a persisting risk premium. Of course at this very moment is the time many may have given up that it doesn't "work" anymore.
I know quality, size, and momentum are other popular ones. I question though how they will hold up in the future. At least with value if the strategy fails you end up with a stock fund that isn't grossly overpriced as you would possibly with momentum.
Are there any known factors beyond Value that you would trust?
- whodidntante
- Posts: 13114
- Joined: Thu Jan 21, 2016 10:11 pm
- Location: outside the echo chamber
Re: Most Trusted Risk Premia Factors?
You might enjoy the rational reminder podcast, then. They dive a bit deeper there.Register44 wrote: ↑Tue Oct 20, 2020 11:06 pmI actually watched Ben Felix's video on the value risk premium last night and that prompted me to ask about it here.whodidntante wrote: ↑Tue Oct 20, 2020 10:54 pm You can also go to the original research and judge for yourself, or you can through an interpreter like Larry Swedroe or Ben Felix. I don't consider Bogleheads a useful factor learning resource for a beginner. There is simply too much misinformation here and your filter is not tuned.
I was just curious if others accepted it or not. Felix stayed pretty neutral, but did not rule out that it was gone.
-
- Posts: 1267
- Joined: Fri Apr 14, 2017 4:25 am
- Location: Israel
Re: Most Trusted Risk Premia Factors?
Aside from market, term, and value? Maybe profitability and investment, but those are just linear components of value anyway. Quality seems somewhat promising as well.
Re: Most Trusted Risk Premia Factors?
I'm not sure if I believe in risk premia (e.g. where is the compensation for inflation risk in fixed income?), but I would say market beta.
Re: Most Trusted Risk Premia Factors?
World Min Vol was down 15.4% in Q1 2020, versus 20.9% for the World Index. Globally defensive sectors did outperform cyclicals. The US market is more tech, tech had an extraordinary run this year, so US low vol stratgies were more muted compared to a prior selloff such as Q4 2018 or the 2015/16 correction.
Amateur Self-Taught Senior Macro Strategist
-
- Posts: 49026
- Joined: Fri May 11, 2007 11:07 am
Re: Most Trusted Risk Premia Factors?
A lot of institutional money has gone into Low Vol since it was first published about. Tended to push up the stocks valuation.Forester wrote: ↑Wed Oct 21, 2020 3:05 amWorld Min Vol was down 15.4% in Q1 2020, versus 20.9% for the World Index. Globally defensive sectors did outperform cyclicals. The US market is more tech, tech had an extraordinary run this year, so US low vol stratgies were more muted compared to a prior selloff such as Q4 2018 or the 2015/16 correction.
-
- Posts: 1267
- Joined: Fri Apr 14, 2017 4:25 am
- Location: Israel
Re: Most Trusted Risk Premia Factors?
That is my fear. I own LV, but the P/Es are not lowValuethinker wrote: ↑Wed Oct 21, 2020 3:29 amA lot of institutional money has gone into Low Vol since it was first published about. Tended to push up the stocks valuation.Forester wrote: ↑Wed Oct 21, 2020 3:05 amWorld Min Vol was down 15.4% in Q1 2020, versus 20.9% for the World Index. Globally defensive sectors did outperform cyclicals. The US market is more tech, tech had an extraordinary run this year, so US low vol stratgies were more muted compared to a prior selloff such as Q4 2018 or the 2015/16 correction.
Re: Most Trusted Risk Premia Factors?
Global Min Vol is slightly cheaper than the index https://www.msci.com/documents/10199/9a ... 7031be35ac
Amateur Self-Taught Senior Macro Strategist
Re: Most Trusted Risk Premia Factors?
That's one of the major issues with risk premia. If something is really a risk premium, it's risky. You hope you'll be rewarded for taking the risk, but there's no guarantee - that's the idea of risk. Risk could include underperformance for a very very long time.Valuethinker wrote: ↑Wed Oct 21, 2020 3:29 amA lot of institutional money has gone into Low Vol since it was first published about. Tended to push up the stocks valuation.Forester wrote: ↑Wed Oct 21, 2020 3:05 amWorld Min Vol was down 15.4% in Q1 2020, versus 20.9% for the World Index. Globally defensive sectors did outperform cyclicals. The US market is more tech, tech had an extraordinary run this year, so US low vol stratgies were more muted compared to a prior selloff such as Q4 2018 or the 2015/16 correction.
Another issue, as Valuethinker points out, is that many investors tend to pile into these investments, leading to higher valuations which, in turn, lowers future returns.
It's also hard to know if something is just a statistical artifact. It may be a genuine risk factor or it may be a coincidence.
Re: Most Trusted Risk Premia Factors?
Dirt cheap market beta, all others are interesting but suspect.
- Uncorrelated
- Posts: 1237
- Joined: Sun Oct 13, 2019 3:16 pm
Re: Most Trusted Risk Premia Factors?
Value (HmL) is quite strong. Some claim it hasn't been as strong recently in the US, but that is not a reason to abandon it.
https://papers.ssrn.com/sol3/papers.cfm ... id=3525096
t-stat is a measure of statistical significance. A value of 2 means that the measured value is 2 standard deviations away from zero. That means we're about 95% sure it exists. Higher is better.
Outside the US, it's very strong in the period 1992-2014, much stronger than the market premium.
https://papers.ssrn.com/sol3/papers.cfm ... id=2601662
BAB (betting against beta) is also quite strong in the US, period 1926-2012:
https://www.sciencedirect.com/science/a ... via%3Dihub
The t-stat in international markets (time period 1984-2012) is barely 2.0 which can be considered mediocre. They provide evidence for BAB in many different asset classes, all evidence taken together the effect appears extremely robust. One advantage of BAB is that it has a very plausible behavioral based explanation. A disadvantage about BAB is that it was discovered relatively recently and the out-of-sample evidence and academic scrutinizing has been limited so far. So far it appears the usual factor shops (read: DFA) are not yet willing to take the plunge. Since everybody seems to copy DFA these days, the fund selection is quite limited.
Then there is size (SmB). The situation with size is complicated: in 1991 it was claimed that SmB was weakly significant. Later, it appeared that some of the evidence was based on survivorship bias in the dataset. The current consensus appears to be that SmB does not have independent positive expected return and the effect is fully explained by CAPM. But oddly, if you control for the investment and profitability factors (Fama & French 4th and 5th factors), then SmB is suddenly very strong. I recommend reading this very readable paper for advanced scrutinizing of SmB: https://www.aqr.com/Insights/Research/J ... ize-Effect. I'm not sure if you should invest in SmB, but you definitely shouldn't do it with single-factor funds.
We also have various quality measures. From Fama/French we have investment (CmA) and profitability (RmW), from AQR research we have Quality (QmJ). They basically do the same thing. Quality is widely used in industry, but is almost never discussed here. I find AQR's literature on QmJ hard to digest.
Finally we have momentum, which is the most controversial factor. One one side you have the academics who claim the evidence for cross section momentum is extremely strong (and it is). On the other side you have the academics that claim momentum does not have a plausible risk-based explanation and is therefore unlikely to persist. Then you have the fund managers, half of which claim momentum is not exploitable after friction cost and half of which claim it is. But that's not all, there is also time series momentum which was claimed to be extremely strong across asset classes (personally, I've always found the to be evidence unconvincing), but there is a recent paper that says the statistical tests used to draw that conclusion are faulty. I won't be surprised if we see something similar for cross-sectional momentum.
Practically speaking, the most popular investment options are value, value + size, value + size + momentum, value + size + quality. Basically anything combined with value. I personally hold VFMF, which is a relatively new multifactor fund that targets value, size, quality, momentum and has a slight BAB screen.
The difficulty is not figuring out which premia are statistically significant, but which funds to select.
https://papers.ssrn.com/sol3/papers.cfm ... id=3525096
t-stat is a measure of statistical significance. A value of 2 means that the measured value is 2 standard deviations away from zero. That means we're about 95% sure it exists. Higher is better.
Outside the US, it's very strong in the period 1992-2014, much stronger than the market premium.
https://papers.ssrn.com/sol3/papers.cfm ... id=2601662
BAB (betting against beta) is also quite strong in the US, period 1926-2012:
https://www.sciencedirect.com/science/a ... via%3Dihub
The t-stat in international markets (time period 1984-2012) is barely 2.0 which can be considered mediocre. They provide evidence for BAB in many different asset classes, all evidence taken together the effect appears extremely robust. One advantage of BAB is that it has a very plausible behavioral based explanation. A disadvantage about BAB is that it was discovered relatively recently and the out-of-sample evidence and academic scrutinizing has been limited so far. So far it appears the usual factor shops (read: DFA) are not yet willing to take the plunge. Since everybody seems to copy DFA these days, the fund selection is quite limited.
Then there is size (SmB). The situation with size is complicated: in 1991 it was claimed that SmB was weakly significant. Later, it appeared that some of the evidence was based on survivorship bias in the dataset. The current consensus appears to be that SmB does not have independent positive expected return and the effect is fully explained by CAPM. But oddly, if you control for the investment and profitability factors (Fama & French 4th and 5th factors), then SmB is suddenly very strong. I recommend reading this very readable paper for advanced scrutinizing of SmB: https://www.aqr.com/Insights/Research/J ... ize-Effect. I'm not sure if you should invest in SmB, but you definitely shouldn't do it with single-factor funds.
We also have various quality measures. From Fama/French we have investment (CmA) and profitability (RmW), from AQR research we have Quality (QmJ). They basically do the same thing. Quality is widely used in industry, but is almost never discussed here. I find AQR's literature on QmJ hard to digest.
Finally we have momentum, which is the most controversial factor. One one side you have the academics who claim the evidence for cross section momentum is extremely strong (and it is). On the other side you have the academics that claim momentum does not have a plausible risk-based explanation and is therefore unlikely to persist. Then you have the fund managers, half of which claim momentum is not exploitable after friction cost and half of which claim it is. But that's not all, there is also time series momentum which was claimed to be extremely strong across asset classes (personally, I've always found the to be evidence unconvincing), but there is a recent paper that says the statistical tests used to draw that conclusion are faulty. I won't be surprised if we see something similar for cross-sectional momentum.
Practically speaking, the most popular investment options are value, value + size, value + size + momentum, value + size + quality. Basically anything combined with value. I personally hold VFMF, which is a relatively new multifactor fund that targets value, size, quality, momentum and has a slight BAB screen.
The difficulty is not figuring out which premia are statistically significant, but which funds to select.
Re: Most Trusted Risk Premia Factors?
That's assuming we're taking a random sample from a stable underlying distribution. Given that things change, including due to actions by market participants, that may not be the safest assumption.Uncorrelated wrote: ↑Wed Oct 21, 2020 5:54 amt-stat is a measure of statistical significance. A value of 2 means that the measured value is 2 standard deviations away from zero. That means we're about 95% sure it exists. Higher is better.
Re: Most Trusted Risk Premia Factors?
It seems that Size and Value work best when paired with Quality. You probably want to combine two or even three factors when screening for stocks. For example, it seemed that the Size premium by itself has gone away but returns with a vengeance with you paired it with Quality. Value worked better with Quality screens and when momentum was set to neutral. Problem is, all of this is known and lots of people are trying to exploit. So the screening that companies use to capture the factors keep getting tweaked.Register44 wrote: ↑Tue Oct 20, 2020 2:31 pm From what I have read here at bogleheads and elsewhere that there is evidence that small value may offer a persisting risk premium. Of course at this very moment is the time many may have given up that it doesn't "work" anymore.
I know quality, size, and momentum are other popular ones. I question though how they will hold up in the future. At least with value if the strategy fails you end up with a stock fund that isn't grossly overpriced as you would possibly with momentum.
Are there any known factors beyond Value that you would trust?
I have a Mid/Small-Cap tilt to my portfolio myself and plan to maintain those tilts. Sticking to a solid investment strategy is more important than picking among the solid investment strategies that are out there. As they say, every dog has his day.
Although I have teased Rick Ferri about backing away from Small/Value tilting strategies, I have come to the conclusion that new investors should start with the Broad Index funds for Stock and Bond Markets. The Broad Indexes would include the "Total" Indexes: Total Stock Market Index, Total Bond Market Index, Total International Stock Index, and Total International Bond Index. The S&P 500 for the U.S. Stock Market would fit my definition of a Broad Index.
The problem is that lots and lots of people know about the factors, the institutional investors are trying to capture them, and there is a lot of nuance in factor investing. There is more to it than just picking a Small Value mutual fund. Some practitioners are better at this than others, if you are buying Index products, some are better constructed than others.
I had a theory about the "anti-factors" which I define as the lottery stocks and the Value traps. My theory is that all I had to do was screen out the "anti-factors" and returns would increase. I am a big fan of the S&P Indexes as they do some screening for Quality, in other words, the truly junky companies just don't get in their indexes. So I went to Portfolio Visualizer and compared the US Total Stock Market Index with a market weight portfolio of the S&P 500, the S&P Mid-Cap 400, and the S&P Small-Cap 600. My thesis is that S&P would screen out the anti-factors that the Total Market didn't. I was right but by very little, it was a tiny fraction of 1%, probably statistically insignificant. So one conclusion I reached is that the markets are pretty efficient though not perfectly so.
My theory is that many of the truly junky stocks are not really very investable, small companies with very low trading volumes, companies that don't trade on the New York Stock Exchange or the NASDAQ. Lots of junk on the Vancouver Stock Exchange, the NASDAQ Bulletin Board, the pink sheets. This is probably what generates a lot of the premiums in the Academic Research. If you use indexes that utilize some sort of screening, particularly the S&P indexes, the junk just won't be there. There are probably 3,300 to 3,500 companies with large enough market cap and high enough trading volumes to be investable by the institutions. My guess is that probably 10,000 stocks are listed in the United States but 2/3 of them aren't worth considering.
A fool and his money are good for business.
-
- Posts: 1267
- Joined: Fri Apr 14, 2017 4:25 am
- Location: Israel
Re: Most Trusted Risk Premia Factors?
Which other factor is riskier that beta?Seasonal wrote: ↑Wed Oct 21, 2020 5:23 amThat's one of the major issues with risk premia. If something is really a risk premium, it's risky. You hope you'll be rewarded for taking the risk, but there's no guarantee - that's the idea of risk. Risk could include underperformance for a very very long time.Valuethinker wrote: ↑Wed Oct 21, 2020 3:29 amA lot of institutional money has gone into Low Vol since it was first published about. Tended to push up the stocks valuation.Forester wrote: ↑Wed Oct 21, 2020 3:05 amWorld Min Vol was down 15.4% in Q1 2020, versus 20.9% for the World Index. Globally defensive sectors did outperform cyclicals. The US market is more tech, tech had an extraordinary run this year, so US low vol stratgies were more muted compared to a prior selloff such as Q4 2018 or the 2015/16 correction.
Another issue, as Valuethinker points out, is that many investors tend to pile into these investments, leading to higher valuations which, in turn, lowers future returns.
It's also hard to know if something is just a statistical artifact. It may be a genuine risk factor or it may be a coincidence.
- patrick013
- Posts: 3301
- Joined: Mon Jul 13, 2015 7:49 pm
Re: Most Trusted Risk Premia Factors?
I still like quality factor like ticker SPHQ. High growth and lowNorthern Flicker wrote: ↑Tue Oct 20, 2020 11:17 pmMarket beta is the most robust and reliable equity risk premium factor.Register44 wrote: ↑Tue Oct 20, 2020 2:31 pm From what I have read here at bogleheads and elsewhere that there is evidence that small value may offer a persisting risk premium. Of course at this very moment is the time many may have given up that it doesn't "work" anymore.
I know quality, size, and momentum are other popular ones. I question though how they will hold up in the future. At least with value if the strategy fails you end up with a stock fund that isn't grossly overpriced as you would possibly with momentum.
Are there any known factors beyond Value that you would trust?
debt. Who else could survive a lengthy low growth period and
regain profitability in a normal business cycle ?
age in bonds, buy-and-hold, 10 year business cycle
Re: Most Trusted Risk Premia Factors?
I understand what you mean, unlike another poster who responded to this comment. However that is only when value underperforms due to valuations spreading between growth and value. It is possible value can underperform due to the growth companies simply growing their earnings a lot more than they expected to relative to the value companies. However I have read that value's recent underperformance is due to the first explanation, valuations widening.Register44 wrote: ↑Tue Oct 20, 2020 2:31 pm ... I question though how they will hold up in the future. At least with value if the strategy fails you end up with a stock fund that isn't grossly overpriced as you would possibly with momentum...
I'm just a fan of the person I got my user name from
- Uncorrelated
- Posts: 1237
- Joined: Sun Oct 13, 2019 3:16 pm
Re: Most Trusted Risk Premia Factors?
That is not a good reason to purchase a specific ETF. You are not the only one that knows high growth and low debt companies are attractive. You can only beat the market by taking risks other people find unattractive, or by being better in selecting the right stocks than other market participants. But unless you spend billions in market research, you can generally forget about the latter.patrick013 wrote: ↑Wed Oct 21, 2020 9:40 amI still like quality factor like ticker SPHQ. High growth and lowNorthern Flicker wrote: ↑Tue Oct 20, 2020 11:17 pmMarket beta is the most robust and reliable equity risk premium factor.Register44 wrote: ↑Tue Oct 20, 2020 2:31 pm From what I have read here at bogleheads and elsewhere that there is evidence that small value may offer a persisting risk premium. Of course at this very moment is the time many may have given up that it doesn't "work" anymore.
I know quality, size, and momentum are other popular ones. I question though how they will hold up in the future. At least with value if the strategy fails you end up with a stock fund that isn't grossly overpriced as you would possibly with momentum.
Are there any known factors beyond Value that you would trust?
debt. Who else could survive a lengthy low growth period and
regain profitability in a normal business cycle ?
- patrick013
- Posts: 3301
- Joined: Mon Jul 13, 2015 7:49 pm
Re: Most Trusted Risk Premia Factors?
I can't follow that advise based on factor statistics. I don't believeUncorrelated wrote: ↑Wed Oct 21, 2020 9:50 amThat is not a good reason to purchase a specific ETF. You are not the only one that knows high growth and low debt companies are attractive. You can only beat the market by taking risks other people find unattractive, or by being better in selecting the right stocks than other market participants. But unless you spend billions in market research, you can generally forget about the latter.patrick013 wrote: ↑Wed Oct 21, 2020 9:40 amI still like quality factor like ticker SPHQ. High growth and lowNorthern Flicker wrote: ↑Tue Oct 20, 2020 11:17 pmMarket beta is the most robust and reliable equity risk premium factor.Register44 wrote: ↑Tue Oct 20, 2020 2:31 pm
Are there any known factors beyond Value that you would trust?
debt. Who else could survive a lengthy low growth period and
regain profitability in a normal business cycle ?
in those. Way too shallow for my tastes.
With energy selling gas at half price, banks receiving trillion
dollar bailouts, tech overpriced, and COVID wrecking everything
else I like being my own advisor. If you snooze you lose, so
follow the market. I'm glad there is a quality factor index for
a defensive tilt available with market equilibrium being
way in the future.
age in bonds, buy-and-hold, 10 year business cycle
Re: Most Trusted Risk Premia Factors?
Define risk.Always passive wrote: ↑Wed Oct 21, 2020 7:41 amWhich other factor is riskier that beta?Seasonal wrote: ↑Wed Oct 21, 2020 5:23 amThat's one of the major issues with risk premia. If something is really a risk premium, it's risky. You hope you'll be rewarded for taking the risk, but there's no guarantee - that's the idea of risk. Risk could include underperformance for a very very long time.Valuethinker wrote: ↑Wed Oct 21, 2020 3:29 amA lot of institutional money has gone into Low Vol since it was first published about. Tended to push up the stocks valuation.Forester wrote: ↑Wed Oct 21, 2020 3:05 amWorld Min Vol was down 15.4% in Q1 2020, versus 20.9% for the World Index. Globally defensive sectors did outperform cyclicals. The US market is more tech, tech had an extraordinary run this year, so US low vol stratgies were more muted compared to a prior selloff such as Q4 2018 or the 2015/16 correction.
Another issue, as Valuethinker points out, is that many investors tend to pile into these investments, leading to higher valuations which, in turn, lowers future returns.
It's also hard to know if something is just a statistical artifact. It may be a genuine risk factor or it may be a coincidence.
Re: Most Trusted Risk Premia Factors?
This aspect of "trust" as the OP puts it is something I've never quite grokked with risk premia factors.
I've read literature (e.g., Swedroe's Black Swans book) demonstrating that a 1/N factor portfolio (where market is just one factor) would have had better risk-adjusted return than market alone in the past. There also appears to be a "risk" (if that's the appropriate word here) that certain factors may not persist and a sense that this "risk" is higher for some factors than others - i.e., the notion of "trust" in this thread. I haven't seen this "risk" modeled or quantified in any mathematically way. Shouldn't that impact how one allocates across premia factors?
I've read literature (e.g., Swedroe's Black Swans book) demonstrating that a 1/N factor portfolio (where market is just one factor) would have had better risk-adjusted return than market alone in the past. There also appears to be a "risk" (if that's the appropriate word here) that certain factors may not persist and a sense that this "risk" is higher for some factors than others - i.e., the notion of "trust" in this thread. I haven't seen this "risk" modeled or quantified in any mathematically way. Shouldn't that impact how one allocates across premia factors?
-
- Posts: 1267
- Joined: Fri Apr 14, 2017 4:25 am
- Location: Israel
Re: Most Trusted Risk Premia Factors?
It seems to me that if you purchase a fund or ETF, risk is volatility.Seasonal wrote: ↑Wed Oct 21, 2020 10:41 amDefine risk.Always passive wrote: ↑Wed Oct 21, 2020 7:41 amWhich other factor is riskier that beta?Seasonal wrote: ↑Wed Oct 21, 2020 5:23 amThat's one of the major issues with risk premia. If something is really a risk premium, it's risky. You hope you'll be rewarded for taking the risk, but there's no guarantee - that's the idea of risk. Risk could include underperformance for a very very long time.Valuethinker wrote: ↑Wed Oct 21, 2020 3:29 amA lot of institutional money has gone into Low Vol since it was first published about. Tended to push up the stocks valuation.Forester wrote: ↑Wed Oct 21, 2020 3:05 am
World Min Vol was down 15.4% in Q1 2020, versus 20.9% for the World Index. Globally defensive sectors did outperform cyclicals. The US market is more tech, tech had an extraordinary run this year, so US low vol stratgies were more muted compared to a prior selloff such as Q4 2018 or the 2015/16 correction.
Another issue, as Valuethinker points out, is that many investors tend to pile into these investments, leading to higher valuations which, in turn, lowers future returns.
It's also hard to know if something is just a statistical artifact. It may be a genuine risk factor or it may be a coincidence.
Re: Most Trusted Risk Premia Factors?
Then it should be easy to determine the riskiest factor. However, multifactor models are usually based on the idea that their specified factors are risk factors. For example, the Fama French three factor model labels small and value as risk factors. If you define risk as volatility and ignore other risk factors in a model that seems to do better, it will often appear that the factors in the better model have better risk adjusted returns than the market, but that's just a result of the definitions. A good explanation is in https://www.nber.org/papers/w7170Always passive wrote: ↑Wed Oct 21, 2020 10:45 amIt seems to me that if you purchase a fund or ETF, risk is volatility.Seasonal wrote: ↑Wed Oct 21, 2020 10:41 amDefine risk.Always passive wrote: ↑Wed Oct 21, 2020 7:41 amWhich other factor is riskier that beta?Seasonal wrote: ↑Wed Oct 21, 2020 5:23 amThat's one of the major issues with risk premia. If something is really a risk premium, it's risky. You hope you'll be rewarded for taking the risk, but there's no guarantee - that's the idea of risk. Risk could include underperformance for a very very long time.Valuethinker wrote: ↑Wed Oct 21, 2020 3:29 am
A lot of institutional money has gone into Low Vol since it was first published about. Tended to push up the stocks valuation.
Another issue, as Valuethinker points out, is that many investors tend to pile into these investments, leading to higher valuations which, in turn, lowers future returns.
It's also hard to know if something is just a statistical artifact. It may be a genuine risk factor or it may be a coincidence.
- Uncorrelated
- Posts: 1237
- Joined: Sun Oct 13, 2019 3:16 pm
Re: Most Trusted Risk Premia Factors?
That's because that risk can't be modeled.wickywack wrote: ↑Wed Oct 21, 2020 10:42 am This aspect of "trust" as the OP puts it is something I've never quite grokked with risk premia factors.
I've read literature (e.g., Swedroe's Black Swans book) demonstrating that a 1/N factor portfolio (where market is just one factor) would have had better risk-adjusted return than market alone in the past. There also appears to be a "risk" (if that's the appropriate word here) that certain factors may not persist and a sense that this "risk" is higher for some factors than others - i.e., the notion of "trust" in this thread. I haven't seen this "risk" modeled or quantified in any mathematically way. Shouldn't that impact how one allocates across premia factors?
A real factor needs four qualities: it needs to be investable, persistent across time, robust to alternative specifications, and there needs to be some clear reason it cannot be arbitraged away. The first three are easy: statistical tests and live funds provide the evidence that we need. The latter is more worrisome. The evidence for value is out there, why are smart people not investing in value, shrinking the premium? Why are high frequency traders not arbitraging momentum away?
I would rank momentum and size as low confidence, and value, quality and BAB as high confidence.