A Gut Punch (Cliff Asness of AQR on the performance of multifactor equity models)

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
HootingSloth
Posts: 1050
Joined: Mon Jan 28, 2019 2:38 pm

Re: A Gut Punch (Cliff Asness of AQR on the performance of multifactor equity models)

Post by HootingSloth »

dumbmoney wrote: Sat Dec 26, 2020 8:22 am
Seasonal wrote: Fri Dec 25, 2020 12:49 pm The question is how much exposure to each factor (source of risk) is appropriate.

The market / representative investor has decided on market cap proportions.

There's a long thread here based on Bill Sharpe's rationale for market proportions. viewtopic.php?t=207804

John Cochrane has a nice paper on the subject of living in a multi-factor world. It concludes "If you're pretty much average, all this thought will lead you right back to holding the market index. To rationalize anything but the market portfolio, you have to be different from the average investor in some
identifiable way." https://eml.berkeley.edu/~craine/EconH1 ... _world.pdf
If you manage to pin down a factor enthusiast, you find that they do believe that "everyone" should be overweight the factors (except beta). Which is inconsistent with saying that the factors are measures of systemic risk. One way to resolve the contradiction would be to say that beta alone is the best measure of systemic risk, but beta alone is not the best explanation of return.
Dumbmoney, for what it is worth, I don't think what you say is true. I might be considered a "factor enthusiast" in that I believe the academic literature on factors is getting at the truth, and I believe the portfolio that will maximize expected return at a given level of volatility is probably tilted towards factors like small and value. I emphatically do not believe that everyone should be overweight factors. For example, I do not tilt towards any factors and do not think that I should. I think that most people probably should not tilt towards factors.

Eugene Fama more or less invented factor investing, and he has repeatedly said that it is not (and cannot be) the case that everyone should tilt towards any given factor. Cochrane also forcefully argues this point in the exact article Seasonal linked to, and he clearly takes the evidence on factors very seriously.
Global Market Portfolio + modest tilt towards volatility (80/20->60/40 as approach FI) + modest tilt away from exchange rate risk (80% global+20% U.S. stocks; currency-hedge bonds) + tax optimization
User avatar
nedsaid
Posts: 19275
Joined: Fri Nov 23, 2012 11:33 am

Re: "Value will have its day."

Post by nedsaid »

Taylor Larimore wrote: Fri Dec 25, 2020 8:24 pm
nedsaid wrote: If you believe, like John Bogle did, that Value and Growth will do about the same over long periods of time, it seems that after such a long period of underperformance relative to Growth that at some point Value will have its day.
nedsaid:

I agree with you that "at some point Value will have its day." That's one of the problems: We don't know when?

Many of us (particularly retirees) cannot risk long periods of underperformance. Why take the chance and agony of underperformance when owning a simple, low-cost, total stock market index fund guarantees we will never underperform the 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. -- In my view, owning the market and holding it forever is the ultimate strategy for winners."
Gosh Taylor, the S&P 500 and the US Total Stock Market was flat from 2000-2012 and during that time these indexes were down 50% twice. These indexes had negative performance during the 2000's. You had been retired for a while back then and somehow you dealt with it. The fact is that good investment approaches have their day and each have time periods when things don't go so well. One just has to have patience. Even the 3 fund approach has had its tough patches but things work themselves out over time. Any good investment approach should work out well over time, there are a few I can think of off the top of my head.

The whole reason that the broad US Indexes have outperformed Value might just boil down to a relative few Tech stocks particularly the FAANG stocks. 6-10 big Tech stocks have made you the smartest man in the world. Give that a thought.
Last edited by nedsaid on Sat Dec 26, 2020 11:55 am, edited 1 time in total.
A fool and his money are good for business.
User avatar
nedsaid
Posts: 19275
Joined: Fri Nov 23, 2012 11:33 am

Re: A Gut Punch (Cliff Asness of AQR on the performance of multifactor equity models)

Post by nedsaid »

bck63 wrote: Fri Dec 25, 2020 4:58 pm
nedsaid wrote: Fri Dec 25, 2020 1:43 pm Even if factor premiums have gone away, you still have reversion to the mean. If you believe, like John Bogle did, that Value and Growth will do about the same over long periods of time, it seems that after such a long period of underperformance relative to Growth that at some point Value will have its day.
So then is it enough to just hold a blended fund of growth and value (total market, S&P 500)? You get whatever is outperforming at the moment. Does one need to tilt in order to benefit from eventual value outperformance?

I'm not referring to SCV, just to growth and value as you mentioned.
The entire reason that an investor would factor tilt his or her portfolio, particularly with Small Value stocks, is the belief that such a factor tilted portfolio would outperform Total Stock Market Index and the S&P 500 over long periods of time. If you perfectly balance your stocks between Growth and Value, this also perfectly cancels out any premium that you hope to achieve. During a time when Value outperforms Growth, your Value premium could more more than cancelled out by the drag created by Growth. Factor tilting is a long term approach.

Factors reflect investor preferences over long periods of time. It is sort of like watching a beauty pageant. It doesn't matter who you think is the prettiest and the most talented, it matters what the judges think. Factors reflect what the judges have liked best over long periods of time. Another way of saying this is that academics have noticed that some types of stocks behave differently than other types of stocks.
A fool and his money are good for business.
User avatar
nedsaid
Posts: 19275
Joined: Fri Nov 23, 2012 11:33 am

Re: A Gut Punch (Cliff Asness of AQR on the performance of multifactor equity models)

Post by nedsaid »

rkhusky wrote: Sat Dec 26, 2020 7:42 am
nedsaid wrote: Fri Dec 25, 2020 1:33 pm
All Seasons wrote: Fri Dec 25, 2020 12:29 pm Because factor portfolios are long-short portfolios.

The market portfolio has no exposure to momentum for instance because all the positive momentum is exactly cancelled out by the negative momentum. The same is true for all the other factors.

The market portfolio has exposure to just one factor: beta.

Merry Christmas everyone.
I want to point out that you can be a long only factor investor, it is just that in theory you can capture more of the premium with a long/short strategy.
That is confusing terminology. Factors were designed to be approximately independent via the long-short structure. Value by itself is not independent of the market and neither is small by itself.

If you define long-only as being a factor, then TSM contains all factors. But if factors are confined to be long-short, then TSM has approximately zero loading to factors other than the market.

Admittedly though, a long-only value fund will have non-zero loading on the value factor, but so will a long-only growth fund.
I am NOT defining long only as a factor. Just saying that you can capture factor premiums with a long only strategy. In theory, a long/short strategy should better capture factor premiums better than a long only strategy but in practice, this hasn't always worked out. Momentum is a great example of why long/short could work better than long only because Momentum can be both on the way up and on the way down. You could in theory best capture the Value premium by being long the stocks with the best Value characteristics and short the stocks with the worst Value characteristics.

One reason that the long/short strategy hasn't always worked is that factors seem to work better when you combine two or three. Quality helps both Size and Value. Momentum set to neutral helps Value. Also lots of debate as to the best proportional combining of factors. What combination works the best? There is complexity with a long only strategy when you combine factors but the complexity is increased when you introduce shorting and then leverage.

The difficulty in executing these strategies can perhaps be explained by too many people chasing too few factor stocks or by saying that the markets are becoming more efficient. A counter argument can be made that a long term Growth trend, we are now 11 years into this, will make it appear that the factors have gone away. Time will tell who is correct.
A fool and his money are good for business.
rkhusky
Posts: 17764
Joined: Thu Aug 18, 2011 8:09 pm

Re: A Gut Punch (Cliff Asness of AQR on the performance of multifactor equity models)

Post by rkhusky »

nedsaid wrote: Sat Dec 26, 2020 11:50 am A counter argument can be made that a long term Growth trend, we are now 11 years into this, will make it appear that the factors have gone away. Time will tell who is correct.
Factor models do not favor either the short or long portion of the long-short structure. And premiums can be both positive and negative. Growth is just as much a part of factor investing as value. Factors will have gone away when growth and value have the same return, and big and small, and high and low momentum, etc., i.e. when market beta alone accurately predicts stock returns.
User avatar
nedsaid
Posts: 19275
Joined: Fri Nov 23, 2012 11:33 am

Re: A Gut Punch (Cliff Asness of AQR on the performance of multifactor equity models)

Post by nedsaid »

rkhusky wrote: Sat Dec 26, 2020 12:05 pm
nedsaid wrote: Sat Dec 26, 2020 11:50 am A counter argument can be made that a long term Growth trend, we are now 11 years into this, will make it appear that the factors have gone away. Time will tell who is correct.
Factor models do not favor either the short or long portion of the long-short structure. And premiums can be both positive and negative. Growth is just as much a part of factor investing as value. Factors will have gone away when growth and value have the same return, and big and small, and high and low momentum, etc., i.e. when market beta alone accurately predicts stock returns.
Growth by itself has not been identified as a factor by the academics. There are at least two factors related to Growth that have been identified: Quality/Profitability and Momentum.
A fool and his money are good for business.
rkhusky
Posts: 17764
Joined: Thu Aug 18, 2011 8:09 pm

Re: A Gut Punch (Cliff Asness of AQR on the performance of multifactor equity models)

Post by rkhusky »

nedsaid wrote: Sat Dec 26, 2020 12:07 pm
rkhusky wrote: Sat Dec 26, 2020 12:05 pm
nedsaid wrote: Sat Dec 26, 2020 11:50 am A counter argument can be made that a long term Growth trend, we are now 11 years into this, will make it appear that the factors have gone away. Time will tell who is correct.
Factor models do not favor either the short or long portion of the long-short structure. And premiums can be both positive and negative. Growth is just as much a part of factor investing as value. Factors will have gone away when growth and value have the same return, and big and small, and high and low momentum, etc., i.e. when market beta alone accurately predicts stock returns.
Growth by itself has not been identified as a factor by the academics. There are at least two factors related to Growth that have been identified: Quality/Profitability and Momentum.
Value is not a factor either. Factors are long-short. A value fund will have positive value factor premium. A growth fund will have negative value factor premium. The signs are simply a convention and could be the opposite with a different equivalent definition of the value factor.

The reason that some like to invest in value stocks (or small stocks) has nothing to do with factor models, but with data-mining past returns. Factor models do not tell you that small will outperform big or that value will outperform growth.
Last edited by rkhusky on Sat Dec 26, 2020 1:39 pm, edited 1 time in total.
Gadget
Posts: 1026
Joined: Fri Mar 17, 2017 1:38 pm

Re: A Gut Punch (Cliff Asness of AQR on the performance of multifactor equity models)

Post by Gadget »

nedsaid wrote: Wed Dec 23, 2020 8:37 pm ...But then again, I was born to be mild.
Sounds like we found a bogleheads theme song.
User avatar
nisiprius
Advisory Board
Posts: 52212
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: A Gut Punch (Cliff Asness of AQR on the performance of multifactor equity models)

Post by nisiprius »

nedsaid wrote: Sat Dec 26, 2020 11:50 amOne reason that the long/short strategy hasn't always worked is that factors seem to work better when you combine two or three. Quality helps both Size and Value. Momentum set to neutral helps Value...
In my flip opinion, because the more factors you have, the better the opportunity for overfitting. If size doesn't "work" by itself, try small + value. If small + value doesn't work, try small + value - utilities. If that doesn't work, try small + quality. Or small + value + quality. Or small + momentum - junk. Or value + momentum + BAB.

If you have five factors to play with, you don't just have five possibilities. If either one of them can be for or against or ignore, then you have 242 combinations to look at. And then you can increase things even more by debating what school you belong to in defining or naming the factors. If value isn't "working," maybe you're dividing price by the Wrong Thing. And momentum versus time series momentum.

Vanguard needed to create nine index funds to take care of just two factors: large growth, large blend, large value, mid-cap growth, mid-cap blend, mid-cap value, small-cap growth, small-cap blend, small-cap value. Imagine how many they would need to have to offer every plausible multifactor combination.
Last edited by nisiprius on Sat Dec 26, 2020 12:39 pm, edited 1 time in total.
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.
tibbitts
Posts: 23716
Joined: Tue Feb 27, 2007 5:50 pm

Re: A Gut Punch (Cliff Asness of AQR on the performance of multifactor equity models)

Post by tibbitts »

bck63 wrote: Sat Dec 26, 2020 10:36 am
Exchme wrote: Sat Dec 26, 2020 8:17 am
bck63 wrote: Fri Dec 25, 2020 7:41 am Given the abysmal returns and ridiculous fees of AQR funds, I'm not really interested in "Cliff's Perspective".
What we all have to realize is that the article is actually an advertisement and the mea culpa is persuasion. I'm sure he is hemorrhaging customers, so being an excellent entrepreneur, he doesn't sit and lick his wounds, he writes about how awful it's been. In the course of the article, he gets to sound smart and explain why his super-secret formula "should" have worked. Roping in a new batch of suckers.
This Clifford dude was actually invited onto a Bogleheads podcast, thus allowing him yet another vehicle to, as you so wisely say, rope in a new batch of suckers.

Shameful, IMHO.
Why is it shameful? Only because he charges "too much" for his funds, or because you disagree with there being any benefit to factor analysis? If the latter, would other security selection processes be acceptable for podcast guests? Or would you have banned Bogle too, based on his anonymous FAJ article? How about Graham, or other active management proponents? Siegel, in light of Wisdom Tree? Who would be acceptable to you?
marcopolo
Posts: 8445
Joined: Sat Dec 03, 2016 9:22 am

Re: A Gut Punch (Cliff Asness of AQR on the performance of multifactor equity models)

Post by marcopolo »

nisiprius wrote: Sat Dec 26, 2020 12:37 pm
nedsaid wrote: Sat Dec 26, 2020 11:50 amOne reason that the long/short strategy hasn't always worked is that factors seem to work better when you combine two or three. Quality helps both Size and Value. Momentum set to neutral helps Value...
In my flip opinion, because the more factors you have, the better the opportunity for overfitting. If size doesn't "work" by itself, try small + value. If small + value doesn't work, try small + value - utilities. If that doesn't work, try small + quality. Or small + value + quality. Or small + momentum - junk. Or value + momentum + BAB.

If you have five factors to play with, you don't just have five possibilities. If either one of them can be for or against or ignore, then you have 242 combinations to look at. And then you can increase things even more by debating what school you belong to in defining or naming the factors. If value isn't "working," maybe you're dividing price by the Wrong Thing. And momentum versus time series momentum.

Vanguard needed to create nine index funds to take care of just two factors: large growth, large blend, large value, mid-cap growth, mid-cap blend, mid-cap value, small-cap growth, small-cap blend, small-cap value. Imagine how many they would need to have to offer every plausible multifactor combination.
Then you backtest those combinations and find the one that "works", then claim it was obvious all along that this was the right combination to use. QED!
Once in a while you get shown the light, in the strangest of places if you look at it right.
User avatar
nedsaid
Posts: 19275
Joined: Fri Nov 23, 2012 11:33 am

Re: A Gut Punch (Cliff Asness of AQR on the performance of multifactor equity models)

Post by nedsaid »

nisiprius wrote: Sat Dec 26, 2020 12:37 pm
nedsaid wrote: Sat Dec 26, 2020 11:50 amOne reason that the long/short strategy hasn't always worked is that factors seem to work better when you combine two or three. Quality helps both Size and Value. Momentum set to neutral helps Value...
In my flip opinion, because the more factors you have, the better the opportunity for overfitting. If size doesn't "work" by itself, try small + value. If small + value doesn't work, try small + value - utilities. If that doesn't work, try small + quality. Or small + value + quality. Or small + momentum - junk. Or value + momentum + BAB.

If you have five factors to play with, you don't just have five possibilities. If either one of them can be for or against or ignore, then you have 242 combinations to look at. And then you can increase things even more by debating what school you belong to in defining or naming the factors. If value isn't "working," maybe you're dividing price by the Wrong Thing. And momentum versus time series momentum.

Vanguard needed to create nine index funds to take care of just two factors: large growth, large blend, large value, mid-cap growth, mid-cap blend, mid-cap value, small-cap growth, small-cap blend, small-cap value. Imagine how many they would need to have to offer every plausible multifactor combination.
You actually have an infinite range of possibilities with factor investing. Not only do you choose what factors to combine but in what proportions.

This is why I say that at some point good enough is good enough and close enough is close enough. For my Small Value investments, my number one holding is the Vanguard Small Cap Value Index ETF (VBR), which is based on the CRSP Small-Cap Value Index. Imperfect has it has a lot of Mid-Caps and a lot of Core Stocks. As one poster said, VBR has moderate loadings on Size and Value or as I put it, sort of Value and sort of Small. Not perfect but plenty well good enough. Number two holding is the iShares S&P 600 Small-Cap Value Index (IJS) which loads better on Size and Value than the CRSP Small Cap Value Index.

If you are interested in getting a better mix of combining factors, let the professionals do it. Avantis offers factor ETFs that are Value oriented but they are a bit coy here but they also combine with Quality.

So for a do-it-yourself Small Value tilters, buy VBR for moderate Size and Value loads, buy IJR for deeper Size and Value loads with a touch of Quality, if you want sometime to optimize the factors for you one could try Avantis Small Cap Value ETF. Depending on how the markets act, any one of these three could work best in a Small Value rebound. I have commented many times regarding the "good" factor products and the "bad" factor products and sometimes the "bad" outperforms the "good".

Pretty much, I am saying not to get too fancy or too cute with this. Even within the Small Value universe of investment products, hard to say what will do best in the future. Just pick a good product, cross your fingers, say a prayer, and hope it all works for the best. An educated and informed guess is better than an uneducated and uninformed guess.
A fool and his money are good for business.
User avatar
nedsaid
Posts: 19275
Joined: Fri Nov 23, 2012 11:33 am

Re: A Gut Punch (Cliff Asness of AQR on the performance of multifactor equity models)

Post by nedsaid »

rkhusky wrote: Sat Dec 26, 2020 12:21 pm
nedsaid wrote: Sat Dec 26, 2020 12:07 pm
rkhusky wrote: Sat Dec 26, 2020 12:05 pm
nedsaid wrote: Sat Dec 26, 2020 11:50 am A counter argument can be made that a long term Growth trend, we are now 11 years into this, will make it appear that the factors have gone away. Time will tell who is correct.
Factor models do not favor either the short or long portion of the long-short structure. And premiums can be both positive and negative. Growth is just as much a part of factor investing as value. Factors will have gone away when growth and value have the same return, and big and small, and high and low momentum, etc., i.e. when market beta alone accurately predicts stock returns.
Growth by itself has not been identified as a factor by the academics. There are at least two factors related to Growth that have been identified: Quality/Profitability and Momentum.
Value is not a factor either. Factors are long-short. A value fund will have positive value factor premium. A growth fund will have negative value factor premium. The signs are simply a convention and could be the opposite with a different equivalent definition of the value factor.

The reason that some like to invest in value stocks (or small stocks) has nothing to do with factors models, but with data-mining past returns. Factors models do not tell you that small will outperform big or that value will outperform growth.
We have to have agreement on what words mean. You are playing semantic games here.

There are literally hundreds of factors out there but there is broad agreement over Market, Size, Value, Quality/Profitability, Momentum, and Low Volatility. Nothing I have read says that Growth is a factor and that Value is not. Nothing I have read says that Long/Short or Long Only are factors, they might better be termed as techniques.

Long/short is contained within the definition of a factor, the purest definition of the Value factor would include being long the stocks with the best Value characteristics and also short the stocks with the worst Value characteristics.

In practice most factor tilters here, execute a long only version for various reasons, number one being cost. There are also limitations to shorting, which works best for the larger and most liquid stocks. When you try to short Small stocks, you deal with lower trading volume and higher bid/ask spreads. You also have to pay the actual owners of the stocks their dividends. When the spreads and other expenses from shorting get large enough, this wipes out whatever premium you might get. I think Larry Swedroe calls this the limits of arbitrage.

So if you buy a Small Value product, it will be almost certainly be a long only product as this is not a good environment for shorting. You also run into the limitation that the Value premium works best with Small Caps.

If you wanted to do Long/Short on Momentum, the best environment would be in the Large Growth area of the market. The MTUM ETF is long only and you see it operate in that Large Growth stylebox.
A fool and his money are good for business.
User avatar
nedsaid
Posts: 19275
Joined: Fri Nov 23, 2012 11:33 am

Re: A Gut Punch (Cliff Asness of AQR on the performance of multifactor equity models)

Post by nedsaid »

marcopolo wrote: Sat Dec 26, 2020 12:58 pm
nisiprius wrote: Sat Dec 26, 2020 12:37 pm
nedsaid wrote: Sat Dec 26, 2020 11:50 amOne reason that the long/short strategy hasn't always worked is that factors seem to work better when you combine two or three. Quality helps both Size and Value. Momentum set to neutral helps Value...
In my flip opinion, because the more factors you have, the better the opportunity for overfitting. If size doesn't "work" by itself, try small + value. If small + value doesn't work, try small + value - utilities. If that doesn't work, try small + quality. Or small + value + quality. Or small + momentum - junk. Or value + momentum + BAB.

If you have five factors to play with, you don't just have five possibilities. If either one of them can be for or against or ignore, then you have 242 combinations to look at. And then you can increase things even more by debating what school you belong to in defining or naming the factors. If value isn't "working," maybe you're dividing price by the Wrong Thing. And momentum versus time series momentum.

Vanguard needed to create nine index funds to take care of just two factors: large growth, large blend, large value, mid-cap growth, mid-cap blend, mid-cap value, small-cap growth, small-cap blend, small-cap value. Imagine how many they would need to have to offer every plausible multifactor combination.
Then you backtest those combinations and find the one that "works", then claim it was obvious all along that this was the right combination to use. QED!
We look to the past as a guide as to what might happen in the future. This even extends to the Equity Risk Premium. We invest in stocks because historically they have better returns than bonds. We buy stocks even though there is no guarantee that the Equity Risk Premium will continue into the future.

No one saying they can guarantee what the best factor combinations are in the future. It is an educated guess just like that it is an educated guess that stocks will outperform bonds in the future.
A fool and his money are good for business.
User avatar
nedsaid
Posts: 19275
Joined: Fri Nov 23, 2012 11:33 am

Re: A Gut Punch (Cliff Asness of AQR on the performance of multifactor equity models)

Post by nedsaid »

vineviz wrote: Thu Dec 24, 2020 10:41 am
djm2001 wrote: Thu Dec 24, 2020 9:27 am
vineviz wrote: Thu Dec 24, 2020 8:44 am There can be no doubt that a portfolio exposed to multiple independent sources of systematic risk is more diversified than a portfolio which is exposed to only one.
I'm confused about something, and hope you can clear it up. I've heard people mention "independent" sources of risk as you just did, but then elsewhere I also hear people say that risk factors (market, value, size, etc.) are not uncorrelated. But if they are correlated, then they can't be independent, right? Am I misunderstanding something fundamental? Am I mixing up different senses of the word "independent"? Or did you just mean "not perfectly correlated" instead of "independent"? Not trying to be pedantic; just trying to resolve a long-held confusion I've had.
If two risk factors had a correlation coefficient of zero they would be completely independent. If the correlation is 1.0 or -1.0 they are not at all independent.

Very few sources of financial risk are COMPLETELY independent but most equity risk factors have cross-correlations closer to zero than to +1 or -1. So if your asset pricing model uses (say) six factors, your total number of independent sources of risk is likely >1 but <6: depending on which factors you’re using, the maximum effective number of independent sources of risk is typically between 3 and 5.

The constraints on long-only portfolio construction means that the effective limit is often closer to two or three, but that’s still a huge improvement over one.
Vineviz has a great response here. I have attempted four factors other than Market but in reality probably two have made a difference. If I had to guess, my factor loadings in order of importance are 1) Market, 2) Size, 3) Value, 4) Quality, and 5) Momentum. As my aggressive growth funds have mostly abandoned an Accelerating Earnings or Earnings Momentum strategy, number 5 has decreased over time. Despite all of the talk of factor tilting, we overwhelmingly rely on Market. I have a very sneaking suspicion that most of us have much less factor loading than we think we have. For example, my Value tilts are actually pretty mild.
A fool and his money are good for business.
Random Walker
Posts: 5561
Joined: Fri Feb 23, 2007 7:21 pm

Re: A Gut Punch (Cliff Asness of AQR on the performance of multifactor equity models)

Post by Random Walker »

nedsaid wrote: Sat Dec 26, 2020 2:49 pm Despite all of the talk of factor tilting, we overwhelmingly rely on Market. I have a very sneaking suspicion that most of us have much less factor loading than we think we have. For example, my Value tilts are actually pretty mild.
I agree. Most all of our portfolios are dominated by market beta. Even a 60/40 portfolio has 85-90% of its risk wrapped up in market beta. Most long only SV funds have a market beta of about 1. And the VG SV funds tend to have modest factor exposures, so the investor using them needs to take on additional market risk to get the factor exposures he desires. This is why I believe that in an effort to create more efficient portfolios, most of us need to focus on diversifying away from the market factor. This is also why a more expensive factor fund with deeper exposures could be worth the cost; the investor needs to buy less of it to get the tilts he wants and he takes on less additional market risk to get the tilts he wants.

Dave

PS And we should likely overwhelmingly rely on market. It makes sense to invest in the factors proportional to ones belief in them. Most all of us likely have the greatest faith in the market factor. (Yet look at the t-stat mentioned way above in this thread for value!!)
Last edited by Random Walker on Sat Dec 26, 2020 3:02 pm, edited 1 time in total.
GoneOnTilt
Posts: 1841
Joined: Fri Sep 28, 2018 4:59 pm

Re: A Gut Punch (Cliff Asness of AQR on the performance of multifactor equity models)

Post by GoneOnTilt »

nedsaid wrote: Sat Dec 26, 2020 11:38 am
bck63 wrote: Fri Dec 25, 2020 4:58 pm
nedsaid wrote: Fri Dec 25, 2020 1:43 pm Even if factor premiums have gone away, you still have reversion to the mean. If you believe, like John Bogle did, that Value and Growth will do about the same over long periods of time, it seems that after such a long period of underperformance relative to Growth that at some point Value will have its day.
So then is it enough to just hold a blended fund of growth and value (total market, S&P 500)? You get whatever is outperforming at the moment. Does one need to tilt in order to benefit from eventual value outperformance?

I'm not referring to SCV, just to growth and value as you mentioned.
The entire reason that an investor would factor tilt his or her portfolio, particularly with Small Value stocks, is the belief that such a factor tilted portfolio would outperform Total Stock Market Index and the S&P 500 over long periods of time. If you perfectly balance your stocks between Growth and Value, this also perfectly cancels out any premium that you hope to achieve. During a time when Value outperforms Growth, your Value premium could more more than cancelled out by the drag created by Growth. Factor tilting is a long term approach.

Factors reflect investor preferences over long periods of time. It is sort of like watching a beauty pageant. It doesn't matter who you think is the prettiest and the most talented, it matters what the judges think. Factors reflect what the judges have liked best over long periods of time. Another way of saying this is that academics have noticed that some types of stocks behave differently than other types of stocks.
Thank you. Very helpful.
User avatar
mrspock
Posts: 2158
Joined: Tue Feb 13, 2018 1:49 am
Location: Vulcan

Re: A Gut Punch (Cliff Asness of AQR on the performance of multifactor equity models)

Post by mrspock »

IMO factor investing is another form of active investing without calling it by name, as you need to actively “pick” your tilts. Because of the need to make a bunch of fairly complex choices, similar challenges and human behavioral flaws apply here as stock picking.

The clever packaging by the industry here is two fold: first you are “indexing” still because they are baskets of funds based on the factors (uhh ok), and second that you shouldn’t compare yourself to the S&P, but of a similarly tilted portfolio. Which is just this generation of asset managers way of doing the cry of “Oh, I never promised to beat the S&P... I was trying to <some nonsense>.”

But that’s never the claim on the outset is it? It’s that you will eek out 0.5% or 1% extra CAGR by implementing all these tilts. And then when they don’t materialize they move the goal posts. Same old tricks.

Somebody prove me wrong here? If 80% of active managers underperform the S&P, what are the stats for asset managers who employ tilting strategies? My guess is the stats are not pretty.
Last edited by mrspock on Sat Dec 26, 2020 4:01 pm, edited 1 time in total.
User avatar
nedsaid
Posts: 19275
Joined: Fri Nov 23, 2012 11:33 am

Re: A Gut Punch (Cliff Asness of AQR on the performance of multifactor equity models)

Post by nedsaid »

Random Walker wrote: Sat Dec 26, 2020 2:59 pm
nedsaid wrote: Sat Dec 26, 2020 2:49 pm Despite all of the talk of factor tilting, we overwhelmingly rely on Market. I have a very sneaking suspicion that most of us have much less factor loading than we think we have. For example, my Value tilts are actually pretty mild.
I agree. Most all of our portfolios are dominated by market beta. Even a 60/40 portfolio has 85-90% of its risk wrapped up in market beta. Most long only SV funds have a market beta of about 1. And the VG SV funds tend to have modest factor exposures, so the investor using them needs to take on additional market risk to get the factor exposures he desires. This is why I believe that in an effort to create more efficient portfolios, most of us need to focus on diversifying away from the market factor. This is also why a more expensive factor fund with deeper exposures could be worth the cost; the investor needs to buy less of it to get the tilts he wants and he takes on less additional market risk to get the tilts he wants.

Dave

PS And we should likely overwhelmingly rely on market. It makes sense to invest in the factors proportional to ones belief in them. Most all of us likely have the greatest faith in the market factor. (Yet look at the t-stat mentioned way above in this thread for value!!)
Here is an example using my individual stocks. I did a 10 year period from 12/25/2010 - 12/25/2020 and used the Fama-French 3 factor model at Portfolio Visualizer. Used the Factor Regression Analysis tool.

Market 1.09
Size -0.19
Value 0.27
Annualized Alpha 1.36%
Market and Size not a surprise as my average market cap is 1.02 relative to the S&P 500.

Compare this to Vanguard Value Index Admiral (VVIAX)

Market 0.91
Size -0.14
Value 0.28
Annualized Alpha -0.13%

Now Compare to DFA US Large Cap Value I (DFLVX)

Market 1.06
Size 0.01
Value 0.41
Annualized Alpha -1.40%

So I had about the same Value loading as Vanguard Value Index and much less than DFA US Large Cap Value I.
A fool and his money are good for business.
User avatar
nedsaid
Posts: 19275
Joined: Fri Nov 23, 2012 11:33 am

Re: A Gut Punch (Cliff Asness of AQR on the performance of multifactor equity models)

Post by nedsaid »

mrspock wrote: Sat Dec 26, 2020 3:28 pm IMO factor investing is another form of active investing without calling it by name, as you need to actively “pick” your tilts. Because of the need to make a bunch of fairly complex choices, similar challenges and human behavioral flaws apply here as stock picking.

The clever packaging by the industry here is two fold: first you are “indexing” still because they are baskets of funds based on the factors (uhh ok), and second that you shouldn’t compare yourself to the S&P, but of a similarly tilted portfolio. Which is just this generation of asset managers way of doing the cry of “Oh, I never promised to beat the S&P... I was trying to <some nonsense>.”

But that’s never the claim on the outset is it? It’s that you will eek out 0.5% or 1% extra CAGR by implementing all these tilts. And then when they don’t materialize they move the goal posts. Same old tricks.

Somebody prove me wrong here? It 80% of active managers underperform the S&P, what are the stats for asset managers who employ tilting strategies? My guess is the stats are not pretty.
I argued with Larry Swedroe about this very point. Factor investing actually is a form of stock picking but best practitioners have pretty low turnover, so I don't really call it active. I place the factor funds like DFA, Avantis, Bridgeway, etc. on a continuum between passive and active, I argued that these factor funds were more passive than active. Larry would say that it is passive because you buy all of the stocks within the universe but a screen was set and the computer selected the stocks. It just seemed that it is active if humans pick stocks and passive if a computer does the same thing. One could say it is active investing with low turnover and more stocks.

I am a believer in the continuum. These factor funds would be more passive than active but not really passive in my view. They should do better than active managers vs. Total Stock Market Index mainly because of costs. There are lower cost and low cost ETFs that do factors pretty well.
A fool and his money are good for business.
GoneOnTilt
Posts: 1841
Joined: Fri Sep 28, 2018 4:59 pm

Re: A Gut Punch (Cliff Asness of AQR on the performance of multifactor equity models)

Post by GoneOnTilt »

tibbitts wrote: Sat Dec 26, 2020 12:38 pm
bck63 wrote: Sat Dec 26, 2020 10:36 am
Exchme wrote: Sat Dec 26, 2020 8:17 am
bck63 wrote: Fri Dec 25, 2020 7:41 am Given the abysmal returns and ridiculous fees of AQR funds, I'm not really interested in "Cliff's Perspective".
What we all have to realize is that the article is actually an advertisement and the mea culpa is persuasion. I'm sure he is hemorrhaging customers, so being an excellent entrepreneur, he doesn't sit and lick his wounds, he writes about how awful it's been. In the course of the article, he gets to sound smart and explain why his super-secret formula "should" have worked. Roping in a new batch of suckers.
This Clifford dude was actually invited onto a Bogleheads podcast, thus allowing him yet another vehicle to, as you so wisely say, rope in a new batch of suckers.

Shameful, IMHO.
Why is it shameful? Only because he charges "too much" for his funds, or because you disagree with there being any benefit to factor analysis? If the latter, would other security selection processes be acceptable for podcast guests? Or would you have banned Bogle too, based on his anonymous FAJ article? How about Graham, or other active management proponents? Siegel, in light of Wisdom Tree? Who would be acceptable to you?
I've climbed off the soapbox and gave it some thought. Your point is taken. I guess a podcast purely about low-fee indexing would get boring pretty fast. People can hear information about anything and come to their own conclusions. That being said, I wouldn't touch his horribly performing funds, with their ridiculously-high fees, with a ten foot pole. In my view, he's ripping people off. That is my personal opinion. Others may disagree. I'll take the market performance, and a 0.02 average ER any day.
Last edited by GoneOnTilt on Sat Dec 26, 2020 4:10 pm, edited 1 time in total.
User avatar
nisiprius
Advisory Board
Posts: 52212
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: A Gut Punch (Cliff Asness of AQR on the performance of multifactor equity models)

Post by nisiprius »

nedsaid wrote: Sat Dec 26, 2020 3:45 pm
mrspock wrote: Sat Dec 26, 2020 3:28 pmIMO factor investing is another form of active investing without calling it by name, as you need to actively “pick” your tilts...
I argued with Larry Swedroe about this very point. Factor investing actually is a form of stock picking but best practitioners have pretty low turnover, so I don't really call it active. I place the factor funds like DFA, Avantis, Bridgeway, etc. on a continuum between passive and active...
I agree about the continuum.

But it is not an extreme view to regard factor investing as active.

Vanguard factor-based funds

Image
Vanguard factor-based funds

...Factor-based funds are a form of active management. They offer the potential to achieve specific risk and return objectives by purposely and explicitly "tilting" portfolios toward certain stock characteristics, like recent momentum, higher quality, or lower stock prices.
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.
User avatar
nedsaid
Posts: 19275
Joined: Fri Nov 23, 2012 11:33 am

Re: A Gut Punch (Cliff Asness of AQR on the performance of multifactor equity models)

Post by nedsaid »

Well, even the S&P indexes have a committee than decides which stocks to include in their indexes. S&P also has criteria for companies to have their stocks listed in their indexes. So even the S&P 500 uses light screening and human input. But I wouldn't argue that the S&P Indexes are a form of active management, they are plenty well passive enough.

This also gets into the territory of Index construction. Some indexes are better than others. At Bogleheads, we don't stock pick, we index pick. Even Vanguard has switched indexes for some of its index funds. Companies will sidestep licensing fees with outfits like S&P by coming up with their own proprietary indexes.

So we get into arguments about which indexes are the best. Again, good enough is good enough and close enough is close enough.

Meanwhile back at the ranch, I wouldn't consider what AQR does as passive. Not quite active but much more active than passive.
A fool and his money are good for business.
marcopolo
Posts: 8445
Joined: Sat Dec 03, 2016 9:22 am

Re: A Gut Punch (Cliff Asness of AQR on the performance of multifactor equity models)

Post by marcopolo »

nedsaid wrote: Sat Dec 26, 2020 1:43 pm
marcopolo wrote: Sat Dec 26, 2020 12:58 pm
nisiprius wrote: Sat Dec 26, 2020 12:37 pm
nedsaid wrote: Sat Dec 26, 2020 11:50 amOne reason that the long/short strategy hasn't always worked is that factors seem to work better when you combine two or three. Quality helps both Size and Value. Momentum set to neutral helps Value...
In my flip opinion, because the more factors you have, the better the opportunity for overfitting. If size doesn't "work" by itself, try small + value. If small + value doesn't work, try small + value - utilities. If that doesn't work, try small + quality. Or small + value + quality. Or small + momentum - junk. Or value + momentum + BAB.

If you have five factors to play with, you don't just have five possibilities. If either one of them can be for or against or ignore, then you have 242 combinations to look at. And then you can increase things even more by debating what school you belong to in defining or naming the factors. If value isn't "working," maybe you're dividing price by the Wrong Thing. And momentum versus time series momentum.

Vanguard needed to create nine index funds to take care of just two factors: large growth, large blend, large value, mid-cap growth, mid-cap blend, mid-cap value, small-cap growth, small-cap blend, small-cap value. Imagine how many they would need to have to offer every plausible multifactor combination.
Then you backtest those combinations and find the one that "works", then claim it was obvious all along that this was the right combination to use. QED!
We look to the past as a guide as to what might happen in the future. This even extends to the Equity Risk Premium. We invest in stocks because historically they have better returns than bonds. We buy stocks even though there is no guarantee that the Equity Risk Premium will continue into the future.

No one saying they can guarantee what the best factor combinations are in the future. It is an educated guess just like that it is an educated guess that stocks will outperform bonds in the future.

Surely, you realize the difference between testing a single hypothesis that equities outperform bonds vs. data mining hundreds of possible combinations of factors and selecting a small number of combinations that "work" and the statistical significance between the two cases being vastly different?
Once in a while you get shown the light, in the strangest of places if you look at it right.
User avatar
nedsaid
Posts: 19275
Joined: Fri Nov 23, 2012 11:33 am

Re: A Gut Punch (Cliff Asness of AQR on the performance of multifactor equity models)

Post by nedsaid »

marcopolo wrote: Sat Dec 26, 2020 4:58 pm
nedsaid wrote: Sat Dec 26, 2020 1:43 pm
marcopolo wrote: Sat Dec 26, 2020 12:58 pm
nisiprius wrote: Sat Dec 26, 2020 12:37 pm
nedsaid wrote: Sat Dec 26, 2020 11:50 amOne reason that the long/short strategy hasn't always worked is that factors seem to work better when you combine two or three. Quality helps both Size and Value. Momentum set to neutral helps Value...
In my flip opinion, because the more factors you have, the better the opportunity for overfitting. If size doesn't "work" by itself, try small + value. If small + value doesn't work, try small + value - utilities. If that doesn't work, try small + quality. Or small + value + quality. Or small + momentum - junk. Or value + momentum + BAB.

If you have five factors to play with, you don't just have five possibilities. If either one of them can be for or against or ignore, then you have 242 combinations to look at. And then you can increase things even more by debating what school you belong to in defining or naming the factors. If value isn't "working," maybe you're dividing price by the Wrong Thing. And momentum versus time series momentum.

Vanguard needed to create nine index funds to take care of just two factors: large growth, large blend, large value, mid-cap growth, mid-cap blend, mid-cap value, small-cap growth, small-cap blend, small-cap value. Imagine how many they would need to have to offer every plausible multifactor combination.
Then you backtest those combinations and find the one that "works", then claim it was obvious all along that this was the right combination to use. QED!
We look to the past as a guide as to what might happen in the future. This even extends to the Equity Risk Premium. We invest in stocks because historically they have better returns than bonds. We buy stocks even though there is no guarantee that the Equity Risk Premium will continue into the future.

No one saying they can guarantee what the best factor combinations are in the future. It is an educated guess just like that it is an educated guess that stocks will outperform bonds in the future.

Surely, you realize the difference between testing a single hypothesis that equities outperform bonds vs. data mining hundreds of possible combinations of factors and selecting a small number of combinations that "work" and the statistical significance between the two cases being vastly different?
The comparisons that I make are not perfect, just making the point that in both cases we are using the past as a guide as to what might happen in the future. There are criticisms of Academic Research, I think that to toss out the phrase "data mining" isn't fair. In an extreme example, I could look at data and find 5 stocks that would have made everyone millionaires if they only purchased them 10 years ago. I don't believe this is what the researchers are doing, their factor strategies cover many years of data collected from all over the world. Also factor strategies are a whole lot broader than picking the 5 stocks that would have made everyone rich. The academics are looking at increasing an investor's chance of success, it isn't a snake oil get rich quick scheme.

There are many reasons to question the data and comparability between eras, I have discussed these issues in depth here. Also arguments that factors are time period dependent. This has been debated on many threads.

The academics still believe in the Size and Value premiums but have tweaked their approach as they learned more. They have tweaked but not abandoned what they had recommended in the beginning.

I first learned about factors and the Academic research back in 2007 at a Paul Merriman seminar. They claimed that their strategy would outperform a similar Vanguard Index portfolio by 1% a year after fees. They also said that in a Growth environment that Vanguard would outperform them even with Vanguard portfolios using Vanguard factor funds. They showed how you could execute their strategies at providers like Fidelity, Vanguard, Schwab, and T. Rowe Price. What they said was that DFA funds were better at capturing the factors than other products because of smaller market caps and better Value characteristics. You can say what you want about Paul Merriman, he was a businessman trying to build his business and I believed he did things in an honorable way. I adopted many of their suggestions but I did not sign up with their firm.

After doing the research on this, I concluded that I would settle for a 0.50% premium per year after expenses and I was hoping for a 1% premium. This wasn't double your money in 60 days.
A fool and his money are good for business.
User avatar
typical.investor
Posts: 5263
Joined: Mon Jun 11, 2018 3:17 am

Re: "The Emperor Has No Clothes"

Post by typical.investor »

nisiprius wrote: Fri Dec 25, 2020 12:14 pm
Taylor Larimore wrote: Fri Dec 25, 2020 11:29 am Bogleheads:

I've never understood why many fund salesmen categorize "beta" (the total stock market) a single "factor" the same as a small-cap stocks factor, value factors, momentum stock factors, etc., which are ALREADY in the total stock market.

Anyone care to explain?
Best wishes of the season to all!

It actually does make sense as far as it goes, Taylor. I don't do factor investing but there isn't any logical impossibility. So forgive me for an analogy.

All food has calories. That is the basic food "factor," analogous to the market factor, it's what prevents starvation.

Protein, carbohydrates, and fat are like other "factors." They all are food, they all have calories, but they are not identical in all of their other properties.
This analogy is extremely misleading I think, and should be disregarded for a proper understanding.

In fact, a Total Food Market would really contain Protein, carbohydrates, and fat. It contains all the nutrients.

Of course, factor proponents would have you believe that you don't have any protein in your diet unless you have concentration your meals to be proportionally more protein and proportionally less carbohydrates.

So if protein and carbohydrates each provide the same proportion of calories in a Total Food Market, factor proponents would say you don't have exposure to protein unless you tilt your diet and reduce the amount of carbohydrates.

That is simply factually incorrect. Reducing your proportion of carbohydrates does not create a new form of nutrition.

Ok, but in your universe I guess the Total Stock Market doesn't include value stocks. It's an interesting theory but one that firmly exists only in a reality defying alternative universe.

[Yes, I do tilt toward value stocks because I think ultimately the difficulty they present will result in a return. It's non-sense though to insist that a Total Market Investor doesn't have value stock exposure simply because they don't have an overweighting in it. Seems like in 2020 a major challenge is for people to recognize non-sense when they hear it]
marcopolo
Posts: 8445
Joined: Sat Dec 03, 2016 9:22 am

Re: A Gut Punch (Cliff Asness of AQR on the performance of multifactor equity models)

Post by marcopolo »

nedsaid wrote: Sat Dec 26, 2020 5:24 pm
...
The academics are looking at increasing an investor's chance of success, it isn't a snake oil get rich quick scheme.
....
I am not criticizing the academics, they are doing research based on historical data (that is all we have!), and making interesting observations.
The vast majority of them include the appropriate caveats around their conclusions. Shiller has discussed the data mining risk, it is my understanding that Fama and French have said there research really should not be a call to invest accordingly. So, I agree they are not doing to get rich in most cases.

I don't think i have ever used the the term "snake oil" or "get rich quick" to describe the fund managers that then implement and market funds based on these strategies. That would probably be going a bit too far. But, lets not kid ourselves that they are not exaggerating the benefits and down playing the caveats to market these strategies to retail investors in order to get rich. They almost always do significantly better than the investors in their products. Maybe that is just a coincidence, but it is certainly not some altruistic academic endeavor on their part.
Once in a while you get shown the light, in the strangest of places if you look at it right.
User avatar
nedsaid
Posts: 19275
Joined: Fri Nov 23, 2012 11:33 am

Re: A Gut Punch (Cliff Asness of AQR on the performance of multifactor equity models)

Post by nedsaid »

marcopolo wrote: Sat Dec 26, 2020 5:56 pm
nedsaid wrote: Sat Dec 26, 2020 5:24 pm
...
The academics are looking at increasing an investor's chance of success, it isn't a snake oil get rich quick scheme.
....
I am not criticizing the academics, they are doing research based on historical data (that is all we have!), and making interesting observations.
The vast majority of them include the appropriate caveats around their conclusions. Shiller has discussed the data mining risk, it is my understanding that Fama and French have said there research really should not be a call to invest accordingly. So, I agree they are not doing to get rich in most cases.

I don't think i have ever used the the term "snake oil" or "get rich quick" to describe the fund managers that then implement and market funds based on these strategies. That would probably be going a bit too far. But, lets not kid ourselves that they are not exaggerating the benefits and down playing the caveats to market these strategies to retail investors in order to get rich. They almost always do significantly better than the investors in their products. Maybe that is just a coincidence, but it is certainly not some altruistic academic endeavor on their part.
There are flaws to everything, one of which is that it is difficult to take bias out of research. We all have the tendency to see what we want to see when looking at data. The human brain wants to see patterns even when those patterns might not exist. So we should have a natural skepticism. But even skepticism can be taken to far, to the point where we don't believe anything.

Certain academics, including Burton Malkiel and Jeremy Seigel have cashed in. The founding of DFA was an act of turning academic ideas into a business. This is capitalism, if done right this should be win/win for everyone involved. But we know that nothing is perfect, even Vanguard. But certainly some people are more reliable sources of information than others and some investment providers are better for individual investors than others. But all have a profit motive.

So it is right and proper to debate these things and hopefully everyone learns. What I would say is that there is a slipperiness to financial data but we live in an imperfect world. Having said that what we have as individual investors is awfully good. Just today, I was at Morningstar and Portfolio Visualizer doing analysis. What I have today is vastly superior to the Stock Page in the newspaper and having to make the trek to the Public Library. We just forget how good we have it.

I don't invest with AQR but the point is that we have choices, and lots of them. We can each invest our own money as we see fit and using the best information that is available to us.

There is big money in distribution and lots of people have gotten rich distributing to the masses goods and services. Certain societies have a huge problem with this and thus have a hard time advancing as envy keeps easy distribution away from their public. Envy becomes a rather self defeating behavior. I would rather have an entrepreneur with a profit motive making a buck from distributing things to me than relying upon corrupt criminal enterprises to get what I need. If my convenience makes Jeff Bezos, Bill Gates, and others rich so be it. Gosh, Michael Bloomberg got rich from distributing financial information through his famous Bloomberg terminals.

Plenty of choices out there, if folks don't like AQR, there are many places to go. I even see lots of complaints regarding Vanguard. There is Fidelity, Schwab, and others. But that is the American Way.
A fool and his money are good for business.
marcopolo
Posts: 8445
Joined: Sat Dec 03, 2016 9:22 am

Re: A Gut Punch (Cliff Asness of AQR on the performance of multifactor equity models)

Post by marcopolo »

nedsaid wrote: Sat Dec 26, 2020 6:34 pm
marcopolo wrote: Sat Dec 26, 2020 5:56 pm
nedsaid wrote: Sat Dec 26, 2020 5:24 pm
...
The academics are looking at increasing an investor's chance of success, it isn't a snake oil get rich quick scheme.
....
I am not criticizing the academics, they are doing research based on historical data (that is all we have!), and making interesting observations.
The vast majority of them include the appropriate caveats around their conclusions. Shiller has discussed the data mining risk, it is my understanding that Fama and French have said there research really should not be a call to invest accordingly. So, I agree they are not doing to get rich in most cases.

I don't think i have ever used the the term "snake oil" or "get rich quick" to describe the fund managers that then implement and market funds based on these strategies. That would probably be going a bit too far. But, lets not kid ourselves that they are not exaggerating the benefits and down playing the caveats to market these strategies to retail investors in order to get rich. They almost always do significantly better than the investors in their products. Maybe that is just a coincidence, but it is certainly not some altruistic academic endeavor on their part.
There are flaws to everything, one of which is that it is difficult to take bias out of research. We all have the tendency to see what we want to see when looking at data. The human brain wants to see patterns even when those patterns might not exist. So we should have a natural skepticism. But even skepticism can be taken to far, to the point where we don't believe anything.

Certain academics, including Burton Malkiel and Jeremy Seigel have cashed in. The founding of DFA was an act of turning academic ideas into a business. This is capitalism, if done right this should be win/win for everyone involved. But we know that nothing is perfect, even Vanguard. But certainly some people are more reliable sources of information than others and some investment providers are better for individual investors than others. But all have a profit motive.

So it is right and proper to debate these things and hopefully everyone learns. What I would say is that there is a slipperiness to financial data but we live in an imperfect world. Having said that what we have as individual investors is awfully good. Just today, I was at Morningstar and Portfolio Visualizer doing analysis. What I have today is vastly superior to the Stock Page in the newspaper and having to make the trek to the Public Library. We just forget how good we have it.

I don't invest with AQR but the point is that we have choices, and lots of them. We can each invest our own money as we see fit and using the best information that is available to us.

There is big money in distribution and lots of people have gotten rich distributing to the masses goods and services. Certain societies have a huge problem with this and thus have a hard time advancing as envy keeps easy distribution away from their public. Envy becomes a rather self defeating behavior. I would rather have an entrepreneur with a profit motive making a buck from distributing things to me than relying upon corrupt criminal enterprises to get what I need. If my convenience makes Jeff Bezos, Bill Gates, and others rich so be it. Gosh, Michael Bloomberg got rich from distributing financial information through his famous Bloomberg terminals.

Plenty of choices out there, if folks don't like AQR, there are many places to go. I even see lots of complaints regarding Vanguard. There is Fidelity, Schwab, and others. But that is the American Way.
I agree with every bit of that.
I mistakenly thought you were saying the firms pushing these strategies were not doing it to get rich.
Once in a while you get shown the light, in the strangest of places if you look at it right.
User avatar
vineviz
Posts: 14921
Joined: Tue May 15, 2018 1:55 pm
Location: Baltimore, MD

Re: "The Emperor Has No Clothes"

Post by vineviz »

typical.investor wrote: Sat Dec 26, 2020 5:36 pm
Of course, factor proponents would have you believe that you don't have any protein in your diet unless you have concentration your meals to be proportionally more protein and proportionally less carbohydrates.

So if protein and carbohydrates each provide the same proportion of calories in a Total Food Market, factor proponents would say you don't have exposure to protein unless you tilt your diet and reduce the amount of carbohydrates.

That is simply factually incorrect. Reducing your proportion of carbohydrates does not create a new form of nutrition.

Ok, but in your universe I guess the Total Stock Market doesn't include value stocks. It's an interesting theory but one that firmly exists only in a reality defying alternative universe.
Are you intentionally misrepresenting the analogy, or are you just momentarily confused?

Obviously the TSM fund includes value STOCKS: it must, since it includes all stocks.

What it does not include is exposure to the value FACTOR. So-called “factor proponents” don’t typically get as confused on this point as others do.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
000
Posts: 8211
Joined: Thu Jul 23, 2020 12:04 am

Re: A Gut Punch (Cliff Asness of AQR on the performance of multifactor equity models)

Post by 000 »

I think true factorheads would just buy individual stocks to get whatever precise exposures they desire.
User avatar
typical.investor
Posts: 5263
Joined: Mon Jun 11, 2018 3:17 am

Re: "The Emperor Has No Clothes"

Post by typical.investor »

vineviz wrote: Sat Dec 26, 2020 7:19 pm
typical.investor wrote: Sat Dec 26, 2020 5:36 pm
Ok, but in your universe I guess the Total Stock Market doesn't include value stocks. It's an interesting theory but one that firmly exists only in a reality defying alternative universe.
Are you intentionally misrepresenting the analogy, or are you just momentarily confused.

Obviously the TSM fund includes value STOCKS: it must, since it includes all stocks.

What it does not include is exposure to the value FACTOR.
I'd counter that the analogy stating Total (food) Market doesn't contain protein is what's misleading. I did no misrepresenting in identifying the flaw of suggesting that the Total Market is analogous to a Total Food Market that is simply calories without the nutrition of protein.

The value FACTOR is simply an overweight to the same value stocks that the total stock market already contains.

Whether of not value will outperform in your holding period is a bet that some want to take. Forcefully advocating that one is not diversified unless you do so is an always will be pure rubbish.
User avatar
vineviz
Posts: 14921
Joined: Tue May 15, 2018 1:55 pm
Location: Baltimore, MD

Re: "The Emperor Has No Clothes"

Post by vineviz »

typical.investor wrote: Sat Dec 26, 2020 7:31 pm
vineviz wrote: Sat Dec 26, 2020 7:19 pm
typical.investor wrote: Sat Dec 26, 2020 5:36 pm
Ok, but in your universe I guess the Total Stock Market doesn't include value stocks. It's an interesting theory but one that firmly exists only in a reality defying alternative universe.
Are you intentionally misrepresenting the analogy, or are you just momentarily confused.

Obviously the TSM fund includes value STOCKS: it must, since it includes all stocks.

What it does not include is exposure to the value FACTOR.
I'd counter that the analogy stating Total (food) Market doesn't contain protein is what's misleading.
I do not see any evidence that anyone said “the total food market doesn’t contain protein”.
typical.investor wrote: Sat Dec 26, 2020 7:31 pm Forcefully advocating that one is not diversified unless you do so is an always will be pure rubbish.
Again, you’re objecting to something that literally no one has said.

Clearly a TSM fund is “well diversified”. It’s just that an investor can easily build a stock portfolio which is MORE diversified by including exposure to more than a single factor.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
Random Walker
Posts: 5561
Joined: Fri Feb 23, 2007 7:21 pm

Re: A Gut Punch (Cliff Asness of AQR on the performance of multifactor equity models)

Post by Random Walker »

000 wrote: Sat Dec 26, 2020 7:28 pm I think true factorheads would just buy individual stocks to get whatever precise exposures they desire.
No, just the opposite. A true Factorhead is interested in diversification. This is where a Factorhead and a TSMer agree: both want to eliminate idiosyncratic single stock risk. A Factorhead wants to go beyond that level of diversification and diversify across unique and independent sources of risk and return. A Factorhead is focused on diversifying across risks that are compensated. A single stock has the same expected return as its asset class but a whole lot more uncompensated idiosyncratic risk than its associated asset class.

Dave
000
Posts: 8211
Joined: Thu Jul 23, 2020 12:04 am

Re: A Gut Punch (Cliff Asness of AQR on the performance of multifactor equity models)

Post by 000 »

Random Walker wrote: Sat Dec 26, 2020 7:42 pm
000 wrote: Sat Dec 26, 2020 7:28 pm I think true factorheads would just buy individual stocks to get whatever precise exposures they desire.
No, just the opposite. A true Factorhead is interested in diversification. This is where a Factorhead and a TSMer agree: both want to eliminate idiosyncratic single stock risk. A Factorhead wants to go beyond that level of diversification and diversify across unique and independent sources of risk and return. A Factorhead is focused on diversifying across risks that are compensated. A single stock has the same expected return as its asset class but a whole lot more uncompensated idiosyncratic risk than its associated asset class.

Dave
Isn't this also true (same degree, different magnitude) of picking subsets of stocks though?
Random Walker
Posts: 5561
Joined: Fri Feb 23, 2007 7:21 pm

Re: A Gut Punch (Cliff Asness of AQR on the performance of multifactor equity models)

Post by Random Walker »

000 wrote: Sat Dec 26, 2020 7:47 pm
Random Walker wrote: Sat Dec 26, 2020 7:42 pm
000 wrote: Sat Dec 26, 2020 7:28 pm I think true factorheads would just buy individual stocks to get whatever precise exposures they desire.
No, just the opposite. A true Factorhead is interested in diversification. This is where a Factorhead and a TSMer agree: both want to eliminate idiosyncratic single stock risk. A Factorhead wants to go beyond that level of diversification and diversify across unique and independent sources of risk and return. A Factorhead is focused on diversifying across risks that are compensated. A single stock has the same expected return as its asset class but a whole lot more uncompensated idiosyncratic risk than its associated asset class.

Dave
Isn't this also true (same degree, different magnitude) of picking subsets of stocks though?
To some extent I have to agree with you, yes. But I believe the dominant goal of the TSM or S&P 500 investor is to eliminate single stock risk. If one has 25 stocks in a portfolio, the potential contribution of any one stock might be 4%. If one has 100 stocks in a portfolio, the potential contribution of any one stock might be 1%. By the time we are talking 500-1000 stocks in a SV fund, 500 stocks in S&P 500, or 3000 stocks in Total Stock Market fund, single stock risk is effectively eliminated in any of them.
If one is looking at diversification in terms of the drivers of stock returns, then he views a TSM fund and a SV fund as both having an exposure to the market factor of about 1, and whether that exposure is through 1000 or 3000 stocks may not be that relevant. Of course I write this having semi-painfully not owned my share of FANG over the last decade :-)

Dave
User avatar
typical.investor
Posts: 5263
Joined: Mon Jun 11, 2018 3:17 am

Re: A Gut Punch (Cliff Asness of AQR on the performance of multifactor equity models)

Post by typical.investor »

Random Walker wrote: Sat Dec 26, 2020 7:42 pm
000 wrote: Sat Dec 26, 2020 7:28 pm I think true factorheads would just buy individual stocks to get whatever precise exposures they desire.
No, just the opposite. A true Factorhead is interested in diversification. This is where a Factorhead and a TSMer agree: both want to eliminate idiosyncratic single stock risk. A Factorhead wants to go beyond that level of diversification and diversify across unique and independent sources of risk and return. A Factorhead is focused on diversifying across risks that are compensated.
Dave
I'm sorry Dave but factor investing isn't really diversify risk.

Wes Grey has a good description of risk factors I think. Some will take the risk and earn an insurance premium and some will pay a premium to avoid the risk. He also points out that sometimes risks show up and those holding the risks have to basically payout a claim (ie. factor underperformance).

Total market investors are already diversified in that they hold both sides. "Diversifying" across risk factors may be safer than concentrating on only one, but it's still less diversified that Total Market which both issues and holds the insurance policy. It doesn't matter if these risks show up and demand an insurance payment or oppositely if the policy actually returns a premium because you hold both.

There is good reason take factor risks I believe, but the real risk is a period like Asness has just noted where all the risk factors are demanding the insurance company pay out a claim at the same time. Unless that situation changes in your holding period, you will have lost money that a total market investor didn't because they were more diversified (and held the insurance policy too).

Anyway, I encourage you to hold your factor tilts and believe they will pay off for you (assuming they can overcome the cost). Abandoning tilts in down times is not a winning strategy.

The simple truth is that factor investing will leave you more exposed to the risks that others are forgoing a premium to avoid. Historically that has worked out over long periods and I hope it continues to be true for both our sake.
Random Walker
Posts: 5561
Joined: Fri Feb 23, 2007 7:21 pm

Re: A Gut Punch (Cliff Asness of AQR on the performance of multifactor equity models)

Post by Random Walker »

typical.investor wrote: Sat Dec 26, 2020 8:34 pm There is good reason take factor risks I believe, but the real risk is a period like Asness has just noted where all the risk factors are demanding the insurance company pay out a claim at the same time. Unless that situation changes in your holding period, you will have lost money that a total market investor didn't because they were more diversified (and held the insurance policy too).

Anyway, I encourage you to hold your factor tilts and believe they will pay off for you (assuming they can overcome the cost). Abandoning tilts in down times is not a winning strategy.

The simple truth is that factor investing will leave you more exposed to the risks that others are forgoing a premium to avoid. Historically that has worked out over long periods and I hope it continues to be true for both our sake.
But there are long periods where market factor underperforms TBills. There have been times where other factors have lessened the pain of market underperformance. If the market does well, a SV fund may underperform the market, but it will do pretty well in its own right since it is dominated by market beta. The below article from Andrew Berkin of Bridgeway may interest you.

https://bridgeway.com/perspectives/stre ... ecessions/


Important to consider the expected return of a portfolio. Need to make apples to apples comparisons. The investor tilted to SV should increase the exposure to safe bonds relative to the TSM investor. In that scenario, the factor investor is more diversified across risks. I’d like to read the Wes Grey description of insurance premiums. Do you have a link? Thanks,

Dave
User avatar
jeffyscott
Posts: 13485
Joined: Tue Feb 27, 2007 8:12 am

Re: A Gut Punch (Cliff Asness of AQR on the performance of multifactor equity models)

Post by jeffyscott »

Random Walker wrote: Sat Dec 26, 2020 8:04 pm
000 wrote: Sat Dec 26, 2020 7:47 pm
Random Walker wrote: Sat Dec 26, 2020 7:42 pm
000 wrote: Sat Dec 26, 2020 7:28 pm I think true factorheads would just buy individual stocks to get whatever precise exposures they desire.
No, just the opposite. A true Factorhead is interested in diversification. This is where a Factorhead and a TSMer agree: both want to eliminate idiosyncratic single stock risk. A Factorhead wants to go beyond that level of diversification and diversify across unique and independent sources of risk and return. A Factorhead is focused on diversifying across risks that are compensated. A single stock has the same expected return as its asset class but a whole lot more uncompensated idiosyncratic risk than its associated asset class.

Dave
Isn't this also true (same degree, different magnitude) of picking subsets of stocks though?
To some extent I have to agree with you, yes. But I believe the dominant goal of the TSM or S&P 500 investor is to eliminate single stock risk. If one has 25 stocks in a portfolio, the potential contribution of any one stock might be 4%. If one has 100 stocks in a portfolio, the potential contribution of any one stock might be 1%. By the time we are talking 500-1000 stocks in a SV fund, 500 stocks in S&P 500, or 3000 stocks in Total Stock Market fund, single stock risk is effectively eliminated in any of them.
If one is looking at diversification in terms of the drivers of stock returns, then he views a TSM fund and a SV fund as both having an exposure to the market factor of about 1, and whether that exposure is through 1000 or 3000 stocks may not be that relevant. Of course I write this having semi-painfully not owned my share of FANG over the last decade :-)

Dave
Rather than FAANG, substitute Microsoft for Netflix and FAAMG would be about 18%. That's about 18% of holdings devoted to just 5 companies, if one holds only US TSM. To me, that makes what used to be called slicing and dicing more diversified than TSM.
User avatar
HomerJ
Posts: 21281
Joined: Fri Jun 06, 2008 12:50 pm

Re: A Gut Punch (Cliff Asness of AQR on the performance of multifactor equity models)

Post by HomerJ »

tibbitts wrote: Sat Dec 26, 2020 12:38 pm
bck63 wrote: Sat Dec 26, 2020 10:36 am
Exchme wrote: Sat Dec 26, 2020 8:17 am
bck63 wrote: Fri Dec 25, 2020 7:41 am Given the abysmal returns and ridiculous fees of AQR funds, I'm not really interested in "Cliff's Perspective".
What we all have to realize is that the article is actually an advertisement and the mea culpa is persuasion. I'm sure he is hemorrhaging customers, so being an excellent entrepreneur, he doesn't sit and lick his wounds, he writes about how awful it's been. In the course of the article, he gets to sound smart and explain why his super-secret formula "should" have worked. Roping in a new batch of suckers.
This Clifford dude was actually invited onto a Bogleheads podcast, thus allowing him yet another vehicle to, as you so wisely say, rope in a new batch of suckers.

Shameful, IMHO.
Why is it shameful? Only because he charges "too much" for his funds, or because you disagree with there being any benefit to factor analysis? If the latter, would other security selection processes be acceptable for podcast guests? Or would you have banned Bogle too, based on his anonymous FAJ article? How about Graham, or other active management proponents? Siegel, in light of Wisdom Tree? Who would be acceptable to you?
The man made $3 billion plus. In a few years... He wasn't on the Forbes list, and then suddenly he was... with $3 billion...

The customers who invested with him LOST money.

Where are the customer's yachts?

Of course that's shameful.

I don't know how anyone can respect Cliff Asness.

Jack Bogle never made that kind of money. And he could have... But he didn't. He gave US the money.

If you make $3 billion, and the people who trust you make money as well, fine... or if you provide a product or a service for that money, fine.

If you make $3 billion, and everyone who invested with you loses money, but you still make $3 billion, you are a horrible human being. $3 billion is an insane amount of money to make, when you don't even provide a service that works.

Bill Gates, Steve Jobs, Jeff Bezos, fine... We all actually got something for the money they made.

Nobody got ANYTHING from Cliff Asness... Pretty much all his investors lost money. Their money got transferred to him.

It's terrible.
Last edited by HomerJ on Sat Dec 26, 2020 9:58 pm, edited 2 times in total.
"The best tools available to us are shovels, not scalpels. Don't get carried away." - vanBogle59
User avatar
HomerJ
Posts: 21281
Joined: Fri Jun 06, 2008 12:50 pm

Re: A Gut Punch (Cliff Asness of AQR on the performance of multifactor equity models)

Post by HomerJ »

nedsaid wrote: Sat Dec 26, 2020 1:22 pm
nisiprius wrote: Sat Dec 26, 2020 12:37 pm
nedsaid wrote: Sat Dec 26, 2020 11:50 amOne reason that the long/short strategy hasn't always worked is that factors seem to work better when you combine two or three. Quality helps both Size and Value. Momentum set to neutral helps Value...
In my flip opinion, because the more factors you have, the better the opportunity for overfitting. If size doesn't "work" by itself, try small + value. If small + value doesn't work, try small + value - utilities. If that doesn't work, try small + quality. Or small + value + quality. Or small + momentum - junk. Or value + momentum + BAB.

If you have five factors to play with, you don't just have five possibilities. If either one of them can be for or against or ignore, then you have 242 combinations to look at. And then you can increase things even more by debating what school you belong to in defining or naming the factors. If value isn't "working," maybe you're dividing price by the Wrong Thing. And momentum versus time series momentum.

Vanguard needed to create nine index funds to take care of just two factors: large growth, large blend, large value, mid-cap growth, mid-cap blend, mid-cap value, small-cap growth, small-cap blend, small-cap value. Imagine how many they would need to have to offer every plausible multifactor combination.
You actually have an infinite range of possibilities with factor investing. Not only do you choose what factors to combine but in what proportions.

This is why I say that at some point good enough is good enough and close enough is close enough. For my Small Value investments, my number one holding is the Vanguard Small Cap Value Index ETF (VBR), which is based on the CRSP Small-Cap Value Index. Imperfect has it has a lot of Mid-Caps and a lot of Core Stocks. As one poster said, VBR has moderate loadings on Size and Value or as I put it, sort of Value and sort of Small. Not perfect but plenty well good enough. Number two holding is the iShares S&P 600 Small-Cap Value Index (IJS) which loads better on Size and Value than the CRSP Small Cap Value Index.

If you are interested in getting a better mix of combining factors, let the professionals do it. Avantis offers factor ETFs that are Value oriented but they are a bit coy here but they also combine with Quality.

So for a do-it-yourself Small Value tilters, buy VBR for moderate Size and Value loads, buy IJR for deeper Size and Value loads with a touch of Quality, if you want sometime to optimize the factors for you one could try Avantis Small Cap Value ETF. Depending on how the markets act, any one of these three could work best in a Small Value rebound. I have commented many times regarding the "good" factor products and the "bad" factor products and sometimes the "bad" outperforms the "good".

Pretty much, I am saying not to get too fancy or too cute with this. Even within the Small Value universe of investment products, hard to say what will do best in the future. Just pick a good product, cross your fingers, say a prayer, and hope it all works for the best. An educated and informed guess is better than an uneducated and uninformed guess.
Just invest in stocks and bonds, and quit messing around. Quit trying to do better.

Rich is rich.
Last edited by HomerJ on Sat Dec 26, 2020 9:52 pm, edited 1 time in total.
"The best tools available to us are shovels, not scalpels. Don't get carried away." - vanBogle59
User avatar
HomerJ
Posts: 21281
Joined: Fri Jun 06, 2008 12:50 pm

Re: A Gut Punch (Cliff Asness of AQR on the performance of multifactor equity models)

Post by HomerJ »

nedsaid wrote: Sat Dec 26, 2020 5:24 pmThe academics are looking at increasing an investor's chance of success, it isn't a snake oil get rich quick scheme.
The "academics" have been proven wrong again and again.

They are not evil, they are not snake oil salesmen.

But man are they deluded and misguided.

Economics isn't physics.

There aren't enough data points, there are a hundred variables, and even the "constants" change occasionally.

You think the academics can figure out the odds? They can't.

It's like if you saw 100 hands of blackjack without knowing the rules, and had to figure out the odds from the hands you saw.

Oh, AND the rules change every 10-20 years/hands.

It's impossible. Stop believing the "academics"
"The best tools available to us are shovels, not scalpels. Don't get carried away." - vanBogle59
Normchad
Posts: 5648
Joined: Thu Mar 03, 2011 6:20 am

Re: A Gut Punch (Cliff Asness of AQR on the performance of multifactor equity models)

Post by Normchad »

HomerJ wrote: Sat Dec 26, 2020 9:45 pm
tibbitts wrote: Sat Dec 26, 2020 12:38 pm
bck63 wrote: Sat Dec 26, 2020 10:36 am
Exchme wrote: Sat Dec 26, 2020 8:17 am
bck63 wrote: Fri Dec 25, 2020 7:41 am Given the abysmal returns and ridiculous fees of AQR funds, I'm not really interested in "Cliff's Perspective".
What we all have to realize is that the article is actually an advertisement and the mea culpa is persuasion. I'm sure he is hemorrhaging customers, so being an excellent entrepreneur, he doesn't sit and lick his wounds, he writes about how awful it's been. In the course of the article, he gets to sound smart and explain why his super-secret formula "should" have worked. Roping in a new batch of suckers.
This Clifford dude was actually invited onto a Bogleheads podcast, thus allowing him yet another vehicle to, as you so wisely say, rope in a new batch of suckers.

Shameful, IMHO.
Why is it shameful? Only because he charges "too much" for his funds, or because you disagree with there being any benefit to factor analysis? If the latter, would other security selection processes be acceptable for podcast guests? Or would you have banned Bogle too, based on his anonymous FAJ article? How about Graham, or other active management proponents? Siegel, in light of Wisdom Tree? Who would be acceptable to you?
The man made $3 billion plus. In a few years... He wasn't on the Forbes list, and then suddenly he was... with $3 billion...

The customers who invested with him LOST money.

Where are the customer's yachts?

Of course that's shameful.

I don't know how anyone can respect Cliff Asness.

Jack Bogle never made that kind of money. And he could have... But he didn't. He gave US the money.

If you make $3 billion, and the people who trust you make money as well, fine... or if you provide a product or a service for that money, fine.

If you make $3 billion, and everyone who invested with you loses money, but you still make $3 billion, you are a horrible human being. $3 billion is an insane amount of money to make, when you don't even provide a service that works.

Bill Gates, Steve Jobs, Jeff Bezos, fine... We all actually got something for the money they made.

Nobody got ANYTHING from Cliff Asness... Pretty much all his investors lost money. Their money got transferred to him.

It's terrible.
Well said. I’m very grateful that I am firmly convinced nobody can out perform the market over long periods of time. It saves me a lot of money and stress.

I suppose people are nervous, because there is a lot of nervous talk these days, and people seemingly are looking for an “authority” that can save them from their fears. There will be plenty of accomplished, smart sounding people like this who will gladly manage their money for them..... but these guys don’t know anything either, they can’t deliver.....
000
Posts: 8211
Joined: Thu Jul 23, 2020 12:04 am

Re: A Gut Punch (Cliff Asness of AQR on the performance of multifactor equity models)

Post by 000 »

Random Walker wrote: Sat Dec 26, 2020 8:04 pm To some extent I have to agree with you, yes. But I believe the dominant goal of the TSM or S&P 500 investor is to eliminate single stock risk. If one has 25 stocks in a portfolio, the potential contribution of any one stock might be 4%. If one has 100 stocks in a portfolio, the potential contribution of any one stock might be 1%. By the time we are talking 500-1000 stocks in a SV fund, 500 stocks in S&P 500, or 3000 stocks in Total Stock Market fund, single stock risk is effectively eliminated in any of them.
If one is looking at diversification in terms of the drivers of stock returns, then he views a TSM fund and a SV fund as both having an exposure to the market factor of about 1, and whether that exposure is through 1000 or 3000 stocks may not be that relevant. Of course I write this having semi-painfully not owned my share of FANG over the last decade :-)

Dave
I agree that it's possible to diversify away from TSM, but where you lose me is the claim that TSM and SV both have the same market factor exposure of 1. Either that is not correct or the definitions are not meaningful: your own example of missing FAANG over a decade is a kind of idiosyncratic risk itself, that of missing the few great stocks.
000
Posts: 8211
Joined: Thu Jul 23, 2020 12:04 am

Re: A Gut Punch (Cliff Asness of AQR on the performance of multifactor equity models)

Post by 000 »

HomerJ wrote: Sat Dec 26, 2020 9:52 pm You think the academics can figure out the odds? They can't.

It's like if you saw 100 hands of blackjack without knowing the rules, and had to figure out the odds from the hands you saw.

Oh, AND the rules change every 10-20 years/hands.
Yep.
User avatar
nedsaid
Posts: 19275
Joined: Fri Nov 23, 2012 11:33 am

Re: A Gut Punch (Cliff Asness of AQR on the performance of multifactor equity models)

Post by nedsaid »

HomerJ wrote: Sat Dec 26, 2020 9:52 pm
nedsaid wrote: Sat Dec 26, 2020 5:24 pmThe academics are looking at increasing an investor's chance of success, it isn't a snake oil get rich quick scheme.
The "academics" have been proven wrong again and again.

They are not evil, they are not snake oil salesmen.

But man are they deluded and misguided.

Economics isn't physics.

There aren't enough data points, there are a hundred variables, and even the "constants" change occasionally.

You think the academics can figure out the odds? They can't.

It's like if you saw 100 hands of blackjack without knowing the rules, and had to figure out the odds from the hands you saw.

Oh, AND the rules change every 10-20 years/hands.

It's impossible. Stop believing the "academics"
I think Homer you are way overstating your case. One reason that I was open to the academic research was that it was consistent with what I already knew or suspected. It was also consistent with my own experiences as an investor. The suggestions to combine Value with Quality and the comments that Value stocks are correlated with negative momentum squared with my own experiences as an investor. If I thought this was a bunch of Ivory Tower hocus pocus, I never would have listened. The Academics have also said things that well known market professionals have been saying for years and years. By the time I had learned about Small Value tilting back in 2007, I already had 23 years of experience investing and 19 years of experience with individual stocks. I had many years of watching the financial markets. So I wasn't a naive individual when I got exposed to this.

Above was just a rant and frankly I am pretty disappointed. I have observed the markets for many years and while I don't know much, I know something.

The economy and the markets are dynamic and it surprises me that you seem to believe something is wrong with people if they don't say the same thing year after year not taking into account that things change. Mr. Bogle himself was fairly consistent in his philosophy but he did say and do some surprising things over his career. He took Wellington from a conservative investment shop to a Go Go Growth Manger after a merger and got fired for it, or in his own words fired with enthusiasm. Wellington returned to its roots. He took this adversity to start Vanguard and later the first retail index fund. Bogle also started factor funds at Vanguard, Vanguard also started sector funds, and he kept around the active funds including Windsor Fund and John Neff who were instrumental in Vanguard's success. The raging success with index funds didn't start coming until the early 1990's. So Mr. Bogle didn't operate in a vacuum, he was affected by the dynamic economic and investing environment and made changes accordingly. Like any smart businessman, he provided products that investors wanted even if it didn't all fit 100% with his philosophies.

I have always taken the attitude that I could learn something from people with whom I disagreed. If I just dismissed what they had to say then I would be a lot less knowledgeable today. I listen to all kinds of people. There is an attitude here to just dismiss everything somebody ever says because they changed their mind on something. It has made the forum a less interesting place.
A fool and his money are good for business.
User avatar
HomerJ
Posts: 21281
Joined: Fri Jun 06, 2008 12:50 pm

Re: A Gut Punch (Cliff Asness of AQR on the performance of multifactor equity models)

Post by HomerJ »

I think it's normal and very understandable to think that the more you study a subject, or the more you strive to learn, the closer you will get to mastering that subject.

It works in almost every other aspect of life.

But not so much investing. The "experts" appear to be just guessing, along with the rest of us.

I don't think that "Nobody knows nothing", but I think that "Nobody knows enough".

I could be wrong. Maybe it is possible to learn enough about investing and factors to consistently make more than the market.

But all my experience (and the experience of most people on this board) tells me otherwise. Most people who seem to know far more than me don't out-perform the market. Predictions from "experts" are almost always wrong.

Look at my sig. That's an "expert" talking there.

Again, I could be wrong. But it sure seems I'm right doesn't it? All these AQR funds came out 7-8 years and most of them have lost money. Cliff Asness is a genius with more geniuses working for him, all of whom study and work this subject all day long and they know far more about the markets and factors than I ever will.

Yet, his funds lost money over all this time.

These were funds that were supposed to balanced across multiple factors, that would give a positive return even if one or two factors performed poorly. It appears his algorithms were incorrect.

I'm not sure that dabbling in understanding factors is useful when a PhD can't even make it work.
"The best tools available to us are shovels, not scalpels. Don't get carried away." - vanBogle59
000
Posts: 8211
Joined: Thu Jul 23, 2020 12:04 am

Re: A Gut Punch (Cliff Asness of AQR on the performance of multifactor equity models)

Post by 000 »

nedsaid wrote: Sun Dec 27, 2020 12:35 am By the time I had learned about Small Value tilting back in 2007, I already had 23 years of experience investing and 19 years of experience with individual stocks. I had many years of watching the financial markets. So I wasn't a naive individual when I got exposed to this.
Yet you bought in AFTER the great performance of SCV.

And you've earned a negative premium since 2007... https://www.portfoliovisualizer.com/bac ... ion2_2=100
User avatar
Anon9001
Posts: 1884
Joined: Fri Dec 20, 2019 8:28 am
Location: بھارت

Re: A Gut Punch (Cliff Asness of AQR on the performance of multifactor equity models)

Post by Anon9001 »

000 wrote: Sun Dec 27, 2020 2:50 am
nedsaid wrote: Sun Dec 27, 2020 12:35 am By the time I had learned about Small Value tilting back in 2007, I already had 23 years of experience investing and 19 years of experience with individual stocks. I had many years of watching the financial markets. So I wasn't a naive individual when I got exposed to this.
Yet you bought in AFTER the great performance of SCV.

And you've earned a negative premium since 2007... https://www.portfoliovisualizer.com/bac ... ion2_2=100
Do you not own Ex-US stocks due to their underperformance since 2007? What relevance does this have on future performance other than to caution people here that TSM will also underperform due to great past performance.
Land/Real Estate:89.4% (Land/RE is Inheritance which will be recieved in 10-20 years) Equities:7.6% Fixed Income:1.7% Gold:0.8% Cryptocurrency:0.5%
User avatar
typical.investor
Posts: 5263
Joined: Mon Jun 11, 2018 3:17 am

Re: A Gut Punch (Cliff Asness of AQR on the performance of multifactor equity models)

Post by typical.investor »

000 wrote: Sun Dec 27, 2020 2:50 am
nedsaid wrote: Sun Dec 27, 2020 12:35 am By the time I had learned about Small Value tilting back in 2007, I already had 23 years of experience investing and 19 years of experience with individual stocks. I had many years of watching the financial markets. So I wasn't a naive individual when I got exposed to this.
Yet you bought in AFTER the great performance of SCV.

And you've earned a negative premium since 2007... https://www.portfoliovisualizer.com/bac ... ion2_2=100
13 years of underperformance shouldn't be so unexpected. Factors have a large amount of kurtosis, correlations can converge, and factor momentum is serial. This all means that you can expect large drawdowns across factors that persist. I believe that view should be commonly known instead of the rosy picture of "factor diversification" eliminating factor investing risk. Of course, there will be a time when those things will work in factor investors' favor too. Hopefully in my holding period!
Post Reply