Skating where the puck was; e.g., commodities. Factors next?

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply
User avatar
Topic Author
CULater
Posts: 2832
Joined: Sun Nov 13, 2016 9:59 am
Location: Hic sunt dracones

Skating where the puck was; e.g., commodities. Factors next?

Post by CULater »

Commodities have gotten crushed with higher volatility than stocks. Not only have the returns been terrible but the correlation benefits were lost during the crisis as well as commodities crashed right alongside stocks.
The lifecycle of a novel investment strategy looks something like this:

- Early investors earn higher returns through a first mover advantage in an unproven strategy.
- Academic research papers highlight the strategy in detail and make it sound too good to be true.
- Wall Street creates a bunch of products to take advantage of said research.
- Retail and institutional investors pile in after all of the huge gains have likely already been extracted.
- Rinse and repeat.
http://awealthofcommonsense.com/2018/01 ... ment-idea/

So, could the same thing happen to "factor investing?" I remember how convincing the case was for commodities back when...
On the internet, nobody knows you're a dog.
livesoft
Posts: 86075
Joined: Thu Mar 01, 2007 7:00 pm

Re: Skating where the puck was; e.g., commodities. Factors next?

Post by livesoft »

Perhaps, but stocks are stocks and all of them are basically found in the Total US Stock Market Index fund or the Total International Stock Market index fund. As such, stocks of a certain factor will not get crushed much more than stocks that aren't classified with that factor. That also means that playing with factors is playing with small differences that may just be in the noise anyways.

And playing in the noise means that any little stupid thing will wipe out any advantage and possibly accentuate any disadvantage. But that goes with total market investing, too. Any behavioral mistakes and one is behind the benchmarks.
Wiki This signature message sponsored by sscritic: Learn to fish.
Random Walker
Posts: 5561
Joined: Fri Feb 23, 2007 7:21 pm

Re: Skating where the puck was; e.g., commodities. Factors next?

Post by Random Walker »

I’m all in on factors, but I’m certainly concerned about skating where the puck was. Factor investing is certainly popular, widely accessible, and more expensive. But there are strong forward looking risk based and behavioral based reasons to believe premia should persist. The market prices risk, so an efficient market should discount the prices of riskier small and value stocks appropriately so that their expected returns are greater. I believe human behavior is stubbornly persistent, so I think there is strong reason to have faith in the purely behavioral cross sectional momentum and time series momentum. I especially have faith in value because there are both risk based and behavioral based explanations for it.
Larry’s 5 criteria for evaluating a factor (really any potential investment) are persistent, pervasive, robust, investable, intuitive. To me persistent, pervasive, robust are all dependent on past data and potentially susceptible to skating where the puck was. Investable sets the investment up for us Bozos to skate where the puck was. But intuitive looks forward. Moreover intuitive is the source of the investor’s conviction to stick with the factor during the inevitable periods of underperformance. Historical data is very useful, important, and helps the investor with conviction. But there is always the nagging concern of skating where the puck was. The risk and behavioral based intuitive explanations for factors look forward.

Dave
asif408
Posts: 2616
Joined: Sun Mar 02, 2014 7:34 am
Location: Florida

Re: Skating where the puck was; e.g., commodities. Factors next?

Post by asif408 »

I think so. If anything, I'd say commodities now are pretty unloved and out of favor (though maybe not quite as much as a few years ago), while factor investing is certainly all the rage. It is interesting to note that commodities correlations with equities has fallen quite a bit in the past few years. So I favor the unloved and out of favor and avoid things that are loved. So I'm a good ole' boring market cap guy with some commodities thrown in.

If you do invest in commodities, you can avoid the issue of negative roll yield (i.e., contango) by investing in the stocks of commodities producers, which along with CCFs have had zero to negative returns the last decade (though not quite as bad as CCFs).
staythecourse
Posts: 6993
Joined: Mon Jan 03, 2011 8:40 am

Re: Skating where the puck was; e.g., commodities. Factors next?

Post by staythecourse »

I have to disagree with trying to correlation the terrible results of commodities to "skating where the puck was".

Commodities return is based on 3 factors: Roll return, collateral return, and spot price. Spot price has always tracked inflation so no real return there. Collateral return (most are done by 30 day tbills) is lousy since interest rates are low so no return there. So all of the return is based on roll return, i.e. hoping one is in backwardation and not in contango. Since we have not had any major supply chain disruptions in oil and oil demand has been so low we have been stuck in contango.

So with that info. why in the world would one expect commodities to do well since it went mainstream around 2006 or so?

Also, Vanguard paper shows that commodities doe great as a short term hedge (3 months) with there are times of unexpected inflation. Have we had any of that during this time period?

Also, if commodities have done so poorly due to so much money comig into the space then why are they still doing so poorly since I am positive there has been so much money leaving the space due to poor returns?

I think trying connect the terrible results to money entering the space doesn't pass the sniff test of common sense. This is a clear case of: correlation does not equal causation.

Good luck.

p.s. One has a better strength on "skating where the puck is" if one can show high correlation of price movement of x with money increasing or decreasing from the available funds in that space.
"The stock market [fluctuation], therefore, is noise. A giant distraction from the business of investing.” | -Jack Bogle
User avatar
nisiprius
Advisory Board
Posts: 52211
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: Skating where the puck was; e.g., commodities. Factors next?

Post by nisiprius »

I think this is already happening. For example, even Berkin and Swedroe's Your Complete Guide to Factor-Based Investing cites academic research showing clear evidence of a decline effect in factor premia after they become widely known. The book presents it in a context of minimizing it, says the size of the decline is only about 30%... but they acknowledge its decline.

The way the factors crowd has moved on from size and value to "fresh woods and pastures new" suggests as much. The size factor is interesting. The factor mavens will not acknowledge it as discredited, but they mention it less and less and as having less and less importance... only shows up as a synergistic booster for value, for example. I think it's widely acknowledged that the paper that brought the size effect to light in 1981 used faulty data and that the effect is much less that it was said to be in 1981. The value factor is often described as "missing in action for twelve years."

It no longer seems to be believed that something like the Coffeehouse Portfolio--e.g. 10% each Vanguard 500 index, Vanguard [Large-Cap] Value Index, Vanguard Small-Cap Index, Vanguard Small-Cap Value Index, 10% Vanguard REIT index, 10% Vanguard Total International, 40% Total Bond--is sufficient for any serious connoisseur. And Larry Swedroe seems to be writing less and less about traditional DFA-centered portfolios and more and more about AQR multifactor leveraged long-short funds and Stone Ridge highly non-traditional interval funds.

More generally, I believe in the reality of something variously described as "the decline effect" (used of scientific research), "mean reversion" (as in John C. Bogle's "The Telltale Chart,") or "Rekenthaler's Rule" ("If the bozos know about it, it doesn't work anymore.")

I regard myself as a bozo. By the time I hear about anything, I am already late to the party. I accept this. There is absolutely no percentage in it for me to kid myself. I can't un-bozo myself just by reading a few articles.

With regard to factors and such, there are two possibilities. One it that it is all illusory, the result of unconscious selection of things to backtest. The other possibility is that it is real, but everything fades and you need to get in on these things "on the ground floor," i.e. before the bozos know about it. Another way to phrase it is that you somehow have to invest in things when they are still poorly known, oddball, and not generally regarded as a good idea. This applies to actively managed funds, too. Most of the legendary active funds had their really impressive results before they became famous... sometimes, of course, before they were even offered to the public.

Here's what I don't know, but I've personally long since thrown in the towel on. Is it possible that a shrewd and well-informed investor could consistently spot fads/trends/novelties/latest things... things that are only likely to last for another 3-10 years... early on, when they still have enough juice left in them to be worth squeezing? I dunno. I think probably not. Not me, anyway. I think the only way to do that is either to have the equivalent of inside information (legal or even illegal) or, better yet, to be the person who publicizes the effect and participates in creating the products based on them.

In Ye Olde Days the literal historic Edward Francis Hutton really did have inside information and really did share it with selected clients. Then the pesky SEC happened. Nowadays I wonder if one of the things advisors are selling is the idea that they have better advance information on valid trends than the average do-it-yourself investor.
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: Skating where the puck was; e.g., commodities. Factors next?

Post by nedsaid »

It is the old "late to the party" problem. Really popular strategies will tend to stop working at least temporarily. No investment strategy works all the time. Pretty much, pick a solid investment approach and stick with it or you will keep switching at precisely the wrong time dooming yourself to underperformance.
A fool and his money are good for business.
golfCaddy
Posts: 728
Joined: Wed Jan 10, 2018 9:02 pm

Re: Skating where the puck was; e.g., commodities. Factors next?

Post by golfCaddy »

What's interesting is if you look at the original momentum paper: http://www.business.unr.edu/faculty/liu ... n_1993.pdf, in table 1, how sensitive the long-short premium is to how its defined: the choice of J, K, and panel A vs. panel B.
fennewaldaj
Posts: 1097
Joined: Sun Oct 22, 2017 11:30 pm

Re: Skating where the puck was; e.g., commodities. Factors next?

Post by fennewaldaj »

So for an effect that is well known if it is to no longer be real due to over fishing we should expect it to deliver outsized returns from all the new money pouring in for a while and then have lower future expected returns. From reading Larry Swedroe's book it appears that might have happened with the low volatility anomaly/factor. Value dose not seem to have followed that path recently though perhaps it did in the early 2000s. Maybe its just not as trendy now. From data posted here recently the spread in valuations between value and growth is as big as it has been in a while. Looking at vanguard funds (info from vanguard website)
Large Cap
Vangaurd value index (VVIAX) P/B 2.3, P/E 19.6
Vangaurd growth index (VIGAX) P/B 5.2, P/E 28.6

Mid Cap
Vangaurd mid cap value index(VMVAX) P/B 2.1, P/E 19.4
Vangaurd mid cap growth index (VMGRX) P/B 4.9, P/E 32.6

Small Cap
Vanguard Small-Cap Value Index Fund Admiral (VSIAX) P/B 1.9, P/E 17.4
Vanguard Small-Cap Growth Index Fund Admiral (VSGAX) P/B 3.5, P/E 29.9

It doesn't really look like value has been overbought from these numbers. Not sure about the other new factors. Quality/profitablity has a free lunch fell much like low vol which could easily make it more popular. Momentum will likely exist as long as there are markets but it will likely be hard to implement.
Dead Man Walking
Posts: 1212
Joined: Wed Nov 07, 2007 5:51 pm

Re: Skating where the puck was; e.g., commodities. Factors next?

Post by Dead Man Walking »

It's been my experience that academics provide entertaining research and theories about many aspects of the human condition; however, some of their work doesn't really apply to the real world human experience. Investing is not an exception to this observation.

DMW
User avatar
saltycaper
Posts: 2650
Joined: Thu Apr 24, 2014 8:47 pm
Location: The Tower

Re: Skating where the puck was; e.g., commodities. Factors next?

Post by saltycaper »

nisiprius wrote: Fri Jan 12, 2018 11:02 am
I regard myself as a bozo. By the time I hear about anything, I am already late to the party. I accept this. There is absolutely no percentage in it for me to kid myself. I can't un-bozo myself just by reading a few articles.
Better sell your stocks. You're late to the party. Every bozo knows they tend to outperform bonds.

I'm partly kidding, of course. It's certainly easier to identify how stocks and bonds differ in risk than, say, SCV stocks and a LCG stocks differ in risk. But just because you know of a risk doesn't remove the risk, and knowing doesn't eliminate the expectation of receiving a higher return, even though no return is guaranteed.
Quod vitae sectabor iter?
User avatar
packer16
Posts: 1488
Joined: Sat Jan 04, 2014 1:28 pm

Re: Skating where the puck was; e.g., commodities. Factors next?

Post by packer16 »

We should expect the value to growth gap to grow over time as the market more efficiently divides stocks into growth (good stocks) & value (crappy stocks). The implicit assumption you are making when you expect the spread to have mean that is being reverted to is the market is no more efficient then when the data you collected for your mean calculation began.

The real question I have is there a way to measure the crappiness (maybe leverage, ROIC, margins or FCF growth) & adjust the multiple for this then you can make a statement about whether value is overbought or not.

Packer
Buy cheap and something good might happen
User avatar
Top99%
Posts: 465
Joined: Sat Apr 22, 2017 9:30 am
Location: Austin, TX

Re: Skating where the puck was; e.g., commodities. Factors next?

Post by Top99% »

I never really understood the case for commodities but I find the evidence for at least the value and momentum factors compelling enough to tilt that way especially as TSM increasingly becomes a big bet on large cap growth at current valuations. I think at today's LCG valuations one could easily argue lots of people are chasing that puck. US LCG may indeed rule the world and this time might be different than the "nifty 50". Or not.
Adapt or perish
User avatar
nisiprius
Advisory Board
Posts: 52211
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: Skating where the puck was; e.g., commodities. Factors next?

Post by nisiprius »

Top99% wrote: Sat Jan 13, 2018 9:32 am...as TSM increasingly becomes a big bet on large cap growth at current valuations...
NO. IT. ISN'T.

That is exactly what makes the total stock market different from every other strategy: it is the only one that cannot be overgrazed. If you have a people in a cafeteria ladling Italian Wedding Soup out of an urn, and every one of them manipulates the ladle to get more than their share of meatballs, over time the composition of the remaining soup changes and people who get in later find fewer meatballs per quart in the urn, and a declining "meatball premium."

But carefully stirring the soup and taking out a ladleful whose composition is the same as the rest of urn does not change the composition of what it left behind, and everybody can take out a ladleful like that; the strategy is sustainable until the soup is all gone.

Not everybody can have a portfolio that is 20% small-cap value because the market is only 2% small-cap value. But everyone can invest in Total Stock and it does not alter or distort the market in any way.

Now, it can be argued that a population of retirement savers automatically buying $X worth of Total Stock Market every month, no matter what they cost, distorts the market as a whole somehow... but it has nothing to do with total stock or cap-weighting, and it has the same effect on all stocks.
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
jhfenton
Posts: 4754
Joined: Sat Feb 07, 2015 10:17 am
Location: Ohio

Re: Skating where the puck was; e.g., commodities. Factors next?

Post by jhfenton »

nisiprius wrote: Sat Jan 13, 2018 11:08 am
Top99% wrote: Sat Jan 13, 2018 9:32 am...as TSM increasingly becomes a big bet on large cap growth at current valuations...
NO. IT. ISN'T.

That is exactly what makes the total stock market different from every other strategy: it is the only one that cannot be overgrazed.
Just to play devil's advocate for a minute...that particular argument isn't that TSM investors make large cap growth dominant, the argument is that active investors have made large cap growth so dominant as a portion of the total market that they drag TSM investors along with them. To follow your analogy, the Meatball Investors buy so many meatballs for the soup that it becomes impossible for a Total Soup Investor to get a ladle of soup that isn't dominated by (overpriced) meatballs.
staythecourse
Posts: 6993
Joined: Mon Jan 03, 2011 8:40 am

Re: Skating where the puck was; e.g., commodities. Factors next?

Post by staythecourse »

nisiprius wrote: Sat Jan 13, 2018 11:08 am
Top99% wrote: Sat Jan 13, 2018 9:32 am...as TSM increasingly becomes a big bet on large cap growth at current valuations...
NO. IT. ISN'T.
Actually it is. TSM is 70% SP500. Look at TSM and SP500 they are basically the same returns, risk, and likely correlation coefficient. The 70% is such a large portion of TSM the 20% in mid, 7% in small, and 3% in micro does not make much of a difference. There is not much difference between DJ, SP500, and TSM. What happens to the Large cap will basically decide what happens to the TSM return.

Good luck.
"The stock market [fluctuation], therefore, is noise. A giant distraction from the business of investing.” | -Jack Bogle
User avatar
BrooklynInvest
Posts: 1183
Joined: Sun Jul 28, 2013 9:23 am

Re: Skating where the puck was; e.g., commodities. Factors next?

Post by BrooklynInvest »

Sorry if this is naive - what's the difference between factor investing and, admittedly sophisticated, backtesting? A dumb example -

Let's say that over the last 30 years, stocks in my (indexed) portfolio at the very beginning of the alphabet - Apple - have performed better than stocks at the end of the alphabet - Xerox. Is "alphabet investing" a factor?

If so I'm opening a hedge fund to some sucker . . . . um, pioneering investors!

Over any given period there has to be a group of outperformers and underperforms linked by some commonality. Value over growth, or the reverse, small over large, dividend growers versus dividend payers yada yada.

Like the old "Dogs of the Dow" - are we implying structural causality and therefore opportunity to mere long term coincidence?
thx1138
Posts: 1164
Joined: Fri Jul 12, 2013 2:14 pm

Re: Skating where the puck was; e.g., commodities. Factors next?

Post by thx1138 »

BrooklynInvest wrote: Sat Jan 13, 2018 12:14 pm Sorry if this is naive - what's the difference between factor investing and, admittedly sophisticated, backtesting? A dumb example -

Let's say that over the last 30 years, stocks in my (indexed) portfolio at the very beginning of the alphabet - Apple - have performed better than stocks at the end of the alphabet - Xerox. Is "alphabet investing" a factor?

If so I'm opening a hedge fund to some sucker . . . . um, pioneering investors!

Over any given period there has to be a group of outperformers and underperforms linked by some commonality. Value over growth, or the reverse, small over large, dividend growers versus dividend payers yada yada.

Like the old "Dogs of the Dow" - are we implying structural causality and therefore opportunity to mere long term coincidence?
The difference is that "factors" should persist across different markets (e.g. US, Asia, Europe), different time periods (e.g. you don't have to fine tune start stop dates to see the effect), modest changes in the selection critieria (e.g. what metric do I use for "value" and what threshold for that metric) and typically have an explanation for them that makes sense and usually isn't a free lunch (e.g. Small-Value has larger volatility and extended periods of underperformance that are "rewarded" by a higher expected return).

That doesn't mean there aren't risks of just being back testing especially as folks search for more and more factors to explain total market returns. But the general red flags for plain old backtesting overfitting are sensitivity to exact market, time period or selection metric. The traditional factors aren't sensitive in that way and so are considered valid models rather than overfitted historical data.
User avatar
nisiprius
Advisory Board
Posts: 52211
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: Skating where the puck was; e.g., commodities. Factors next?

Post by nisiprius »

BrooklynInvest wrote: Sat Jan 13, 2018 12:14 pm Sorry if this is naive - what's the difference between factor investing and, admittedly sophisticated, backtesting? A dumb example -

Let's say that over the last 30 years, stocks in my (indexed) portfolio at the very beginning of the alphabet - Apple - have performed better than stocks at the end of the alphabet - Xerox. Is "alphabet investing" a factor?

If so I'm opening a hedge fund to some sucker . . . . um, pioneering investors!

Over any given period there has to be a group of outperformers and underperforms linked by some commonality. Value over growth, or the reverse, small over large, dividend growers versus dividend payers yada yada.

Like the old "Dogs of the Dow" - are we implying structural causality and therefore opportunity to mere long term coincidence?
It's hard to get away from backtesting because we don't have the patience to wait for investing experiments.

Speaking as someone who is a skeptic, there is, nevertheless, a big difference between factor investing and your "alphabet" investing. The idea behind factors is to discover large groups of stocks that behave similarly to each other... and different from the rest. In the days of the three-factor Fama-French model, we would have said: when we know that Alphabet is a stock, we know 70% of all that can be known about it. It is not some completely unique thing, it's just 70% "it's a stock" and 30% "completely random behavior we can't predict, due to the efficient market having squeezed everything predictable out of it."

By exploring for factors, they tried to find general, common... well... factors that could be used to explain or predict some of that 30%. So instead of saying, one extreme, "Alphabet is a stock and that's all we can say about it," or, other extreme, "Alphabet is everything in its annual report and ten thousand other details..." we can say "Alphabet is +1.29 market, -0.46 size, -0.92 value. THAT tells us 90% of all that can be known about it, and the other 10% is completely random."

Once we have our factors (if it's all true!) we see the world in terms of factors, and stop trying to pick individual stocks and instead start trying to pick some mix of the three factors.

As to where it fits--again, if it's all true--it is an important step in the "prescientific" stage, the stage in which investigators find useful classifications that have some predictive value. Oh, it's in this column of the periodic table, therefore it is probably not going to react with anything; it's over here, so it's a metal, so it will probably conduct electricity. It's not quite being able to solve the quantum wave equations for the atom, but it's more than saying "so much earth, so much air, so much fire, it burns because it's full of the fire nature."
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
nisiprius
Advisory Board
Posts: 52211
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: Skating where the puck was; e.g., commodities. Factors next?

Post by nisiprius »

staythecourse wrote: Sat Jan 13, 2018 11:52 am
nisiprius wrote: Sat Jan 13, 2018 11:08 am
Top99% wrote: Sat Jan 13, 2018 9:32 am...as TSM increasingly becomes a big bet on large cap growth at current valuations...
NO. IT. ISN'T.
Actually it is. TSM is 70% SP500. Look at TSM and SP500 they are basically the same returns, risk, and likely correlation coefficient. The 70% is such a large portion of TSM the 20% in mid, 7% in small, and 3% in micro does not make much of a difference. There is not much difference between DJ, SP500, and TSM. What happens to the Large cap will basically decide what happens to the TSM return.

Good luck.
It's not a "bet." We are not "betting" that the market will beat the market. We are depending on the sure thing that the market will tie the market.

We put most of our dollars in big companies, not because are betting on our belief in the superiority of big companies, but it's because it's where the market has chosen to put most of its dollars.

Factor investors are making intelligent bets founded on rational analysis that certain systematic departures from the market will beat the market.
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.
staythecourse
Posts: 6993
Joined: Mon Jan 03, 2011 8:40 am

Re: Skating where the puck was; e.g., commodities. Factors next?

Post by staythecourse »

nisiprius wrote: Sat Jan 13, 2018 1:11 pm
staythecourse wrote: Sat Jan 13, 2018 11:52 am
nisiprius wrote: Sat Jan 13, 2018 11:08 am
Top99% wrote: Sat Jan 13, 2018 9:32 am...as TSM increasingly becomes a big bet on large cap growth at current valuations...
NO. IT. ISN'T.
Actually it is. TSM is 70% SP500. Look at TSM and SP500 they are basically the same returns, risk, and likely correlation coefficient. The 70% is such a large portion of TSM the 20% in mid, 7% in small, and 3% in micro does not make much of a difference. There is not much difference between DJ, SP500, and TSM. What happens to the Large cap will basically decide what happens to the TSM return.

Good luck.
It's not a "bet." We are not "betting" that the market will beat the market. We are depending on the sure thing that the market will tie the market.

We put most of our dollars in big companies, not because are betting on our belief in the superiority of big companies, but it's because it's where the market has chosen to put most of its dollars.

Factor investors are making intelligent bets founded on rational analysis that certain systematic departures from the market will beat the market.
I don't think he factor folks consider it a bet any more then a bet on beta. Anyone can believe what they want to believe, but there is plenty of academic evidence which is accepted as fact that premiums exits. If I or your or anyone want to minimize that and say it is a "bet" then it is no different then someone else stating beta is a bet as well. All of them are factors it just matter which one's you want to invest in.

Just because I do or don't invest in certain premiums doesn't mean I believe they don't exist as strong as the ERP does.

Good luck.
"The stock market [fluctuation], therefore, is noise. A giant distraction from the business of investing.” | -Jack Bogle
User avatar
Topic Author
CULater
Posts: 2832
Joined: Sun Nov 13, 2016 9:59 am
Location: Hic sunt dracones

Re: Skating where the puck was; e.g., commodities. Factors next?

Post by CULater »

nisiprius wrote: Sat Jan 13, 2018 1:11 pm
staythecourse wrote: Sat Jan 13, 2018 11:52 am
nisiprius wrote: Sat Jan 13, 2018 11:08 am
Top99% wrote: Sat Jan 13, 2018 9:32 am...as TSM increasingly becomes a big bet on large cap growth at current valuations...
NO. IT. ISN'T.
Actually it is. TSM is 70% SP500. Look at TSM and SP500 they are basically the same returns, risk, and likely correlation coefficient. The 70% is such a large portion of TSM the 20% in mid, 7% in small, and 3% in micro does not make much of a difference. There is not much difference between DJ, SP500, and TSM. What happens to the Large cap will basically decide what happens to the TSM return.

Good luck.
It's not a "bet." We are not "betting" that the market will beat the market. We are depending on the sure thing that the market will tie the market.

We put most of our dollars in big companies, not because are betting on our belief in the superiority of big companies, but it's because it's where the market has chosen to put most of its dollars.

Factor investors are making intelligent bets founded on rational analysis that certain systematic departures from the market will beat the market.
It's not quite that, IMO. One claim made by factor investors is that you can use factors to tailor your portfolio asset allocation to achieve certain objectives. For example, Larry Swedroe has discussed the notion of investing a smaller equity percentage into SCV (small and value factors), allowing a larger percentage to be invested in treasuries. This might allow the portfolio to achieve the same cumulative return with less volatility. For example, from 2005 - 2013 SCV and TSM had about the same return. But if you invested 30% in SCV + 70% in Intermediate Treasuries, you got the same portfolio return as 60% in TSM + 40% Treasuries but with 2/3 the volatility and a worst drawdown of 12% vs. 29%.

The Small and Value factors didn't beat TSM during this period, but by heavily tilting toward these factors you could have configured your portfolio differently to deliver the same return as a larger allocation to stocks, with consequent lower portfolio volatility and drawdown.
On the internet, nobody knows you're a dog.
User avatar
HomerJ
Posts: 21281
Joined: Fri Jun 06, 2008 12:50 pm

Re: Skating where the puck was; e.g., commodities. Factors next?

Post by HomerJ »

A funny observation CULater, is that even valuations of P/E ratios somehow seem to have stopped working as a prediction factor.

As soon as they were quantified and became known, they stopped working. PE ratios have been far above the historical average for 25 years.

That's even during TWO stock market crashes, one of which was nearly another Great Depression with the collapse of financial markets. And even then, PE ratios barely dropped below the historical NORM, nowhere near the historical low points.

Now, I'm not saying we're not going to see a crash... Of course, we will... I just find it interesting that valuations have been a terrible predictor of stock market returns ever since the model was quantified.
User avatar
nisiprius
Advisory Board
Posts: 52211
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: Skating where the puck was; e.g., commodities. Factors next?

Post by nisiprius »

CULater wrote: Sat Jan 13, 2018 2:19 pm
nisiprius wrote: Sat Jan 13, 2018 1:11 pm... Factor investors are making intelligent bets founded on rational analysis that certain systematic departures from the market will beat the market...
It's not quite that, IMO. One claim made by factor investors is that you can use factors to tailor your portfolio asset allocation to achieve certain objectives. For example, Larry Swedroe has discussed the notion of investing a smaller equity percentage into SCV (small and value factors), allowing a larger percentage to be invested in treasuries. This might allow the portfolio to achieve the same cumulative return with less volatility. For example, from 2005 - 2013 SCV and TSM had about the same return. But if you invested 30% in SCV + 70% in Intermediate Treasuries, you got the same portfolio return as 60% in TSM + 40% Treasuries but with 2/3 the volatility and a worst drawdown of 12% vs. 29%.

The Small and Value factors didn't beat TSM during this period, but by heavily tilting toward these factors you could have configured your portfolio differently to deliver the same return as a larger allocation to stocks, with consequent lower portfolio volatility and drawdown.
Suppose I amend my remark and say:

"Factor investors are making intelligent bets founded on rational analysis that certain systematic departures from the market will beat the market in risk-adjusted return?"

In the case of the Larry Portfolio, his goal is to hold return constant and reduce risk, hence his goal is superior risk-adjusted return--which he prefers to take in the form of lower risk rather than higher return.
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.
lack_ey
Posts: 6701
Joined: Wed Nov 19, 2014 10:55 pm

Re: Skating where the puck was; e.g., commodities. Factors next?

Post by lack_ey »

nisiprius wrote: Sat Jan 13, 2018 3:59 pm Suppose I amend my remark and say:

"Factor investors are making intelligent bets founded on rational analysis that certain systematic departures from the market will beat the market in risk-adjusted return?"

In the case of the Larry Portfolio, his goal is to hold return constant and reduce risk, hence his goal is superior risk-adjusted return--which he prefers to take in the form of lower risk rather than higher return.
Let's go with "...arguably intelligent bets..." to not give quite as much credit and then maybe "will enhance the odds of a more desirable outcome" (so there's wiggle room as to the intent, what counts as more desirable, as that will change based on strategy and investor; low vol strats for example may seek to improve risk/return, and there are other ideas about shaping the distribution of outcomes).

So
Factor investors are making arguably intelligent bets founded on rational analysis that certain systematic departures from the market will enhance the odds of a more desirable outcome.
randomguy
Posts: 11295
Joined: Wed Sep 17, 2014 9:00 am

Re: Skating where the puck was; e.g., commodities. Factors next?

Post by randomguy »

You know if you stay in place eventually the puck comes back.:) Skating where the puck was isn't that bad as long as you stop when you get there (i.e. don't switch from value to momentum to quality as each posts nice 3 year returns)
golfCaddy
Posts: 728
Joined: Wed Jan 10, 2018 9:02 pm

Re: Skating where the puck was; e.g., commodities. Factors next?

Post by golfCaddy »

thx1138 wrote: Sat Jan 13, 2018 12:36 pm That doesn't mean there aren't risks of just being back testing especially as folks search for more and more factors to explain total market returns. But the general red flags for plain old backtesting overfitting are sensitivity to exact market, time period or selection metric. The traditional factors aren't sensitive in that way and so are considered valid models rather than overfitted historical data.
I don't know what you consider the traditional factors, but the return to momentum is significantly sensitive to the selection criteria as I posted earlier.
User avatar
triceratop
Posts: 5838
Joined: Tue Aug 04, 2015 8:20 pm
Location: la la land

Re: Skating where the puck was; e.g., commodities. Factors next?

Post by triceratop »

randomguy wrote: Sat Jan 13, 2018 4:29 pm You know if you stay in place eventually the puck comes back.:) Skating where the puck was isn't that bad as long as you stop when you get there (i.e. don't switch from value to momentum to quality as each posts nice 3 year returns)
I have to admit it's amusing to see people talking so confidently about factors being arb'ed away when this is how value has been performing of late:

Image

source: yardeni
"To play the stock market is to play musical chairs under the chord progression of a bid-ask spread."
selters
Posts: 702
Joined: Thu Feb 27, 2014 8:26 am

Re: Skating where the puck was; e.g., commodities. Factors next?

Post by selters »

The iShares MSCI Diversified Multiple Factor ETFs outperformed cap weighted ETFs in developed markets ex US and in emerging markets in the last 1 and 3 years. This applies both to small caps and large caps. In the US, small caps have outperformed in the last 3 years, but not in the last 1 years. For US large caps, there's been no difference in last 1 and 3 years.

The indexes have also outperformed in the last 5, 10 and 15 years, but that's to be expected, since they're backtested, so I wouldn't put any emphasis on that.

The ETFs have only been on the market for three years, so it's too early to say if factor investing has stopped working. My impression is that is hasn't, at least it seems to be working outside of the US large cap space.

If you take away value and growth funds, less than 1% of ETFs inflows go into factor ETFs. If you take away dividend and minimum volatility ETFs too, the number is closer to 0.5% of inflows. If ETF flows are indicative of how crowded factor investing is, then factor investing isn't crowded at all.
User avatar
Topic Author
CULater
Posts: 2832
Joined: Sun Nov 13, 2016 9:59 am
Location: Hic sunt dracones

Re: Skating where the puck was; e.g., commodities. Factors next?

Post by CULater »

nisiprius wrote: Sat Jan 13, 2018 3:59 pm
CULater wrote: Sat Jan 13, 2018 2:19 pm
nisiprius wrote: Sat Jan 13, 2018 1:11 pm... Factor investors are making intelligent bets founded on rational analysis that certain systematic departures from the market will beat the market...
It's not quite that, IMO. One claim made by factor investors is that you can use factors to tailor your portfolio asset allocation to achieve certain objectives. For example, Larry Swedroe has discussed the notion of investing a smaller equity percentage into SCV (small and value factors), allowing a larger percentage to be invested in treasuries. This might allow the portfolio to achieve the same cumulative return with less volatility. For example, from 2005 - 2013 SCV and TSM had about the same return. But if you invested 30% in SCV + 70% in Intermediate Treasuries, you got the same portfolio return as 60% in TSM + 40% Treasuries but with 2/3 the volatility and a worst drawdown of 12% vs. 29%.

The Small and Value factors didn't beat TSM during this period, but by heavily tilting toward these factors you could have configured your portfolio differently to deliver the same return as a larger allocation to stocks, with consequent lower portfolio volatility and drawdown.
Suppose I amend my remark and say:

"Factor investors are making intelligent bets founded on rational analysis that certain systematic departures from the market will beat the market in risk-adjusted return?"

In the case of the Larry Portfolio, his goal is to hold return constant and reduce risk, hence his goal is superior risk-adjusted return--which he prefers to take in the form of lower risk rather than higher return.
So, this is what Larry has to say, which appears to be in agreement, mostly. The LP is a low-beta, high-tilt portfolio that is designed to achieve the returns of a higher-beta, low-tilt portfolio with reduced risk; e.g., 30/70 SCV tilt has historically matched the returns of 60/40 TSM non-tilted portfolio, but with lower SD, drawdowns, and consequent higher Sharpe.
...the “conventional way” to use the factor exposure, is to increase the expected return of the portfolio. And because these factors have very low correlations to each other, tilted portfolios diversified across factors historically have earned higher risk-adjusted returns.

There is, however, a second way to use “tilting.” Instead of focusing on increasing expected returns, an investor can focus on reducing risk while holding expected returns about the same. This is accomplished by lowering exposure to market beta while at the same time increasing exposure to the size and value premiums. An investor would need less exposure to market beta to achieve the same expected returns because the equities held have a higher expected return than the market portfolio. The lower exposure to market beta is obtained by owning less in stocks and more in bonds. The result is that the portfolio has now become more diversified in terms of its exposure to factors (exposure to market beta went down, while exposure to the other factors went up). In addition, because the allocation to safe bonds increased, the portfolio’s exposure to term risk went up, further diversifying the portfolio across factors.
Berkin, Andrew L.. Your Complete Guide to Factor-Based Investing: The Way Smart Money Invests Today (Kindle Locations 1792-1796). BAM ALLIANCE Press. Kindle Edition.
On the internet, nobody knows you're a dog.
User avatar
BuyAndHoldOn
Posts: 295
Joined: Mon Mar 30, 2015 6:51 pm

Re: Skating where the puck was; e.g., commodities. Factors next?

Post by BuyAndHoldOn »

Many commodities (oil, industrial metals, even gold) are enjoying a resurgences in price. Spot prices are above the futures prices ("backwardation"), which creates positive "roll yield". Trouble is: How long it will continue?

I have read industrial metals are a good bet for this year. But are they a better bet than Global Equities, particularly for the next 10-20 years? Hard to buy and hold with commodities....And I would imagine with certain "factors" as well.
I’d trade it all for a little more | -C Montgomery Burns
staythecourse
Posts: 6993
Joined: Mon Jan 03, 2011 8:40 am

Re: Skating where the puck was; e.g., commodities. Factors next?

Post by staythecourse »

triceratop wrote: Sat Jan 13, 2018 6:05 pm
randomguy wrote: Sat Jan 13, 2018 4:29 pm You know if you stay in place eventually the puck comes back.:) Skating where the puck was isn't that bad as long as you stop when you get there (i.e. don't switch from value to momentum to quality as each posts nice 3 year returns)
I have to admit it's amusing to see people talking so confidently about factors being arb'ed away when this is how value has been performing of late:

Image

source: yardeni
Who knows who is right or wrong, but considering Ben Graham was writing about ideas of value investing in 1930's and is the most well known investor of ALL TIME with many financial professionals following his work (Buffett being just one of them) and in the value premium is STILL around tells me it will exist forever. If it has been there from pre Great Depression what is different now that suggests it won't be there down the road? J

When it comes down to it who in the world wants to own my solo primary doc business if I offered it at the same share price as buying shares of Amazon? NO ONE. Folks will want a discount to persuade them to buy into my practice, no? So I don't see how small or value or both will/ can be arbitraged away. Now the more reasonable argument would be: Despite the factors being present does the delta difference in return vs. market get lost on trading/ taxes/ and other cost to lose any advantage. That may have some legs.

Good luck.
"The stock market [fluctuation], therefore, is noise. A giant distraction from the business of investing.” | -Jack Bogle
thx1138
Posts: 1164
Joined: Fri Jul 12, 2013 2:14 pm

Re: Skating where the puck was; e.g., commodities. Factors next?

Post by thx1138 »

golfCaddy wrote: Sat Jan 13, 2018 5:50 pm
thx1138 wrote: Sat Jan 13, 2018 12:36 pm That doesn't mean there aren't risks of just being back testing especially as folks search for more and more factors to explain total market returns. But the general red flags for plain old backtesting overfitting are sensitivity to exact market, time period or selection metric. The traditional factors aren't sensitive in that way and so are considered valid models rather than overfitted historical data.
I don't know what you consider the traditional factors, but the return to momentum is significantly sensitive to the selection criteria as I posted earlier.
I meant the original Fama-French Three Factors (Large/Small and Growth/Value). They have had the most testing across a wide range of markets and time periods. Since FF3 explains almost 90% of the market returns it shouldn't be surprising any additional factor found starts to look more and more like backtest fitting. That's true of almost any fitting of any data - don't fit with too many parameters.

It is important to note though that this still says nothing about the persistence of the factors after they are discovered. That is also very difficult to prove since just like the original "beta" from CAPM the factors can "underperform" for very long periods, longer than the time since their discovery. So if they are "missing" is that just normal variation or their disappearance entirely because of their published discovery. I was only trying to answer what made the factors different from "Dogs of the Dow" or other backtested rules. No one has ever claimed "Dogs of the Dow" disappeared because it was published. It disappeared because it was backtested nonsense to begin with.

So somewhere in between CAPM and "Dogs of the Dow" is a big grey area...
User avatar
nisiprius
Advisory Board
Posts: 52211
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: Skating where the puck was; e.g., commodities. Factors next?

Post by nisiprius »

staythecourse wrote: Sat Jan 13, 2018 8:04 pmWho knows who is right or wrong, but considering Ben Graham was writing about ideas of value investing in 1930's and is the most well known investor of ALL TIME with many financial professionals following his work (Buffett being just one of them) and in the value premium is STILL around tells me it will exist forever. If it has been there from pre Great Depression what is different now that suggests it won't be there down the road?...
It's not enough for there to be a "premium," if there is also additional risk. In order to be useful, value stocks either need to have a higher risk-adjusted return than the portfolio it is diversifying, or it has to have about the same risk-adjusted return but low correlation--meaningfully low correlation, robust and persistent low correlation, count-on-it low correlation.*

As for Graham, do not forget what he said in 1976:
Q: In selecting the common stock portfolio, do you advise careful study of and selectivity among different issues?

A: In general, no. I am no longer an advocate of elaborate techniques of security analysis in order to find superior value opportunities. This was a rewarding activity, say, 40 years ago, when our textbook "Graham and Dodd" was first published; but the situation has changed a great deal since then. In the old days any well-trained security analyst could do a good professional job of selecting undervalued issues through detailed studies; but in the light of the enormous amount of research now being carried on, I doubt whether in most cases such extensive efforts will generate sufficiently superior selections to justify their cost.
If that isn't "arbitraged away" (or "declined as a result of Graham and Dodd publishing about it"), it's close.

*That's an imprecise statement. A more precise statement is that for a diversifying asset X to increase the Sharpe ratio of portfolio Y, the correlation between X and Y must be lower than min(Sharpe(X), Sharpe(Y)) / max(Sharpe(X), Sharpe(Y)).
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
Topic Author
CULater
Posts: 2832
Joined: Sun Nov 13, 2016 9:59 am
Location: Hic sunt dracones

Re: Skating where the puck was; e.g., commodities. Factors next?

Post by CULater »

When I read about factors, and really investing ideas in general, I get an "Alice in Wonderland" feeling. Factors are real, but they can "disappear" especially when a lot of people are looking at them. Then they can "reappear" when everyone has gone on to other things and forgotten about them. Since they can disappear for long time periods, you have to "believe" they are real in order to reap the benefit of investing in them. If you don't patiently keep the faith, the factor fairies won't lead you to the pot of gold. No, I'm not making this up. You'll be forgiven if you think it sounds more like Harry Potter than financial economics. :D :D :D
On the internet, nobody knows you're a dog.
User avatar
nedsaid
Posts: 19275
Joined: Fri Nov 23, 2012 11:33 am

Re: Skating where the puck was; e.g., commodities. Factors next?

Post by nedsaid »

CULater wrote: Sat Jan 13, 2018 8:26 pm When I read about factors, and really investing ideas in general, I get an "Alice in Wonderland" feeling. Factors are real, but they can "disappear" especially when a lot of people are looking at them. Then they can "reappear" when everyone has gone on to other things and forgotten about them. Since they can disappear for long time periods, you have to "believe" they are real in order to reap the benefit of investing in them. If you don't patiently keep the faith, the factor fairies won't lead you to the pot of gold. No, I'm not making this up. You'll be forgiven if you think it sounds more like Harry Potter than financial economics. :D :D :D
Actually, the act of investing itself is an act of faith. It is belief in human progress, the advance of civilization, economic growth and progress. Pretty much stocks grow because the economy grows over time. Nothing says that stocks have to outperform bonds, they have in the past and likely will do so in the future. Nothing says that the economy has to grow over time or even that human beings will achieve progress in civilization. We do have history and some compelling data to go on but in the final analysis plunking down your hard earned money on investments with uncertain futures does take some faith.
A fool and his money are good for business.
Post Reply